Applications of Big Data and Machine Learning in Smart Agriculture

In agricultural production, rapid and accurate detection of peach blossom bloom plays a crucial role in yield prediction, and is the foundation for automatic thinning. The currently available manual operation-based detection and counting methods are extremely time-consuming and labor-intensive, and are prone to human error. In response to the above issues, this paper proposes a natural environment peach blossom detection model based on the YOLOv5 model. First, a cascaded network is used to add an output layer specifically for small target detection on the basis of the original three output layers. Second, a combined context extraction module (CAM) and feature refinement module (FSM) are added. Finally, the network clusters and statistically analyzes the range of multi-scale channel elements using the K-means++ algorithm, obtaining candidate box sizes that are suitable for the dataset. A novel bounding box regression loss function (SIoU) is used to fuse the directional information between the real box and the predicted box to improve detection accuracy. The experimental results show that, compared with the original YOLOv5s model, our model has correspondingly improved AP values for identifying three different peach blossom shapes, namely, bud, flower, and falling flower, by 7.8%, 10.1%, and 3.4%, respectively, while the final mAP value for peach blossom recognition increases by 7.1%. Good results are achieved in the detection of peach blossom flowering volume. The proposed model provides an effective method for obtaining more intuitive and accurate data sources during the process of peach yield prediction, and lays a theoretical foundation for the development of thinning robots. Full article

The leaf area of pak choi is a critical indicator of growth rate, nutrient absorption, and photosynthetic efficiency, and it is required to be precisely measured for an optimal agricultural output. Traditional methods often fail to deliver the necessary accuracy and efficiency. We propose a method for calculating the leaf area of pak choi based on an improved Mask R-CNN. We have enhanced Mask R-CNN by integrating an advanced attention mechanism and a two-layer fully convolutional network (FCN) into its segmentation branch. This integration significantly improves the model’s ability to detect and segment leaf edges with increased precision. By extracting the contours of reference objects, the conversion coefficient between the pixel area and the actual area is calculated. Using the mask segmentation output from the model, the area of each leaf is calculated. Experimental results demonstrate that the improved model achieves mean average precision (mAP) scores of 0.9136 and 0.9132 in detection and segmentation tasks, respectively, representing improvements of 1.01% and 1.02% over the original Mask R-CNN. The model demonstrates excellent recognition and segmentation capabilities for pak choi leaves. The error between the calculation result of the segmented leaf area and the actual measured area is less than 4.47%. These results indicate that the proposed method provides a reliable segmentation and prediction performance. It eliminates the need for detached leaf measurements, making it suitable for real-life leaf area measurement scenarios and providing valuable support for automated production technologies in plant factories. Full article

Potato diseases and pests have a serious impact on the quality and yield of potatoes, and timely prevention and control of potato diseases and pests is essential. A rich knowledge reserve of potato diseases and pests is one of the most important prevention and control measures; however, valuable knowledge is buried in the massive data of potato diseases and pests, making it difficult for potato growers and managers to obtain and use it in a timely manner and to develop the potential of knowledge. Therefore, this paper explores the construction method of a knowledge graph for automatic knowledge extraction, which extracts the knowledge of potato diseases and pests scattered in heterogeneous data from multiple sources, organises it into a semantically related knowledge base, and provides potato growers with professional knowledge and timely guidance to effectively prevent and control potato diseases and pests. In this paper, a data corpus on potato diseases and pests, called PotatoRE, is first constructed. Then, a model of ALBert-BiLSTM-Self_Att-CRF is designed to extract knowledge from the corpus to form a triplet structure, which is imported into the Neo4j graph database for storage and visualisation. Furthermore, the performance of the model constructed in this paper is compared and verified using the datasets PotatoRE and People’s Daily. The results show that compared to the SOTA models of ALBert BiLSTM-CRF and ALBert BiGRU-CRF, the accuracy of our model has been improved by 2.92% and 3.12%, respectively, using PotatoRE. Compared to the Bert BiLSTM-CRF model on two datasets, our model not only improves the accuracy, recall, and F1 values, but also has a higher efficiency. The model in this paper solves the problem of the difficult recognition of nested entities. On this basis, through comparative experiments, the TransH model is used to effectively represent the constructed knowledge graph, which lays the foundation for achieving inference, extension, and automatic updating of the knowledge base. The achievements of the thesis have made certain contributions to the automatic construction of large-scale knowledge bases. Full article

In apple cultivation, the total nitrogen content is an important indicator of plant growth, fruit quality, and yield. Timely monitoring of growth becomes imperative, since an imbalance, either in deficiency or excess nitrogen, can result in physiological disorders, adversely impacting both the quantity and quality of fruit. Leaf nitrogen content can be determined using simple chlorophyll meters or destructive testing; however, these methods are time-consuming. However, by employing spectral imaging technology, it is possible to swiftly predict leaf nitrogen content. This study estimated the total nitrogen content in apple trees via hyperspectral imaging and machine learning-based regression analysis (partial least-squares regression (PLSR), support vector regression (SVR), and eXtreme gradient boosting regression (XGBoost). Additionally, to reduce computational costs and improve reproducibility, spectral binning was divided into three stages (4, 8, and 16 bins), and models were compared with a 2-binning estimation model. The analysis focused on green, red, red edge, and near-infrared (NIR) spectra, with 5–10 selected wavelengths, and the SVR-based prediction model showed a similar or greater performance to that of the full spectrum. At 4- and 8-binning, the selected wavelengths were similar to those at 2-binning, maintaining similar prediction model performance. However, at 16 bp, the performance of the prediction model decreased owing to spectral data loss, leading to a significant reduction in wavelengths for nitrogen content estimation. These results can support informed nitrogen fertilization decisions, enabling precise, real-time monitoring of nitrogen content for enhanced plant growth, fruit quality, and yield in apple trees. Additionally, the selected wavelengths can be considered in the development of new types of multispectral sensors. Full article

The absence of accurate measurement or calculation techniques for crop water requirements in greenhouses frequently results in over- or under-irrigation. In order to find a better method, this study analyzed the accuracy, data consistency and practicability of the Penman–Monteith (PM), Hargreaves–Samani (HS), Pan Evaporation (PAN), and Artificial Neural Network (ANN) models. Model-calculated crop evapotranspiration (ET_C) was compared with lysimeter-measured crop evapotranspiration (ET_C) in the National Precision Agriculture Demonstration Station in Beijing, China. The results showed that the actual ET_C over the entire experimental period was 176.67 mm. The ET_C calculated with the PM, HS, PAN, and ANN model were 146.07 mm, 189.45 mm, 197.03 mm, and 174.7 mm, respectively, which were different from the actual value by −17.32%, 7.23%, 11.52%, and −1.12%, respectively. The order of the calculation accuracy for the four models is as follows: ANN model > PAN model > PM model > HS model. By comprehensively evaluating the statistical indicators of each model, the ANN model was found to have a significantly higher calculation accuracy compared to the other three models. Therefore, the ANN model is recommended for estimating ET_C under greenhouse conditions. The PM and PAN models can also be used after improvement. Full article

The application of data science (DS) techniques has become increasingly essential in various fields, including epidemiology and climatology in agricultural production systems. In this sector, traditionally large amounts of data are acquired, but not well-managed and -analyzed as a basis for evidence-based decision-making processes. Here, we present a comprehensive step-by-step guide that explores the use of DS in managing epidemiological and climatological data within rice production systems under tropical conditions. Our work focuses on using the multi-temporal dataset associated with the monitoring of diseases and climate variables in rice in Colombia during eight years (2012–2019). The study comprises four main phases: (I) data cleaning and organization to ensure the integrity and consistency of the dataset; (II) data management involving web-scraping techniques to acquire climate information from free databases, like WordClim and Chelsa, validation against in situ weather stations, and bias removal to enrich the dataset; (III) data visualization techniques to effectively represent the gathered information, and (IV) a basic analysis related to the clustering and climatic characterization of rice-producing areas in Colombia. In our work, a process of evaluation and the validation of climate data are conducted based on errors (r, R^2, MAE, RSME) and bias evaluation metrics. In addition, in phase II, climate clustering was conducted based on a PCA and K-means algorithm. Understanding the association of climatic and epidemiological data is pivotal in predicting and mitigating disease outbreaks in rice production areas. Our research underscores the significance of DS in managing epidemiological and climatological data for rice production systems. By applying a protocol responsible for DS tools, our study provides a solid foundation for further research into disease dynamics and climate interactions in rice-producing regions and other crops, ultimately contributing to more informed decision-making processes in agriculture. Full article

Quality assurance through visual inspection plays a pivotal role in agriculture. In recent years, deep learning techniques (DL) have demonstrated promising results in object recognition. Despite this progress, few studies have focused on assessing human visual inspection and DL for defect identification. This study aims to evaluate visual human inspection and the suitability of using DL for defect identification in products of the floriculture industry. We used a sample of defective and correct decorative wreaths to conduct an attribute agreement analysis between inspectors and quality standards. Additionally, we computed the precision, accuracy, and Kappa statistics. For the DL approach, a dataset of wreath images was curated for training and testing the performance of YOLOv4-tiny, YOLOv5, YOLOv8, and ResNet50 models for defect identification. When assessing five classes, inspectors showed an overall precision of 92.4% and an accuracy of 97%, just below the precision of 93.8% obtained using YOLOv8 and YOLOv5 with accuracies of 99.9% and 99.8%, respectively. With a Kappa value of 0.941, our findings reveal an adequate agreement between inspectors and the standard. The results evidence that the models presented a similar performance to humans in terms of precision and accuracy, highlighting the suitability of DL in assisting humans with defect identification in artisanal-made products from floriculture. Therefore, by assisting humans with digital technologies, organizations can embrace the full potential of Industry 4.0, making the inspection process more intelligent and reliable. Full article

Excessive total nitrogen (TN) content in topsoil is a major cause of eutrophication when nitrogen flows into water systems from soil losses. Therefore, TN content prediction is essential for establishing topsoil management systems and protecting aquatic ecosystems. Recently, hyperspectral imaging (HSI) has been used as a rapid, nondestructive technique for quantifying various soil properties. This study developed a machine and deep learning-based model using hyperspectral imaging to rapidly measure TN contents. A total of 139 topsoil samples were collected from the four major rivers in the Republic of Korea. Visible-to-near-infrared (VNIR) and near-infrared (NIR) hyperspectral imaging data were acquired in the 400–1000 nm and 895–1720 nm ranges, respectively. Prediction models for predicting the TN content in the topsoil were developed using partial least square regression (PLSR) and one-dimensional convolutional neural networks (1D-CNNs). From the total number of pixels in each topsoil sample, 12.5, 25, and 50% of the pixels were randomly selected, and the data were augmented 10 times to improve the performance of the 1D-CNN model. The performances of the models were evaluated by estimating the coefficients of determination (R²) and root mean squared errors (RMSE). The R_p² values of the optimal PLSR (with maximum normalization preprocessing) and 1D-CNN (with SNV preprocessing) models were 0.72 and 0.92, respectively. Therefore, HSI can be used to estimate TN content in topsoil and build a topsoil database to develop conservation strategies. Full article

The respiratory status of dairy cows can reflect their heat stress and health conditions. It is widely used in the precision farming of dairy cows. To realize intelligent monitoring of cow respiratory status, a system based on infrared thermography was constructed. First, the YOLO v8 model was used to detect and track the nose of cows in thermal images. Three instance segmentation models, Mask2Former, Mask R-CNN and SOLOv2, were used to segment the nostrils from the nose area. Second, the hash algorithm was used to extract the temperature of each pixel in the nostril area of a cow to obtain the temperature change curve. Finally, the sliding window approach was used to detect the peaks of the filtered temperature curve to obtain the respiratory rate of cows. Totally 81 infrared thermography videos were used to test the system, and the results showed that the AP₅₀ of nose detection reached 98.6%, and the AP₅₀ of nostril segmentation reached 75.71%. The accuracy of the respiratory rate was 94.58%, and the correlation coefficient R was 0.95. Combining infrared thermography technology with deep learning models can improve the accuracy and usability of the respiratory monitoring system for dairy cows. Full article

Weeds have a significant impact on the growth of rice. Accurate information about weed infestations can provide farmers with important information to facilitate the precise use of chemicals. In this study, we utilized visible light images captured by UAVs to extract information about weeds in areas of two densities on farmland. First, the UAV images were segmented using an optimal segmentation scale, and the spectral, texture, index, and geometric features of each segmented object were extracted. Cross-validation and recursive feature elimination techniques were combined to reduce the dimensionality of all features to obtain a better feature set. Finally, we analyzed the extraction effect of different feature dimensions based on the random forest (RF) algorithm to determine the best feature dimensions, and then we further analyzed the classification result of machine learning algorithms, such as random forest, support vector machine (SVM), decision tree (DT), and K-nearest neighbors (KNN) and compared them based on the best feature dimensions. Using the extraction results of the best classifier, we created a zoning map of the weed infestations in the study area. The results indicated that the best feature subset achieved the highest accuracy, with respective overall accuracies of 95.38% and 91.33% for areas with dense and sparse weed densities, respectively, and F1-scores of 94.20% and 90.57. Random forest provided the best extraction results for each machine learning algorithm in the two experimental areas. When compared to the other algorithms, it improved the overall accuracy by 1.74–12.14% and 7.51–11.56% for areas with dense and sparse weed densities, respectively. The F1-score improved by 1.89–17.40% and 7.85–10.80%. Therefore, the combination of object-based image analysis (OBIA) and random forest based on UAV remote sensing accurately extracted information about weeds in areas with different weed densities for farmland, providing effective information support for weed management. Full article

Data processing of low-altitude remote sensing visible images from UAVs is one of the hot research topics in precision agriculture aviation. In order to solve the problems of large model size with slow detection speed that lead to the inability to process images in real time, this paper proposes a lightweight target detector CURI-YOLOv7 based on YOLOv7tiny which is suitable for individual citrus tree detection from UAV remote sensing imagery. This paper augmented the dataset with morphological changes and Mosica with Mixup. A backbone based on depthwise separable convolution and the MobileOne-block module was designed to replace the backbone of YOLOv7tiny. SPPF (spatial pyramid pooling fast) was used to replace the original spatial pyramid pooling structure. Additionally, we redesigned the neck by adding GSConv and depth-separable convolution and deleted its input layer from the backbone with a size of (80, 80) and its output layer from the head with a size of (80, 80). A new ELAN structure was designed, and the redundant convolutional layers were deleted. The experimental results show that the GFLOPs = 1.976, the parameters = 1.018 M, the weights = 3.98 MB, and the mAP = 90.34% for CURI-YOLOv7 in the UAV remote sensing imagery of the citrus trees dataset. The detection speed of a single image is 128.83 on computer and 27.01 on embedded devices. Therefore, the CURI-YOLOv7 model can basically achieve the function of individual tree detection in UAV remote sensing imagery on embedded devices. This forms a foundation for the subsequent UAV real-time identification of the citrus tree with its geographic coordinates positioning, which is conducive to the study of precise agricultural management of citrus orchards. Full article

Photosynthesis is one of the key issues for vertical cultivation in plant factories, and efficient natural sunlight utilization requires predicting the light falling on each seedbed in a real-time manner. However, public weather services neither provide sunshine data nor meet spatial resolution requirement. Facing these short-term and small-area weather forecasting challenges, we propose a cross-scale approach to infer seedbed-sized areas of sunshine from the city-level public weather services, and then design a seedbed rotation scheduling system for optimal natural sunlight utilization. First, an end-edge-cloud coordinated computing architecture was employed to concurrently aggregate the multi-scale data from weather satellites to sunshine sensors in the plant factory. Second, the small area of sunshine deterministically depends on the meteorological data given a fixed environment, and this correlation was described by a hybrid mapping model, which combined the long short-term memory (LSTM) and gradient boosting decision tree (GBDT) algorithms to form the LSTM-GBDT hybrid prediction algorithm (LGHPA). By training the LGHPA with historical local sensory sunshine and the city-scale meteorological data, the hourly sunshine on a seedbed can be predicted from the public weather forecasting service. Finally, a dynamic seedbed scheduling scheme was constructed to provide uniform solar energy absorption according to the one-hour-ahead radiation estimation. Experiment results show that the hourly sunshine prediction error was less than 18.44% over a seasonal period and the deviation for different solar absorption by seedbeds with rotation capability is less than 7.1%. Consequently, it was demonstrated that the application of short-term, small-area sunshine forecasting improved the performance of seedbed rotation for uniformly absorbed solar radiation. The proposed method verifies the feasibility of precisely predicting small-area sunshine down to the seedbed scale by leveraging a model-based approach and a cloud-edge-end merged cybernetic computing paradigm. Full article

This study proposes an improved link prediction model for predicting the “suitable for people” relationship within the knowledge graph of tea. The relationships between various types of tea and suitable target groups have yet to be fully explored, and the existing InteractE model still does not adequately capture a portion of the complex information around the interactions between entities and relationships. In this study, we integrate SENet into the feature layer of the InteractE model to enhance the capturing of helpful information in the feature channels. Additionally, the GCN layer is employed as the encoder, and the SENet-integrated InteractE model is used as the decoder to further capture the neighbour node information in the knowledge graph. Furthermore, our proposed improved model demonstrates significant improvements compared to several standard models, including the original model from public datasets (WN18RR, Kinship). Finally, we construct a tea dataset comprising 6698 records, including 330 types of tea and 29 relationship types. We predict the “suitable for people” relationship in the tea dataset through transfer learning. When comparing our model with the original model, we observed an improvement of 1.4% in H@10 for the WN18RR dataset, a 7.6% improvement in H@1 for the Kinship dataset, and a 5.2% improvement in MRR. Regarding the tea dataset, we achieved a 4.1% increase in H@3 and a 2.5% increase in H@10. This study will help to fully exploit the value potential of tea varieties and provide a reference for studies assessing healthy tea drinking. Full article

Automatic weight detection is an essential step in the factory production of Pleurotus eryngii. In this study, a data set containing 1154 Pleurotus eryngii images was created, and then machine vision technology was used to extract eight two-dimensional features from the images. Because the fruiting bodies of Pleurotus eryngii have different shapes, these features were less correlated with weight. This paper proposed a multidimensional feature derivation method and an Attention-Based CNN model to solve this problem. This study aimed to realize the traditional feature screening task by deep learning algorithms and built an estimation model. Compared with different regression algorithms, the R², RMSE, MAE, and MAPE of the Attention-Based CNN were 0.971, 7.77, 5.69, and 5.87%, respectively, and showed the best performance. Therefore, it can be used as an accurate, objective, and effective method for automatic weight measurements of Pleurotus eryngii. Full article

The internal temperature of the pigsty has a great impact on the pigs. Keeping the temperature in the pigsty within a certain range is a pressing problem in environmental control. The current pigsty temperature regulation method is based mainly on manual and simple automatic control. There is rarely intelligent control, and such direct methods have problems such as low control accuracy, high energy consumption and untimeliness, which can easily lead to the occurrence of heat stress conditions. Therefore, this paper proposed an improved sparrow search algorithm (ISSA) based on a multi-strategy improvement to optimize the least squares support vector machine (LSSVM) to form a pigsty temperature prediction model. In the optimization process of the sparrow search algorithm (SSA), the initial position of the sparrow population was first generated by using the reverse good point set; secondly, the population number update formula was proposed to automatically adjust the number of discoverers and followers based on the number of iterations to improve the search ability of the algorithm; finally, the adaptive t-distribution was applied to the discoverer position variation to refine the discoverer population and further improve the search ability of the algorithm. Tests were conducted using 23 benchmark functions, and the results showed that ISSA outperformed SSA. By comparing it with the LSSVM models optimized by four standard algorithms, the prediction effect of the ISSA-LSSVM model was tested. In the end, the ISSA-LSSVM temperature prediction model had MSE of 0.0766, MAE of 0.2105, and R² of 0.9818. The results showed that the proposed prediction model had the best prediction performance and prediction accuracy, and can provide accurate data support for the prediction and control of the internal temperature of the pigsty. Full article

The total nonstructural carbohydrate (TNC) concentration is an important indicator of the growth period and health of fruit trees. Remote sensing can be applied to monitor the TNC concentration in crops in a non-destructive manner. In this study, hyperspectral imaging from an unmanned aerial vehicle was applied to estimate the TNC concentration in apple trees. Partial least-squares regression, ridge regression, and Gaussian process regression (GP) were used to develop estimation models, and their effectiveness using selected key bands as opposed to full bands was evaluated in an effort to reduce computational costs and improve reproducibility. Nine key bands were identified, and the GP-based model using these key bands performed almost as well as the models using full bands. These results can be combined with previous studies on estimating the nitrogen concentration to provide useful information for more precise nutrient management to improve the yield and quality of apple trees. Full article

Crop nutrient biochemical information (mainly including chlorophyll class and nutrient elements mainly nitrogen, phosphorus and potassium) is an important basis for revealing crop growth and development patterns and their relationship with the environment. Hyperspectral technology has been rapidly developed and applied in crop nutrient biochemical information monitoring research. This paper firstly describes the theoretical basis of hyperspectral technology for monitoring crop nutrients and biochemical information. Then, the research progress of hyperspectral technology in monitoring nutrient and biochemical information of crops in different growth periods or different growth environments is outlined. Meanwhile, the shortcomings of the current technology in these research directions and the future research trends are discussed. Finally, the modeling methods for building crop nutrient biochemical information monitoring models by applying hyperspectral data are systematically outlined. And the effects of different spectral pre-processing methods, spectral effective information extraction methods and modeling algorithms on the accuracy of monitoring models are analyzed. On this basis, the challenges and prospects of hyperspectral technology in monitoring crop nutrient biochemical information are presented, aiming to provide relevant theoretical basis and technical reference for the research related to monitoring and inversion of crop physiological parameters based on hyperspectral technology. Full article

In image classification, few-shot learning deals with recognizing visual categories from a few tagged examples. The degree of expressiveness of the encoded features in this scenario is a crucial question that needs to be addressed in the models being trained. Recent approaches have achieved encouraging results in improving few-shot models in deep learning, but designing a competitive and simple architecture is challenging, especially considering its requirement in many practical applications. This work proposes an improved few-shot model based on a multi-layer feature fusion (FMLF) method. The presented approach includes extended feature extraction and fusion mechanisms in the Convolutional Neural Network (CNN) backbone, as well as an effective metric to compute the divergences in the end. In order to evaluate the proposed method, a challenging visual classification problem, maize crop insect classification with specific pests and beneficial categories, is addressed, serving both as a test of our model and as a means to propose a novel dataset. Experiments were carried out to compare the results with ResNet50, VGG16, and MobileNetv2, used as feature extraction backbones, and the FMLF method demonstrated higher accuracy with fewer parameters. The proposed FMLF method improved accuracy scores by up to 3.62% in one-shot and 2.82% in five-shot classification tasks compared to a traditional backbone, which uses only global image features. Full article

Existing deep learning methods usually adopt deeper and wider network structures to achieve better performance. However, we found that this rule does not apply well to crop disease identification tasks, which inspired us to rethink the design paradigm of disease identification models. Crop diseases belong to fine-grained features and lack obvious patterns. Deeper and wider network structures will cause information loss of features, which will damage identification efficiency. Based on this, this paper designs a very lightweight disease identification network called VLDNet. The basic module VLDBlock of VLDNet extracts intrinsic features through 1 × 1 convolution, and uses cheap linear operations to supplement redundant features to improve feature extraction efficiency. In inference, reparameterization technology is used to further reduce the model size and improve inference speed. VLDNet achieves state-of-the-art model (SOTA) latency-accuracy trade-offs on self-built and public datasets, such as equivalent performance to Swin-Tiny with a parameter size of 0.097 MB and 0.04 G floating point operations (FLOPs), while reducing parameter size and FLOPs by 297 times and 111 times, respectively. In actual testing, VLDNet can recognize 221 images per second, which is far superior to similar accuracy models. This work is expected to further promote the application of deep learning-based crop disease identification methods in practical production. Full article

Seed treatment with micronutrients is a crucial strategy for providing early seedling supply during development, and is commonly employed in soybean cultivation. However, responses to micronutrient treatment may vary based on seed vigor levels. Therefore, this study aimed to assess the potential of hyperspectral imaging combined with preprocessing and machine learning, compared to X-ray fluorescence spectroscopy, in evaluating the dynamics of micronutrient uptake during the germination of soybean seeds with varying levels of vigor. Two seed lots with differing levels of vigor were utilized for the analysis. The absorption of micronutrients by the seeds was evaluated using X-ray fluorescence spectroscopy (XRF), microprobe X-ray fluorescence spectroscopy (μ-XRF) and hyperspectral imaging (HSI) in two regions of interest (cotyledons and the embryonic axis). Artificial neural network (ANN), decision tree (DT) and partial least squares–discriminant analysis (PLS-DA) classification models, along with the Savitzky–Golay (SG), standard normal variation (SNV) and multiplicative scatter correction (MSC) methods, were employed to determine seed vigor based on the impact of micronutrient treatment. XRF identified higher concentrations of micronutrients in the treated seeds, with zinc being the predominant element. μ-XRF analysis revealed that a significant proportion of the micronutrients remained adhered to the hilum and seed coat, irrespective of seed vigor. The PLS-DA classification model using spectral data exhibited higher accuracy in classifying soybean seeds with high and low vigor, regardless of seed treatment with micronutrients and the analyzed region. Full article

Accurately estimating aboveground dry biomass (ADB) is crucial. The ADB of rice has primarily been estimated using vegetation indices with several discrete bands; nevertheless, these indices cannot take advantage of continuous bands available with hyperspectral remote sensing. This study analyzed the quantitative relationship between canopy hyperspectral characteristic parameters (HCPs) and the ADB of rice. Twenty HCPs were used, including red edge area (SDr), blue edge area (SDb), and others. The variable-screening methods involved stepwise regression (SR), a regression coefficient (RC), variable importance in projection (vip), and random forest (RF). Stepwise and partial least squares regression methods were employed with traditional linear regression as well as machine learning methods including random forest (RF), a support vector machine (SVM), a BP artificial neural network (BPNN), and an extreme learning machine. Whole- and screening-variable models were constructed to estimate rice ADB at jointing, booting, heading, and maturing stages and across growth stages. Screening-variable models include SVM models based on SR (SVM-sr), RF models based on vip (RF-vip), and others. The results show that the HCPs had a significant correlation with ADB containing elements in the red edge region, namely SDr, SDr/SDb, and (SDr − SDb)/(SDr + SDb) at each growth stage. In addition, the screening performance of vip and SR was better than that of RC and RF, and fewer variables were screened. Moreover, the HCPs of the red edge region were screened using different screening methods at each growth stage. Among them, SDr/SDb and (SDr − SDb)/(SDr + SDb) appeared frequently, indicating they are important. Furthermore, at each growth stage, ADB could be well-estimated using diverse models with the RF modeling method based on vip screening variables found to be the best modeling method for ADB estimation; the independent variables of the RF-vip model involved the (SDr − SDb)/(SDr + SDb) at each growth stage. Full article

In recent years, the domain of diagnosing plant afflictions has predominantly relied upon the utilization of deep learning techniques for classifying images of diseased specimens; however, these classification algorithms remain insufficient for instances where a single plant exhibits multiple ailments. Consequently, we view the region afflicted by the malady of rice leaves as a minuscule issue of target detection, and then avail ourselves of a computational approach to vision to identify the affected area. In this paper, we advance a proposal for a Dense Higher-Level Composition Feature Pyramid Network (DHLC-FPN) that is integrated into the Detection Transformer (DETR) algorithm, thereby proffering a novel Dense Higher-Level Composition Detection Transformer (DHLC-DETR) methodology which can effectively detect three diseases: sheath blight, rice blast, and flax spot. Initially, the proposed DHLC-FPN is utilized to supersede the backbone network of DETR through amalgamation with Res2Net, thus forming a feature extraction network. Res2Net then extracts five feature scales, which are coalesced through the deployment of high-density rank hybrid sampling by the DHLC-FPN architecture. The fused features, in concert with the location encoding, are then fed into the transformer to produce predictions of classes and prediction boxes. Lastly, the prediction classes and the prediction boxes are subjected to binary matching through the application of the Hungarian algorithm. On the IDADP datasets, the DHLC-DETR model, through the utilization of data enhancement, elevated mean Average Precision (mAP) by 17.3% in comparison to the DETR model. Additionally, mAP for small target detection was improved by 9.5%, and the magnitude of hyperparameters was reduced by 324.9 M. The empirical outcomes demonstrate that the optimized structure for feature extraction can significantly enhance the average detection accuracy and small target detection accuracy of the model, achieving an average accuracy of 97.44% on the IDADP rice disease dataset. Full article

Rapidly increasing numbers of the plastic-covered greenhouse (PCG) worldwide ensure food security but threaten environmental security; thus, accurate monitoring of the spatiotemporal pattern in plastic-covered greenhouses (PCGs) is necessary for modern agricultural management and environmental protection. However, many urgent issues still exist in PCG mapping, such as multi-source data combination, classification accuracy improvement, spatiotemporal scale expansion, and dynamic trend quantification. To address these problems, this study proposed a new framework that progressed layer by layer from multi-feature scenario construction, classifier and feature scenario preliminary screening, feature optimization, and spatiotemporal mapping, to rapidly identify large-scale PCGs by integrating multi-source data using Google Earth Engine (GEE), and the framework was first applied to Central Yunnan Province (CYP), where PCGs are concentrated but no relevant research exists. The results suggested that: (1) combining the random forest (RF) classifier and spectrum (S) + backscatter (B) + index (I) + texture (T) + terrain (Tr) feature scenario produced the highest F-score (95.60%) and overall accuracy (88.04%). (2) The feature optimization for the S + I + T + B + Tr scenario positively impacted PCG recognition, increasing the average F-score by 1.03% (96.63% vs. 95.60%). (3) The 6-year average F-score of the PCGs extracted by the combined RF algorithm and the optimal feature subset exceeded 95.00%, and its spatiotemporal mapping results indicated that PCGs were prominently agglomerated in the central CYP and continuously expanded by an average of 65.45 km²/yr from 2016 to 2021. The research reveals that based on the GEE platform, multi-source data can be integrated through a feature optimization algorithm to more efficiently map PCG spatiotemporal information in complex regions. Full article

Satellite hyperspectral imagery is an important data source for large-scale refined land cover classification and mapping, but the high spatial heterogeneity and spectral variability at low spatial resolution and the high computation cost for massive data remain challenges in the research community. In recent years, convolutional neural network (CNN) models with the capability for feature extraction have been widely used in hyperspectral image classification. However, incomplete feature extraction, inappropriate feature fusion, and high time consumption are still the major problems for CNN applications in large-scale fine land cover mapping. In this study, a Spatial-Convolution Spectral-Transformer Interactive Network (SCSTIN) was proposed to integrate 2D-CNN and Transformer into a dual-branch network to enhance feature extraction capabilities by exploring spatial context information and spectral sequence signatures in a targeted manner. In addition, spatial-spectral interactive fusion (SSIF) units and category-adaptive weighting (CAW) as two feature fusion modules were also adopted between and after the two feature extraction branches to improve efficiency in feature fusion. The ZY1-02D hyperspectral imagery was collected to conduct the experiments in the study area of the eastern foothills of the Helan Mountains (EFHLM), covering an area of about 8800 km², which is the largest hyperspectral dataset as far as we know. To explore the potential of the proposed network in terms of accuracy and efficiency, SCSTIN models with different depths (SCSTIN-4 and SCSTIN-2) were performed. The results suggest that compared with the previous eight advanced hyperspectral image classifiers, both SCSTIN models achieved satisfactory performance in accuracy and efficiency aspects with low complexity, where SCSTIN-4 achieved the highest accuracy and SCSTIN-2 obtained higher efficiency. Accordingly, the SCSTIN models are reliable for large-scale fast refined land cover classification and mapping. In addition, the spatial distribution pattern of diverse ground objects in EFHLM is also analyzed. Full article

Many large dams built on the Çoruh River have resulted in the inundation of olive groves in Artvin Province, Turkey. This research sets out to identify suitable locations for olive cultivation in Artvin using the random forest (RF) algorithm. A total of 575 plots currently listed in the Farmer Registration System, where olive cultivation is practiced, were used as inventory data in the training and validation of the RF model. In order to determine the areas where olive cultivation can be carried out, a land suitability map was created by taking into account 10 parameters including the average annual temperature, average annual precipitation, slope, aspect, land use capability class, land use capability sub-class, soil depth, other soil properties, solar radiation, and land cover. According to this map, an area of 53,994.57 hectares was detected as suitable for olive production within the study region. To validate the created model, the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) were utilized. As a result, the AUC value was determined to be 0.978, indicating that the RF method may be successfully used in determining suitable lands for olive cultivation in particular, as well as crop-based land suitability research in general. Full article

Unmanned aerial vehicles (UAVs) are increasingly being integrated into the domain of precision agriculture, revolutionizing the agricultural landscape. Specifically, UAVs are being used in conjunction with machine learning techniques to solve a variety of complex agricultural problems. This paper provides a careful survey of more than 70 studies that have applied machine learning techniques utilizing UAV imagery to solve agricultural problems. The survey examines the models employed, their applications, and their performance, spanning a wide range of agricultural tasks, including crop classification, crop and weed detection, cropland mapping, and field segmentation. Comparisons are made among supervised, semi-supervised, and unsupervised machine learning approaches, including traditional machine learning classifiers, convolutional neural networks (CNNs), single-stage detectors, two-stage detectors, and transformers. Lastly, future advancements and prospects for UAV utilization in precision agriculture are highlighted and discussed. The general findings of the paper demonstrate that, for simple classification problems, traditional machine learning techniques, CNNs, and transformers can be used, with CNNs being the optimal choice. For segmentation tasks, UNETs are by far the preferred approach. For detection tasks, two-stage detectors delivered the best performance. On the other hand, for dataset augmentation and enhancement, generative adversarial networks (GANs) were the most popular choice. Full article

The aim of this study was to evaluate and compare the performance of multivariate classification algorithms, specifically Partial Least Squares Discriminant Analysis (PLS-DA) and machine learning algorithms, in the classification of Monthong durian pulp based on its dry matter content (DMC) and soluble solid content (SSC), using the inline acquisition of near-infrared (NIR) spectra. A total of 415 durian pulp samples were collected and analyzed. Raw spectra were preprocessed using five different combinations of spectral preprocessing techniques: Moving Average with Standard Normal Variate (MA+SNV), Savitzky–Golay Smoothing with Standard Normal Variate (SG+SNV), Mean Normalization (SG+MN), Baseline Correction (SG+BC), and Multiplicative Scatter Correction (SG+MSC). The results revealed that the SG+SNV preprocessing technique produced the best performance with both the PLS-DA and machine learning algorithms. The optimized wide neural network algorithm of machine learning achieved the highest overall classification accuracy of 85.3%, outperforming the PLS-DA model, with overall classification accuracy of 81.4%. Additionally, evaluation metrics such as recall, precision, specificity, F1-score, AUC ROC, and kappa were calculated and compared between the two models. The findings of this study demonstrate the potential of machine learning algorithms to provide similar or better performance compared to PLS-DA in classifying Monthong durian pulp based on DMC and SSC using NIR spectroscopy, and they can be applied in the quality control and management of durian pulp production and storage. Full article

The segmentation and positioning of tea buds are the basis for intelligent picking robots to pick tea buds accurately. Tea images were collected in a complex environment, and median filtering was carried out to obtain tea bud images with smooth edges. Four semantic segmentation algorithms, U-Net, high-resolution network (HRNet_W18), fast semantic segmentation network (Fast-SCNN), and Deeplabv3+, were selected for processing images. The centroid of the tea buds and the image center of the minimum external rectangle were calculated. The farthest point from the centroid was extracted from the tea stalk orientation, which was the final picking point for tea buds. The experimental results showed that the mean intersection over union (mIoU) of HRNet_W18 was 0.81, and for a kernel with a median filter size of 3 × 3, the proportion of abnormal tea buds was only 11.6%. The average prediction accuracy of picking points with different tea stalk orientations was 57%. This study proposed a fresh tea bud segmentation and picking point location method based on a high-resolution network model. In addition, the cloud platform can be used for data sharing and real-time calculation of tea bud coordinates, reducing the computational burden of picking robots. Full article

Canola and wheat are the main oilseed crop and grain crop, respectively, and they often have similar phenological stages. The understanding of the interactions between microwave signals with wheat and canola in different stages is important for their monitoring using synthetic aperture radar (SAR) imagery. This paper investigates the characteristics of canola and wheat through the use of backscattering profiles from multi-year Sentinel-1 images. Large fluctuations are observed for the temporal backscattering profiles of canola and wheat in different growth statuses induced by agrometeorological conditions in different years. The capability and stability of Sentinel-1 for wheat and canola mapping is further investigated using single- and multi-temporal SAR images hosted in Google Earth Engine (GEE) using the random forest classifier. Although different agrometeorological conditions and field managements make the temporal profiles of backscattering variations, the large difference in canopy structure allows SAR images to make the separability of canola and wheat stable on Sentinel-1 images in different phenology stages. The classification accuracies and the feature importance scores from multi-temporal classification in different years show that the backscattering features obtained at flowering to maturity stages make more contributions to the good-quality mapping of canola and wheat than those at other stages. The F1 scores of canola and wheat achieve 0.95 during the canola flowering and podding period, and the minimum F1 scores of 0.85 were also obtained at other stages. These findings show that SAR images have great potential in the good-quality mapping of canola and wheat in a wide phenology window. Full article

Wheat yellow mosaic disease is a low-temperature and soil-borne disease. Crop infection by the yellow mosaic virus can lead to severe yield and economic losses. It is easily confused with nitrogen deficiency based on the plant’s morphological characteristics. Timely disease detection and crop management in the field require the precise identification of crop stress types. However, the detection of crop stress is often underappreciated. Wheat nitrogen deficiency and yellow mosaic disease were investigated in the field and wheat physiological and biochemical experiments were conducted to collect agronomic indicators, four years of reflectance spectral data at green-up and jointing were collected, and then studies for the detection of nitrogen deficiency and yellow mosaic disease stresses were carried out. The continuous removal (CR), first-order derivative (FD), standard normal variate (SNV), and spectral separation of soil and vegetation (3SV) preprocessing methods and 96 spectral indices were evaluated. The threshold method and variance inflation factor (TVIF) were used as feature selection methods combined with machine learning to develop a crop stress detection method. The results show that the most sensitive wavelengths are found in the 725–1000 nm region, while the sensitivity of the spectrum in the 400–725 nm region is lower. The PRI_670,570, B, and RARSa spectral indices can detect nitrogen deficiency and yellow leaf disease stress, and the OA and Kappa values are 93.87% and 0.873, respectively, for PRI_670,570, which is the best index. A 3SV-TVIF-SVM stress detection method was then proposed, using OA and Kappa values of 96.97% and 0.931, respectively, for field data validation. The results of the study can provide technical support and a theoretical basis for the accurate control of yellow mosaic disease and nitrogen fertilizer management in the field. Full article

Soil erosion triggered by water and wind pose a great threat to the sustainable development of Pakistan. In this study, a combination of geographic information systems (GISs) and machine learning approaches were used to predict soil water erosion rates. The Revised Wind Erosion Equation (RWEQ) model was used to evaluate soil wind erosion, map erosion factors, and analyze the soil erosion rates for each land use type. Finally, the maps of soil water and wind erosion were spatially integrated to identify erosion risk regions and recommend land use management in Pakistan. According to our estimates, the Potohar Plateau and its surrounding regions were mostly impacted by water erosion and have a soil erosion rate of 2500–5000 t·km⁻²·a⁻¹; on the other hand, wind erosion predominated the Kharan Desert and the Thar Desert, with a soil erosion rate exceeding 15,000 t·km⁻²·a⁻¹. The Sulaiman and Kirthar Mountain Ranges were susceptible to wind–water compound erosion, which was more than 8000 t·km⁻²·a⁻¹. This study offers new perspectives on the geographic pattern of individual and integrated water–wind erosion threats in Pakistan and provides high-precision data and a scientific foundation for designing rational soil and water conservation practices. Full article

Artificial intelligence, specifically machine learning (ML), serves as a valuable tool for decision support in crop management under ongoing climate change. However, ML implementation to predict maize yield is still limited in Central Europe, especially in Hungary. In this context, we assessed the performance of four ML algorithms (Bagging (BG), Decision Table (DT), Random Forest (RF) and Artificial Neural Network-Multi Layer Perceptron (ANN-MLP)) in predicting maize yield based on four different input scenarios. The collected data included both agricultural data (production (PROD) (ton) and maize cropped area (AREA) (ha)) and climate data (annual mean temperature °C (Tmean), precipitation (PRCP) (mm), rainy days (RD), frosty days (FD) and hot days (HD)). This research adopted four scenarios, as follows: SC1: AREA+ PROD+ Tmean+ PRCP+ RD+ FD+ HD; SC2: AREA+ PROD; SC3: Tmean+ PRCP+ RD+ FD+ HD; and SC4: AREA+ PROD+ Tmean+ PRCP. In the training stage, ANN-MLP-_SC1 and ANN-MLP-_SC4 outperformed other ML algorithms; the correlation coefficient (r) was 0.99 for both, while the root mean squared errors (RMSEs) were 107.9 (ANN-MLP-_SC1) and 110.7 (ANN-MLP-_SC4). In the testing phase, the ANN-MLP-_SC4 had the highest r value (0.96), followed by ANN-MLP-_SC1 (0.94) and RF-_SC2 (0.94). The 10-fold cross validation also revealed that the ANN-MLP-_SC4 and ANN-MLP-_SC1 have the highest performance. We further evaluated the performance of the ANN-MLP-_SC4 in predicting maize yield on a regional scale (Budapest). The ANN-MLP-_SC4 succeeded in reaching a high-performance standard (r = 0.98, relative absolute error = 21.87%, root relative squared error = 20.4399% and RMSE = 423.23). This research promotes the use of ANN as an efficient tool for predicting maize yield, which could be highly beneficial for planners and decision makers in developing sustainable plans for crop management. Full article

Phosphorus (P) is a macronutrient that often limits the productivity and growth of terrestrial ecosystems, but it is also one of the main causes of eutrophication in aquatic systems at both local and global levels. P content in soils can vary largely, but usually, only a small fraction is plant-available or in an organic form for biological utilization because it is bound in incompletely weathered mineral particles or adsorbed on mineral surfaces. Furthermore, in agricultural ecosystems, plant-available P content in topsoil is mainly controlled by fertilization and land management. To understand, model, and predict P dynamics at the landscape level, the availability of detailed observation-based P data is extremely valuable. We used more than 388,000 topsoil plant-available P samples from the period 2005 to 2021 to study spatial and temporal variability and land-use effect on soil P. We developed a mapping approach based on existing databases of soil, land-use, and fragmentary soil P measurements by land-use classes to provide spatially explicit high-resolution estimates of topsoil P at the national level. The modeled spatially detailed (1:10,000 scale) GIS dataset of topsoil P is useful for precision farming to optimize nutrient application and to increase productivity; it can also be used as input for biogeochemical models and to assess P load in inland waters and sea. Full article

The demand for safe and edible meat has led to the advancement of freeze-storage techniques, but falsely labeled thawed meat remains an issue. Many methods have been proposed for this purpose, but they all destroy the sample and can only be performed in the laboratory by skilled personnel. In this study, hyperspectral image data were used to construct a machine learning (ML) model to discriminate between freshly refrigerated, long-term refrigerated, and thawed beef meat samples. With four pre-processing methods, a total of five datasets were prepared to construct an ML model. The PLS-DA and SVM techniques were used to construct the models, and the performance was highest for the SVM model applying scatter correction and the RBF kernel function. These results suggest that it is possible to construct a prediction model to distinguish between fresh and non-fresh meat using the spectra obtained by purifying hyperspectral image data cubes, which can be a rapid and non-invasive method for routine analyses of the meat storage state. Full article

Maize is one of the main grain reserve crops, which directly affects the food security of the country. It is extremely important to evaluate the growth status of maize in a timely and accurate manner. Canopy Chlorophyll Density (CCD) is closely related to crop health status. A timely and accurate estimation of CCD is helpful for managers to take measures to avoid yield loss. Thus, many methods have been developed to estimate CCD with remote sensing data. However, the relationship between the CCD and the features used in these CCD estimation methods at different growth stages is unclear. In addition, the CCD was directly estimated from remote sensing data in most previous studies. If the CCD can be accurately estimated from the estimation results of Leaf Chlorophyll Density (LCD) and Leaf Area Index (LAI) remains to be explored. In this study, Random Forest (RF), Support Vector Machines (SVM), and Multivariable Linear Regression (MLR) were used to develop CCD, LCD, and LAI estimation models by integrating multiple features derived from unmanned aerial vehicle (UAV) multispectral images. Firstly, the performances of the RF, SVM, and MLR trained over spectral features (including vegetation indices and band reflectance; dataset I), texture features (dataset II), wavelet coefficient features (dataset III), and multiple features (dataset IV, including all the above datasets) were analyzed, respectively. Secondly, the CCD_P was calculated from the estimated LCD and estimated LAI, and then the CCD was estimated based on multiple features and the CCD_P was compared. The results show that the correlation between CCD and different features is significantly different at every growth stage. The RF model trained over dataset IV yielded the best performance for the estimation of LCD, LAI, and CCD (R² values were 0.91, 0.97, and 0.97, and RMSE values were 6.59 μg/cm², 0.35, and 24.85 μg/cm²). The CCD directly estimated from dataset IV is slightly closer to the ground truth CCD than the CCD_P (R² = 0.96, RMSE = 26.85 μg/cm²) calculated from LCD and LAI. The results indicated that the CCD of maize can be accurately estimated from multiple multispectral image features at the whole growth stage, and both CCD estimation strategies can be used to estimate the CCD accurately. This study provides a new reference for accurate CCD evaluation in precision agriculture. Full article

Maize is among the most important grain crops. Aboveground biomass (AGB) is a key agroecological indicator for crop yield prediction and growth status monitoring, etc. In this study, we propose two new methods, improved crop biomass algorithm (iCBA) and iCBA with piecewise function (iCBA-PF), to estimate maize AGB. Multispectral (MS) images, visible-band (RGB) images, and light detection and ranging (LiDAR) data were collected using unmanned aerial vehicles (UAVs). Vegetation indices (VIs) and the VI-weighted canopy volume model (CVM_VI) were calculated and used as input variables for AGB estimation. The two proposed methods and three benchmark methods were compared. Results demonstrated that: (1) The performance of MS and RGB data in AGB estimation was similar. (2) AGB was estimated with higher accuracy using CVM_VI than using VI, probably because the temporal trends of CVM_VI and AGB were similar in the maize growing season. (3) The best estimation method was the iCBA-PF (R² = 0.90 ± 0.02, RMSE = 190.01 ± 21.55 g/m²), indicating that AGB before and after maize heading should be estimated with different methods. Our method and findings are possibly applicable to other crops with a heading stage. Full article

Apples are ranked third, after bananas and oranges, in global fruit production. Fresh apples are more likely to be appreciated by consumers during the marketing process. However, apples inevitably suffer mechanical damage during transport, which can affect their economic performance. Therefore, the timely detection of apples with surface defects can effectively reduce economic losses. In this paper, we propose an apple surface defect detection method based on weight contrast transfer and the MobileNetV3 model. By means of an acquisition device, a thermal, infrared, and visible apple surface defect dataset is constructed. In addition, a model training strategy for weight contrast transfer is proposed in this paper. The MobileNetV3 model with weight comparison transfer (Weight Compare-MobileNetV3, WC-MobileNetV3) showed a 16% improvement in accuracy, 14.68% improvement in precision, 14.4% improvement in recall, and 15.39% improvement in F1-score. WC-MobileNetV3 compared to MobileNetV3 with fine-tuning improved accuracy by 2.4%, precision by 2.67%, recall by 2.42% and F1-score by 2.56% compared to the classical neural networks AlexNet, ResNet50, DenseNet169, and EfficientNetV2. The experimental results show that the WC-MobileNetV3 model adequately balances accuracy and detection time and achieves better performance. In summary, the proposed method achieves high accuracy for apple surface defect detection and can meet the demand of online apple grading. Full article

Rice, the world’s most extensively cultivated cereal crop, serves as a staple food and energy source for over half of the global population. A variety of abiotic and biotic factors such as weather conditions, soil quality, temperature, insects, pathogens, and viruses can greatly impact the quantity and quality of rice grains. Studies have established that plant infections have a significant impact on rice crops, resulting in substantial financial losses in agriculture. To accurately diagnose and manage the diseases affecting rice plants, plant pathologists are seeking efficient and reliable methods. Traditional disease detection techniques, employed by farmers, involve time-consuming visual inspections and result in inadequate farming practices. With advancements in agricultural technology, the identification of pathogenic organisms in rice plants has become significantly more manageable through techniques such as machine learning and deep learning, which are receiving significant attention in crop disease research. In this paper, we used the transfer learning approach on 15 pre-trained CNN models for the automatic identification of Rice leave diseases. Results showed that the InceptionV3 model is outperforming with an average accuracy of 99.64% with Precision, Recall, F1-Score, and Specificity as 98.23, 98.21, 98.20, and 99.80, and the AlexNet model resulted in poor performance with average accuracy of 97.35% among others. Full article

Due to their rapid development and wide application in modern agriculture, robots, mobile terminals, and intelligent devices have become vital technologies and fundamental research topics for the development of intelligent and precision agriculture. Accurate and efficient target detection technology is required for mobile inspection terminals, picking robots, and intelligent sorting equipment in tomato production and management in plant factories. However, due to the limitations of computer power, storage capacity, and the complexity of the plant factory (PF) environment, the precision of small-target detection for tomatoes in real-world applications is inadequate. Therefore, we propose an improved Small MobileNet YOLOv5 (SM-YOLOv5) detection algorithm and model based on YOLOv5 for target detection by tomato-picking robots in plant factories. Firstly, MobileNetV3-Large was used as the backbone network to make the model structure lightweight and improve its running performance. Secondly, a small-target detection layer was added to improve the accuracy of small-target detection for tomatoes. The constructed PF tomato dataset was used for training. Compared with the YOLOv5 baseline model, the mAP of the improved SM-YOLOv5 model was increased by 1.4%, reaching 98.8%. The model size was only 6.33 MB, which was 42.48% that of YOLOv5, and it required only 7.6 GFLOPs, which was half that required by YOLOv5. The experiment showed that the improved SM-YOLOv5 model had a precision of 97.8% and a recall rate of 96.7%. The model is lightweight and has excellent detection performance, and so it can meet the real-time detection requirements of tomato-picking robots in plant factories. Full article

Due to the dense distribution of tomato fruit with similar morphologies and colors, it is difficult to recognize the maturity stages when the tomato fruit is harvested. In this study, a tomato maturity recognition model, YOLOv5s-tomato, is proposed based on improved YOLOv5 to recognize the four types of different tomato maturity stages: mature green, breaker, pink, and red. Tomato maturity datasets were established using tomato fruit images collected at different maturing stages in the greenhouse. The small-target detection performance of the model was improved by Mosaic data enhancement. Focus and Cross Stage Partial Network (CSPNet) were adopted to improve the speed of network training and reasoning. The Efficient IoU (EIoU) loss was used to replace the Complete IoU (CIoU) loss to optimize the regression process of the prediction box. Finally, the improved algorithm was compared with the original YOLOv5 algorithm on the tomato maturity dataset. The experiment results show that the YOLOv5s-tomato reaches a precision of 95.58% and the mean Average Precision (mAP) is 97.42%; they are improved by 0.11% and 0.66%, respectively, compared with the original YOLOv5s model. The per-image detection speed is 9.2 ms, and the size is 23.9 MB. The proposed YOLOv5s-tomato can effectively solve the problem of low recognition accuracy for occluded and small-target tomatoes, and it also can meet the accuracy and speed requirements of tomato maturity recognition in greenhouses, making it suitable for deployment on mobile agricultural devices to provide technical support for the precise operation of tomato-picking machines. Full article

Soybeans (Glycine max (L.) Merr.), a popular food resource worldwide, have various uses throughout the industry, from everyday foods and health functional foods to cosmetics. Soybeans are vulnerable to pests such as stink bugs, beetles, mites, and moths, which reduce yields. Riptortus pedestris (R. pedestris) has been reported to cause damage to pods and leaves throughout the soybean growing season. In this study, an experiment was conducted to detect R. pedestris according to three different environmental conditions (pod filling stage, maturity stage, artificial cage) by developing a surveillance platform based on an unmanned ground vehicle (UGV) GoPro CAM. Deep learning technology (MRCNN, YOLOv3, Detectron2)-based models used in this experiment can be quickly challenged (i.e., built with lightweight parameter) immediately through a web application. The image dataset was distributed by random selection for training, validation, and testing and then preprocessed by labeling the image for annotation. The deep learning model localized and classified the R. pedestris individuals through a bounding box and masking in the image data. The model achieved high performances, at 0.952, 0.716, and 0.873, respectively, represented through the calculated means of average precision (mAP) value. The manufactured model will enable the identification of R. pedestris in the field and can be an effective tool for insect forecasting in the early stage of pest outbreaks in crop production. Full article

LNC (leaf nitrogen content) in crops is significant for diagnosing the crop growth status and guiding fertilization decisions. Currently, UAV (unmanned aerial vehicles) remote sensing has played an important role in estimating the nitrogen nutrition of crops at the field scale. However, many existing methods of evaluating crop nitrogen based on UAV imaging techniques usually have used a single type of imagery such as RGB or multispectral images, seldom considering the usage of information fusion from different types of UAV imagery for assessing the crop nitrogen status. In this study, GS (Gram–Schmidt Pan Sharpening) was utilized to fuse images from two sensors of digital RGB and multispectral cameras mounted on UAV, and the specific bands of the multispectral cameras are blue, green, red, rededge and NIR. The color space transformation method, HSV (Hue-Saturation-Value), was used to separate soil background noise from crops due to the high spatial resolution of UAV images. Two methods of optimizing feature variables, the Successive Projection Algorithm (SPA) and the Competitive Adaptive Reweighted Sampling method (CARS), combined with two regularization regression algorithms, LASSO and RIDGE, were adopted to estimate the LNC, compared to the commonly used Random Forest algorithm. The results showed that: (1) the accuracy of LNC estimation using the fusion image is improved distinctly by a comparison to the original multispectral image; (2) the denoised images performed better than the original multispectral images in evaluating LNC in rice; (3) the RIDGE-SPA combined method, using SPA to select the MCARI, SAVI and OSAVI, had the best performance for LNC in rice, with an R² of 0.76 and an RMSE of 10.33%. It can be demonstrated that the information fusion of multiple-sensor imagery from UAV coupling with the methods of optimizing feature variables can estimate the rice LNC more effectively, which can also provide a reference for guiding the decision making of fertilization in rice fields. Full article

As the habitat areas of Tibetan antelopes usually exhibit poaching and unpredictable risks, combining target recognition and tracking with intelligent Unmanned Aerial Vehicle (UAV) technology is necessary to obtain the real-time location of injured Tibetan antelopes to better protect and rescue them. (1) Background: The most common way to track an object is to detect each frame of it, and it is not necessary to run the object tracker and classifier at the same rate, because the speed for them to change class is slower than objects move. Especially in the edge reasoning scene, UAV real-time monitoring requires to seek a balance between the frame rate, latency, and accuracy. (2) Methods: A backtracking tracker is proposed to recognize Tibetan antelopes which generates motion vectors through stored optical flow, achieving faster target detection. The lightweight You Only Look Once X (YOLOX) is selected as the baseline model to reduce the dependence on hardware configuration and calculation cost while ensuring detection accuracy. Region-of-Interest (ROI)-to-centroid tracking technology is employed to reduce the processing cost of motion interpolation, and the overall processing frame rate is smoothed by pre-calculating the motions of different objects recognized. The On-Line Object Tracking (OLOT) system with adaptive search area selection is adopted to dynamically adjust the frame rate to reduce energy waste. (3) Results: using YOLOX to trace back in the native Darkenet can reduce latency by 3.75 times, and the latency is only 2.82 ms after about 10 frame hops, with the accuracy being higher than YOLOv3. Compared with traditional algorithms, the proposed algorithm can reduce the tracking latency of UAVs by 50%. By running and comparing in the onboard computer, although the proposed tracker is inferior to KCF in FPS, it is significantly higher than other trackers and is obviously superior to KCF in accuracy. (4) Conclusion: A UAV equipped with the proposed tracker effectively reduces reasoning latency in monitoring Tibetan antelopes, achieving high recognition accuracy. Therefore, it is expected to help better protection of Tibetan antelopes. Full article

Post-harvest diseases are one of the main causes of economical losses in the apple fruit production sector. Therefore, this paper presents an application of a knowledge-based expert system to diagnose post-harvest diseases of apple. Specifically, we detail the process of domain knowledge elicitation for constructing a Bayesian network reasoning system. We describe the developed expert system, dubbed BN-DSSApple, and the diagnostic mechanism given the evidence provided by the user, as well as a likelihood evidence method, learned from the estimated consensus of users’ and expert’s interactions, to effectively transfer the performance of the model to different cohorts of users. Finally, we detail a novel technique for explaining the provided diagnosis, thus increasing the trust in the system. We evaluate BN-DSSApple with three different types of user studies, involving real diseased apples, where the ground truth of the target instances was established by microbiological and DNA analysis. The experiments demonstrate the performance differences in the knowledge-based reasoning mechanism due to heterogeneous users interacting with the system under various conditions and the capability of the likelihood-based method to improve the diagnostic performance in different environments. Full article

Increasing agricultural production is a major concern that aims to increase income, reduce hunger, and improve other measures of well-being. Recently, the prediction of soil-suitability has become a primary topic of rising concern among academics, policymakers, and socio-economic analysts to assess dynamics of the agricultural production. This work aims to use physico-chemical and remotely sensed phenological parameters to produce soil-suitability maps (SSM) based on Machine Learning (ML) Algorithms in a semi-arid and arid region. Towards this goal an inventory of 238 suitability points has been carried out in addition to14 physico-chemical and 4 phenological parameters that have been used as inputs of machine-learning approaches which are five MLA prediction, namely RF, XgbTree, ANN, KNN and SVM. The results showed that phenological parameters were found to be the most influential in soil-suitability prediction. The validation of the Receiver Operating Characteristics (ROC) curve approach indicates an area under the curve and an AUC of more than 0.82 for all models. The best results were obtained using the XgbTree with an AUC = 0.97 in comparison to other MLA. Our findings demonstrate an excellent ability for ML models to predict the soil-suitability using physico-chemical and phenological parameters. The approach developed to map the soil-suitability is a valuable tool for sustainable agricultural development, and it can play an effective role in ensuring food security and conducting a land agriculture assessment. Full article

Leaf Area Index (LAI) is an important parameter which can be used for crop growth monitoring and yield estimation. Many studies have been carried out to estimate LAI with remote sensing data obtained by sensors mounted on Unmanned Aerial Vehicles (UAVs) in major crops; however, most of the studies used only a single type of sensor, and the comparative study of different sensors and sensor combinations in the model construction of LAI was rarely reported, especially in soybean. In this study, three types of sensors, i.e., hyperspectral, multispectral, and LiDAR, were used to collect remote sensing data at three growth stages in soybean. Six typical machine learning algorithms, including Unary Linear Regression (ULR), Multiple Linear Regression (MLR), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Back Propagation (BP), were used to construct prediction models of LAI. The results indicated that the hyperspectral and LiDAR data did not significantly improve the prediction accuracy of LAI. Comparison of different sensors and sensor combinations showed that the fusion of the hyperspectral and multispectral data could significantly improve the predictive ability of the models, and among all the prediction models constructed by different algorithms, the prediction model built by XGBoost based on multimodal data showed the best performance. Comparison of the models for different growth stages showed that the XGBoost-LAI model for the flowering stage and the universal models of the XGBoost-LAI and RF-LAI for three growth stages showed the best performances. The results of this study might provide some ideas for the accurate estimation of LAI, and also provide novel insights toward high-throughput phenotyping of soybean with multi-modal remote sensing data. Full article

In this work, we have proposed a novel methodology for greenhouse tomato yield prediction, which is based on a hybrid of an explanatory biophysical model—the Tomgro model, and a machine learning model called CNN-RNN. The Tomgro and CNN-RNN models are calibrated/trained for predicting tomato yields while different fusion approaches (linear, Bayesian, neural network, random forest and gradient boosting) are exploited for fusing the prediction result of individual models for obtaining the final prediction results. The experimental results have shown that the model fusion approach achieves more accurate prediction results than the explanatory biophysical model or the machine learning model. Moreover, out of different model fusion approaches, the neural network one produced the most accurate tomato prediction results, with means and standard deviations of root mean square error (RMSE), r2-coefficient, Nash-Sutcliffe efficiency (NSE) and percent bias (PBIAS) being 17.69 ± 3.47 g/m${}^2$, 0.9995 ± 0.0002, 0.9989 ± 0.0004 and 0.1791 ± 0.6837, respectively. Full article

Predicting leaf nitrogen content (LNC) using unmanned aerial vehicle (UAV) images is of great significance. Traditional LNC prediction methods based on empirical and mechanistic models have limitations. This study aimed to propose a new LNC prediction method based on combining deep learning methods and mechanistic models. Wheat field experiments were conducted to make plants with different LNC values. The LNC and UAV hyperspectral images were collected during the critical growth stages of wheat. Based on these data, a method combining the deep multitask learning method and the N-based PROSAIL model was proposed and compared with traditional LNC prediction methods, including spectral index (SI), partial least squares regression (PLSR) and artificial neural network (ANN) methods. The results show that the new proposed method obtained the best LNC prediction results, with R², RMSE and RMSE% values of 0.79, 20.86 μg/cm² and 18.63%, respectively, during calibration and 0.82, 18.40 μg/cm² and 16.92%, respectively, during validation. The other methods obtained R², RMSE and RMSE% values between 0.29 and 0.68, 25.71 and 38.52 μg/cm² and 22.95 and 34.39%, respectively, during calibration and between 0.43 and 0.74, 22.79 and 33.55 μg/cm² and 20.96 and 30.86%, respectively, during validation. Thus, this study provides an accurate LNC prediction tool for precise nitrogen (N) management in the field. Full article

A phenotype is the composite of an observable expression of a genome for traits in a given environment. The trajectories of phenotypes computed from an image sequence and timing of important events in a plant’s life cycle can be viewed as temporal phenotypes and indicative of the plant’s growth pattern and vigor. In this paper, we introduce a novel method called FlowerPhenoNet, which uses deep neural networks for detecting flowers from multiview image sequences for high-throughput temporal plant phenotyping analysis. Following flower detection, a set of novel flower-based phenotypes are computed, e.g., the day of emergence of the first flower in a plant’s life cycle, the total number of flowers present in the plant at a given time, the highest number of flowers bloomed in the plant, growth trajectory of a flower, and the blooming trajectory of a plant. To develop a new algorithm and facilitate performance evaluation based on experimental analysis, a benchmark dataset is indispensable. Thus, we introduce a benchmark dataset called FlowerPheno, which comprises image sequences of three flowering plant species, e.g., sunflower, coleus, and canna, captured by a visible light camera in a high-throughput plant phenotyping platform from multiple view angles. The experimental analyses on the FlowerPheno dataset demonstrate the efficacy of the FlowerPhenoNet. Full article

The mechanisms of lettuce growth in plant factories under artificial light (PFALs) are well known, whereby the crop is generally used as a model in horticultural science. Deep learning has also been tested several times using PFALs. Despite their numerous advantages, the performance of deep learning models is commonly evaluated based only on their accuracy. Therefore, the objective of this study was to train deep neural networks and analyze the deeper abstraction of the trained models. In total, 443 images of three lettuce cultivars were used for model training, and several deep learning algorithms were compared using multivariate linear regression. Except for linear regression, all models showed adequate accuracies for the given task, and the convolutional neural network (ConvNet) model showed the highest accuracy. Based on color mapping and the distribution of the two-dimensional t-distributed stochastic neighbor embedding (t-SNE) results, ConvNet effectively perceived the differences among the lettuce cultivars under analysis. The extension of the target domain knowledge with complex models and sufficient data, similar to ConvNet with multitask learning, is possible. Therefore, deep learning algorithms should be investigated from the perspective of feature extraction. Full article

Non-invasive techniques for the detection of apple fruit damages are central to the correct operation of sorting lines ensuring storability of the collected fruit batches. The choice of optimal method of fruit imaging and efficient image processing method is still a subject of debate. Here, we have dissected the information content of hyperspectral images focusing on either spectral component, spatial component, or both. We have employed random forest (RF) classifiers using different parameters as inputs: reflectance spectra, vegetation indices (VIs), and spatial texture descriptors (local binary patterns, or LBP), comparing their performance in the task of damage detection in apple fruit. The amount of information in raw hypercubes was found to be over an order of magnitude excessive for the end-to-end problem of classification. Converting spectra to vegetation indices has resulted in a 60-fold compression with no significant loss of information relevant for phenotyping and more robust performance with respect to varying illumination conditions. We concluded that the advanced machine learning approaches could be more efficient if complemented by spectral information about the objects in question. We discuss the potential advantages and pitfalls of the different approaches to the machine learning-based processing of hyperspectral data for fruit grading. Full article

Spodoptera frugiperda (fall armyworm, FAW) is a global agriculture pest. Adults have a strong migratory ability and larvae feed on the host stalks, which pose a serious threat for maize and other crops. Identification and counting of different instar larvae in the fields is important for effective pest management and forecasting emergence and migration time of adults. Usually, the technicians identify the larval instars according to the larva morphological features with the naked eye or stereoscope in the lab. The manual identification method is complex, professional and inefficient. In order to intelligently, quickly and accurately identify the larval instar, we design a portable image acquisition device using a mobile phone with a macro lens and collect 1st-6th instar larval images. The YOLOv4 detection method and improved MRES-UNet++ segmentation methods are used to locate the larvae and segment the background. The larval length and head capsule width are automatically measured by some graphics algorithms, and the larval image features are extracted by SIFT descriptors. The random forest model improved by Boruta feature selection and grid search method is used to identify the larval instars of FAWs. The test results show that high-definition images can be easily collected by using the portable device (Shenzhen, China). The MRES-UNet++ segmentation method can accurately segment the larvae from the background. The average measurement error of the head capsule width and body length of moth larvae is less than 5%, and the overall identification accuracy of 1st–6th instar larvae reached 92.22%. Our method provides a convenient, intelligent and accurate tool for technicians to identify the larval instars of FAWs. Full article

Affected by geographical location and climatic conditions, crop classification in the Yunnan Plateau of China is greatly restricted by the low utilization rate of annual optical data, complex crop planting structure, and broken cultivated land. This paper combines monthly Sentinel-2 optical remote sensing data with Sentinel-1 radar data to minimize cloud interference to conduct crop classification for plateau areas. However, pixel classification will inevitably produce a “different spectrum of the same object, foreign objects in the same spectrum”. A principal component feature synthesis method is developed for multi-source remote sensing data (PCA-MR) to improve classification accuracy. In order to compare and analyze the classification effect of PCA-MR combined with multi-source remote sensing data, we constructed 11 classification scenarios using the Google Earth Engine platform and random forest algorithm (RF). The results show that: (1) the classification accuracy is 79.98% by using Sentinel-1 data and 91.18% when using Sentinel-2 data. When integrating Sentinel-1 and Sentinel-2 data, the accuracy is 92.31%. By analyzing the influence of texture features on classification under different feature combinations, it was found that optical texture features affected the recognition accuracy of rice to a lesser extent. (2) The errors will be reduced if the PCA-MR feature is involved in the classification, and the classification accuracy and Kappa coefficient are improved to 93.47% and 0.92, respectively. Full article

Soybean is one of the most important agricultural commodities in the world, thus making it important for global food security. However, widely used process-based crop models, such as the GIS-based Environmental Policy Integrated Climate (GEPIC) model, tend to underestimate the impacts of extreme climate events on soybean, which brings large uncertainties. This study proposed an approach of hybrid models to constrain such uncertainties by coupling the GEPIC model and extreme climate indicators using machine learning. Subsequently, the key extreme climate indicators for the globe and main soybean producing countries are explored, and future soybean yield changes and variability are analyzed using the proposed hybrid model. The results show the coupled GEPIC and Random Forest (GEPIC+RF) model (R: 0.812, RMSD: 0.716 t/ha and rRMSD: 36.62%) significantly eliminated uncertainties and underestimation of climate extremes from the GEPIC model (R: 0.138, RMSD: 1.401 t/ha and rRMSD: 71.57%) compared to the other five hybrid models (R: 0.365–0.612, RMSD: 0.928–1.021 and rRMSD: 47.48–52.24%) during the historical period. For global soybean yield and those in Brazil and Argentina, low-temperature-related indices are the main restriction factors, whereas drought is the constraining factor in the USA and China, and combined drought–heat disaster in India. The GEPIC model would overestimate soybean yields by 13.40–27.23%. The GEPIC+RF model reduced uncertainty by 28.45–41.83% for the period of 2040–2099. Our results imply that extreme climate events will possibly cause more losses in soybean in the future than we have expected, which would help policymakers prepare for future agriculture risk and food security under climate change. Full article

Quality grading in antler mushroom industrial production is a labor-intensive operation. For a long time, manual grading has been used for grading, which produces various problems such as insufficient reliability, low production efficiency, and high mushroom body damage. Automatic grading is a problem to be solved urgently for antler mushroom industrial development with increasing labor costs. To solve the problem, this paper deeply integrates the single-stage object detection of YOLOv5 and the semantic segmentation of PSPNet, and proposes a Y-PNet model for real-time object detection and an image segmentation network. This article also proposes an evaluation model for antler mushroom’s size, which eliminates subjective judgment and achieves quality grading. Moreover, to meet the needs of efficient and accurate hierarchical detection in the factory, this study uses the lightweight network model to construct a lightweight YOLOv5 single-stage object detection model. The MobileNetV3 network model embedded with a CBAM module is used as the backbone extractor in PSPNet to reduce the model’s size and improve the model’s efficiency and accuracy for segmentation. Experiments show that the proposed system can perform real-time grading successfully, which can provide instructive and practical references in industry. Full article

Due to large line spacing and planting distances, the adoption of continuous and uniform pesticide spraying in vegetable farming can lead to pesticide waste, thus increasing cost and environmental pollution. In this paper, by applying deep learning and online identification methods, control technology for target-oriented spraying is studied with cabbages as the research object. To overcome motion blur and low average precision under strong light conditions during the operation of sprayers, an innovative YOLOV5 model implanted with a transformer module is utilized to achieve accurate online identification for cabbage fields under complex environments. Based on this concept, a new target-oriented spray system is built on an NVIDIA Jetson Xavier NX. Indoor test results show that the average precision is 96.14% and the image processing time is 51.07 ms. When motion blur occurs, the average precision for the target is 90.31%. Then, in a field experiment, when the light intensity is within the range of 3.76–12.34 wlx, the advance opening distance is less than 3.51 cm, the delay closing distance is less than 2.05 cm, and the average identification error for the cabbage diameter is less than 1.45 cm. The experimental results indicate that changes in light intensity have no significant impact on the identification effect. The average precision is 98.65%, and the savings rate reaches 54.04%. In general, the target-oriented spray system designed in this study achieves the expected experimental results and can provide technical support for field target spraying. Full article

By studying the spectral information of cotton leaf nitrogen content, sensitive feature bands and spectral indices for leaf nitrogen content were screened, and different methods were used to model the screened feature bands and indices to find a method with higher accuracy and stability of the inversion model, which provides a theoretical basis and technical support for remote sensing estimation of cotton nitrogen content in Xinjiang. The experiment was conducted in 2019–2020 at the Second Company of Shihezi University Teaching Experimental Farm in Xinjiang, China, with six fertilization treatments (0, 120, 240, 360, 480 kg/hm pure N), sampled at five key fertility stages of cotton (squaring stage, full budding stage, flowering, boll stage, and boll opening stage), and the obtained data were used in two modeling approaches (eigenbands and spectral indices) to establish a cotton nitrogen estimation model and estimate the cotton leaf N content. The results showed that the nonlinear model using SVR was validated with an R² of 0.71 and an RMSE of 3.91. The linear models of MLR and PLS were developed for the feature bands screened by SPA and RF, respectively, and the best modeling result was achieved by SPA-PLS with a validated R² of 0.722 and an RMSE of 3.83. The existing spectral indices were optimized by screening the central wavelength and the simple linear regression model was constructed. The inversion effect of the SVR model with the characteristic spectral modeling was better than the index results. However, compared with the direct use of the characteristic wavelengths and the SVR way of modeling, the accuracy of leaf N content estimation by the model built by optimizing the spectral indices was reduced but the stability was greatly improved, and it can be used as a hyperspectral model for leaf N content at full fertility. The hyperspectral estimation of leaf N content in cotton can be used as a hyperspectral estimation method for the whole fertility period. Full article

UAV imaging provides an efficient and non-destructive tool for characterizing farm information, but the quality of the UAV model is often affected by the image’s spatial resolution. In this paper, the predictability of models established using UAV multispectral images with different spatial resolutions for nitrogen content of winter wheat was evaluated during the critical growth stages of winter wheat over the period 2021–2022. Feature selection based on UAV image reflectance, vegetation indices, and texture was conducted using the competitive adaptive reweighted sampling, and the random forest machine learning method was used to construct the prediction model with the optimized features. Results showed that model performance increased with decreasing image spatial resolution with a R², a RMSE, a MAE and a RPD of 0.84, 4.57 g m⁻², 2.50 g m⁻² and 2.34 (0.01 m spatial resolution image), 0.86, 4.15 g m⁻², 2.82 g m⁻² and 2.65 (0.02 m), and 0.92, 3.17 g m⁻², 2.45 g m⁻² and 2.86 (0.05 m), respectively. Further, the transferability of models differed when applied to images with coarser (upscaling) or finer (downscaling) resolutions. For upscaling, the model established with the 0.01 m images had a R² of 0.84 and 0.89 when applied to images with 0.02 m and 0.05 m resolutions, respectively. For downscaling, the model established with the 0.05 m image features had a R² of 0.86 and 0.83 when applied to images of 0.01 m and 0.02 m resolutions. Though the image spatial resolution affects image texture features more than the spectral features and the effects of image spatial resolution on model performance and transferability decrease with increasing plant wetness under irrigation treatment, it can be concluded that all the UAV images acquired in this study with different resolutions could achieve good predictions and transferability of the nitrogen content of winter wheat plants. Full article

The early and accurate acquisition of crop yields is of great significance for maintaining food market stability and ensuring global food security. Unmanned aerial vehicle (UAV) remote sensing offers the possibility of predicting crop yields with its advantages of flexibility and high resolution. However, most of the existing remote sensing yield estimation studies focused solely on crops but did not fully consider the influence of soil on yield formation. As an integrated system, the status of crop and soil together determines the final yield. Compared to crop-only yield prediction, the approach that additionally considers soil background information will effectively improve the accuracy and reduce bias in the results. In this study, a novel method for segmenting crop and soil spectral images based on different vegetation coverage is first proposed, in which pixels of crop and soil can be accurately identified by determining the discriminant value Q. On the basis of extracting crop and soil waveband’s information by individual pixel, an innovative approach, projected non-negative matrix factorization based on good point set and matrix cross fusion (PNMF-MCF), was developed to effectively extract and fuse the yield-related features of crop and soil. The experimental results on winter wheat show that the proposed segmentation method can accurately distinguish crop and soil pixels under complex soil background of four different growth periods. Compared with the single reflectance of crop or soil and the simple combination of crop and soil reflectance, the fused yield features spectral matrix F_P obtained with PNMF−MCF achieved the best performance in yield prediction at the flowering, flag leaf and pustulation stages, with R² higher than 0.7 in these three stages. Especially at the flowering stage, the yield prediction model based on PNMF-MCF had the highest R² with 0.8516 and the lowest RMSE with 0.0744 kg/m². Correlation analysis with key biochemical parameters (nitrogen and carbon, pigments and biomass) of yield formation showed that the flowering stage was the most vigorous season for photosynthesis and the most critical stage for yield prediction. This study provides a new perspective and complete framework for high-precision crop yield forecasting using UAV remote sensing technology. Full article

As the second largest rice producer, India contributes about 20% of the world’s rice production. Timely, accurate, and reliable rice yield prediction in India is crucial for global food security and health issues. Deep learning models have achieved excellent performances in predicting crop yield. However, the interpretation of deep learning models is still rare. In this study, we proposed a transformer-based model, Informer, to predict rice yield across the Indian Indo-Gangetic Plains by integrating time-series satellite data, environmental variables, and rice yield records from 2001 to 2016. The results showed that Informer had better performance (R² = 0.81, RMSE = 0.41 t/ha) than four other machine learning and deep learning models for end-of-season prediction. For within-season prediction, the Informer model could achieve stable performances (R² ≈ 0.78) after late September, which indicated that the optimal prediction could be achieved 2 months before rice maturity. In addition, we interpreted the prediction models by evaluating the input feature importance and analyzing hidden features. The evaluation of feature importance indicated that NIR_V was the most critical factor, while intervals 6 (mid-August) and 12 (mid-November) were the key periods for rice yield prediction. The hidden feature analysis demonstrated that the attention-based long short-term memory (AtLSTM) model accumulated the information of each growth period, while the Informer model focused on the information around intervals 5 to 6 (August) and 11 to 12 (November). Our findings provided a reliable and simple framework for crop yield prediction and a new perspective for explaining the internal mechanism of deep learning models. Full article

Flower thinning for fruit trees in time is an important operation to keep a suitable quantity of fruits and guarantee the quality of fruits. Accurate detection of flower density is the premise of precise flower thinning, and machine vision provides an effective approach to achieving the accurate identification of flower density. To detect the flower density on the proximal side of Y-shaped densely planted peach trees accurately, this study proposed a method based on an RGBD camera and a convolutional neural network that incorporated an attention mechanism and multi-scale feature fusion. Firstly, image acquisition and preprocessing were performed with the RGBD camera, and the complex background and distal flowers were filtered out through depth information. Then, a convolutional neural network for flower density detection based on an attention mechanism and multi-scale feature fusion, named the flower counting network (FC-Net), was constructed and tested. Results showed that the coefficient of determination (R²) between the estimated number of flowers by the FC-Net and the real values reached 0.95, the mean absolute error (MAE) was 4.3, the root mean square error (RMSE) was 5.65, the counting error rate (Er) was 0.02%, and the processing time of one image was 0.12 s. The proposed FC-Net can provide visual support for intelligent mechanical flower thinning operations. Full article

Soil moisture (SM), as a crucial variable in the soil–vegetation–atmosphere continuum, plays an important role in the terrestrial water cycle. Analyzing SM’s variation and driver factors is crucial to maintaining ecosystem diversity on the Tibetan Plateau (TP) and ensuring food security as well as water supply balance in developing countries. Gradual wetting of the soil has been detected and attributed to precipitation in this area. However, there is still a gap in understanding the potential mechanisms. It is unclear whether the greening, glacier melting, and different vegetation degradation caused by asymmetrical climate change and intensified human activities have significantly affected the balance of SM. Here, to test the hypothesis that heterogeneous SM caused by precipitation was subject to temperatures and anthropogenic constraints, GLDAS-2.1 (Global Land Data Assimilation System-2.1) SM products combined with the statistical downscaling and Geographic detectors were applied. The results revealed that: (1) Seasonal SM gradually increased (p < 0.05), while SM deficit frequently appeared with exposure to extreme climates, such as in the summer of 2010 and 2013, and changed into a pattern of precipitation transport to western dry lands in autumn. (2) There was a synergistic reaction between greening and local moisture in autumn. SM was dominated by low temperature (TMN) in winter, warming indirectly regulated SM by exacerbating the thawing of glaciers and permafrost. The spatial coupling between the faster rising rate of TMN and the frozen soil might further aggravate the imbalance of SM. (3) The land cover’s mutual transformation principally affected SM in spring and autumn, and degradation accelerated the loss of SM replenished by precipitation. (4) Land cover responses were different; SM in grassland was less affected by external disturbance, while degraded woodland and shrub performed adaptive feedback under dry environments, SM increased by 0.05 and 0.04 m³/(m³ 10a), respectively. Our research provides a scientific basis for improving hydrological models and developing vegetation restoration strategies for long-term adaptation to TP-changing environments. Full article

Considering the slow response and unstable velocity of agricultural machinery caused by soil resistance, actuator delay, environmental change, velocity fluctuation, and other internal and external factors under real working conditions, a kind of agricultural machinery following a control system that considers variable load and control delay was proposed. Taking distance-keeping, velocity-following, and acceleration-following as parameters, the controller model was deduced, and the influence of different values of model parameters on the driving stability of agricultural machinery was analyzed in detail. In addition, this paper describes a kind of agricultural machinery following a strategy that can realize the graded adjustment of vehicle distance with the dynamic increase in vehicle weight. Then, the following strategy, under the influence of velocity and quality, was simulated and verified using MATLAB/Simulink (MATLAB2016a, mathworks: Natick, Massachusetts, USA). When the crop harvester was at 1.5 m/s and the amplitude of velocity fluctuation was 0.3 m and 1.3 m, respectively, the grain truck could adjust its velocity to keep up with the crop harvester to complete the operation task. Simulation verification was carried out for the proposed graded adjustment of vehicle distance of agricultural machinery following strategy. The unit mass of the crops was set at 360 kg, and the vehicle distance changed at 18s to adapt to the graded adjustment of the vehicle distance following strategy. Finally, a real-vehicle validation test was carried out, and the results show that the grain truck velocity can keep up with the change of crop harvester velocity on the basis of maintaining the desired vehicle distance, the grain truck velocity can keep up with the change of crop harvester velocity on the road condition, which verifies the effectiveness and feasibility of the proposed method. Full article

Soil organic matter (SOM) is one of the main sources of plant nutrition and promotes plant growth and development. The content of SOM varies in different areas of the field. In this study, a method based on pyrolysis and electronic nose combined with multi-feature data fusion optimization was proposed to realize rapid, accurate and low-cost measurement of SOM content. Firstly, an electronic nose was used to collect response data from the soil pyrolysis gas, and the sensor features (10 × 6) were extracted to form the original feature space. Secondly, Pearson correlation coefficient (PCC), one-way analysis of variance (One-Way ANOVA), principal component analysis algorithm (PCA), linear discriminant analysis algorithm (LDA), and genetic algorithm-backpropagation neural network algorithm (GA-BP) were used to realize multi-feature data fusion optimization. Thirdly, the optimized feature space was used to train the PLSR models, and the predictive performance of the models were used as an indicator to evaluate different feature optimization algorithms. The results showed that the PLSR model with GA-BP for feature optimization had the best predictive performance (R² = 0.90) and could achieve accurate quantitative prediction of SOM content. The dimensionality of the optimized feature space was reduced to 30 and there was no redundancy in the sensor array. Full article

The body size of pigs is a vital evaluation indicator for growth monitoring and selective breeding. The detection of joint points is critical for accurately estimating pig body size. However, most joint point detection methods focus on improving detection accuracy while neglecting detection speed and model parameters. In this study, we propose an HRNet with Swin Transformer block (HRST) based on HRNet for detecting the joint points of pigs. It can improve model accuracy while significantly reducing model parameters by replacing the fourth stage of parameter redundancy in HRNet with a Swin Transformer block. Moreover, we implemented joint point detection for multiple pigs following two steps: first, CenterNet was used to detect pig posture (lying or standing); then, HRST was used for joint point detection for standing pigs. The results indicated that CenterNet achieved an average precision (AP) of 86.5%, and HRST achieved an AP of 77.4% and a real-time detection speed of 40 images per second. Compared with HRNet, the AP of HRST improved by 6.8%, while the number of model parameters and the calculated amount reduced by 72.8% and 41.7%, respectively. The study provides technical support for the accurate and rapid detection of pig joint points, which can be used for contact-free body size estimation of pigs. Full article

A high-efficiency, nondestructive, rapid, and automatic crop disease classification method is essential for the modernization of agriculture. To more accurately extract and fit citrus disease image features, we designed a new 13-layer convolutional neural network (CNN13) consisting of multiple convolutional layer stacks and dropout in this study. To address the problem created by the uneven number of disease images in each category, we used the VGG16 network module for transfer learning, which we combined with the proposed CNN13 to form a new joint network, which we called OplusVNet. To verify the performance of the proposed OplusVNet network, we collected 1869 citrus pest and disease images and 202 normal citrus images from the field. The experimental results showed that the proposed OplusVNet can more effectively solve the problem caused by uneven data volume and has higher recognition accuracy, especially for image categories with a relatively small data volume. Compared with the state of the art networks, the generalization ability of the proposed OplusVNet network is stronger for classifying diseases. The classification accuracy of the model prediction results was 0.99, indicating the model can be used as a reference for crop image classification. Full article

To ensure continued food security and economic development in Africa, it is very important to address and adapt to climate change. Excessive dependence on rainfed agricultural production makes Africa more vulnerable to climate change effects. Weather information and services are essential for farmers to more effectively survive the increasing occurrence of extreme weather events due to climate change. Weather information is important for resource management in agricultural production and helps farmers plan their farming activities in advance. Machine Learning is one of the technologies used in agriculture for weather forecasting and crop disease detection among others. The objective of this study is to develop Machine Learning-based models adapted to the context of daily weather forecasting for Rainfall, Relative Humidity, and Maximum and Minimum Temperature in Senegal. In this study, we made a comparison of ten Machine Learning Regressors with our Ensemble Model. These models were evaluated based on Mean Absolute Error, Mean Squared Error, Root Mean Squared Error and Coefficient of Determination. The results show that the Ensemble Model performs better than the ten base models. The Ensemble Model results for each parameter are as follows; Relative Humidity: Mean Absolute Error was 4.0126, Mean Squared Error was 29.9885, Root Mean Squared Error was 5.4428 and Coefficient of Determination was 0.9335. For Minimum Temperature: Mean Absolute Error was 0.7908, Mean Squared Error was 1.1329, Root Mean Squared Error was 1.0515 and Coefficient of Determination was 0.9018. For Maximum Temperature: Mean Absolute Error was 1.2515, Mean Squared Error was 2.8038, Root Mean Squared Error was 1.6591 and Coefficient of Determination was 0.8205. For Rainfall: Mean Absolute Error was 0.2142, Mean Squared Error was 0.1681, Root Mean Squared Error was 0.4100 and Coefficient of Determination was 0.7733. From the present study, it has been observed that the Ensemble Model is a feasible model to be used for Rainfall, Relative Humidity, and Maximum and Minimum Temperature forecasting. Full article

Cotton constitutes 81% of the world’s natural fibers. Accurate and rapid cotton yield estimation is important for cotton trade and agricultural policy development. Therefore, we developed a remote sensing index that can intuitively represent cotton boll characteristics and support cotton yield estimation by extracting cotton boll pixels. In our study, the Density of open Cotton boll Pixels (DCPs) was extracted by designing different cotton boll indices combined with the threshold segmentation method. The relationship between DCP and field survey datasets, the Density of Total Cotton bolls (DTC), and yield were compared and analyzed. Five common yield estimation models, Linear Regression (LR), Support Vector Regression (SVR), Classification and Regression Trees (CART), Random Forest (RF), and K-Nearest Neighbors (KNN), were implemented and evaluated. The results showed that DCP had a strong correlation with yield, with a Pearson correlation coefficient of 0.84. The RF method exhibited the best yield estimation performance, with average R² and rRMSE values of 0.77 and 7.5%, respectively (five-fold cross-validation). This study showed that RedGreenBlue (RGB) and Near Infrared Red (NIR) normalized, a normalized form index consisting of the RGB and NIR bands, performed best. Full article

Rapid, accurate insect identification is the first and most critical step of pest management and vital to agriculture for determining optimal management strategies. In many instances, classification is necessary within a short developmental window. Two examples, the tobacco budworm, Chloridea virescens, and bollworm, Helicoverpa zea, both have <5 days from oviposition until hatching. H. zea has evolved resistance to Bt-transgenic crops and requires farmers to decide about insecticide application during the ovipositional window. The eggs of these species are small, approximately 0.5 mm in diameter, and often require a trained biologist and microscope to resolve morphological differences between species. In this work, we designed, built, and validated a machine learning approach to insect egg identification with >99% accuracy using a convolutional neural architecture to classify the two species of caterpillars. A gigapixel scale parallelized microscope, referred to as the Multi-Camera Array Microscope (MCAM™), and automated image-processing pipeline allowed us to rapidly build a dataset of ~5500 images for training and testing the network. In the future, applications could be developed enabling farmers to photograph eggs on a leaf and receive an immediate species identification before the eggs hatch. Full article

Cassava is a typical staple food in the tropics, and cassava leaf disease can cause massive yield reductions in cassava, resulting in substantial economic losses and a lack of staple foods. However, the existing convolutional neural network (CNN) for cassava leaf disease classification is easily affected by environmental background noise, which makes the CNN unable to extract robust features of cassava leaf disease. To solve the above problems, this paper introduces a transformer structure into the cassava leaf disease classification task for the first time and proposes a transformer-embedded ResNet (T-RNet) model, which enhances the focus on the target region by modeling global information and suppressing the interference of background noise. In addition, a novel loss function called focal angular margin penalty softmax loss (FAMP-Softmax) is proposed, which can guide the model to learn strict classification boundaries while fighting the unbalanced nature of the cassava leaf disease dataset. Compared to the Xception, VGG16 Inception-v3, ResNet-50, and DenseNet121 models, the proposed method achieves performance improvements of 3.05%, 2.62%, 3.13%, 2.12%, and 2.62% in recognition accuracy, respectively. Meanwhile, the extracted feature maps are visualized and analyzed by gradient-weighted class activation map (Grad_CAM) and 2D T-SNE, which provides interpretability for the final classification results. Extensive experimental results demonstrate that the method proposed in this paper can extract robust features from complex non-balanced disease datasets and effectively carry out the classification of cassava leaf disease. Full article

Considering the population growth rate of recent years, a doubling of the current worldwide crop productivity is expected to be needed by 2050. Pests and diseases are a major obstacle to achieving this productivity outcome. Therefore, it is very important to develop efficient methods for the automatic detection, identification, and prediction of pests and diseases in agricultural crops. To perform such automation, Machine Learning (ML) techniques can be used to derive knowledge and relationships from the data that is being worked on. This paper presents a literature review on ML techniques used in the agricultural sector, focusing on the tasks of classification, detection, and prediction of diseases and pests, with an emphasis on tomato crops. This survey aims to contribute to the development of smart farming and precision agriculture by promoting the development of techniques that will allow farmers to decrease the use of pesticides and chemicals while preserving and improving their crop quality and production. Full article

Food security depends not only on the extent of climate change but also on the compensatory potentials of agronomic improvements. However, the separate contribution of these agronomic factors to rice yield remains largely unknown. Here we distinguished the impacts and relative contributions on rice yield based on statistical models and machine learning by using an observation database collected from 52 agro-meteorological stations in China from 1981 to 2018. Agronomic improvements are responsible for more than 40% of the observed rice yield change, ranging from 42.9% to 96.5% in different cropping types, and the effect increased with the latitude. Among the management considered, sowing date adjustment contributes most to late and early rice yield. Response of rice yield to nighttime temperature was stronger than that to daytime temperature, and wind speed is the main climatic contributing factor to early rice yield. The effects of wind speed on rice yield should be considered for the adaptation measures. This observation-based evidence may help guide agricultural priorities in mitigating the impact of climate change on rice yield. Full article

The existing precision grazing technology helps to improve the utilization rate of livestock to pasture, but it is still at the level of “collectivization” and cannot provide more accurate grazing management and control. (1) Background: In recent years, with the rapid development of agent-related technologies such as deep learning, visual navigation and tracking, more and more lightweight edge computing cell target detection algorithms have been proposed. (2) Methods: In this study, the improved YOLOv5 detector combined with the extended dataset realized the accurate identification and location of domestic cattle; with the help of the kernel correlation filter (KCF) automatic tracking framework, the long-term cyclic convolution network (LRCN) was used to analyze the texture characteristics of animal fur and effectively distinguish the individual cattle. (3) Results: The intelligent UAV equipped with an AGX Xavier high-performance computing unit ran the above algorithm through edge computing and effectively realized the individual identification and positioning of cattle during the actual flight. (4) Conclusion: The UAV platform based on airborne depth reasoning is expected to help the development of smart ecological animal husbandry and provide better precision services for herdsmen. Full article