AgriEngineering, Volume 5, Issue 2 (June 2023) – 29 articles

Corn is a staple food crop grown in over 100 countries worldwide. To meet the growing demand for corn, losses in its quality and quantity should be minimized. One of the potential threats to the quality and viability of corn is mechanical damage during harvesting and handling. Despite extensive research on corn, there is a lack of reliable data on the damage its seeds undergo when they are subjected to mechanical impact against different surfaces during handling and transportation. This study is designed to investigate the effects of (a) drop height (5, 10, and 15 m) during free fall, (b) impact surface (concrete, metal, and seed to seed), seed moisture content (10, 15, 20, and 25% w.b), and ambient temperature (−10 and 20 °C) on the percentage of physical damage (PPD) and physiological damage to corn seeds. The PPD and the extent of physiological damage were determined as the percentage of seed breakage and the percentage of loss in germination (PLG), respectively. The latter parameter was specifically chosen to evaluate seeds that showed no visible external damage, thus enabling the assessment of purely internal damage that PPD did not capture. This approach enabled a comprehensive analysis of free fall’s influence on the seeds’ quality and viability, providing a complete picture of the overall impact. Total damage was then calculated as the sum of PPD and PLG. An evaluation and modeling process was undertaken to assess how corn seed damage depends on variables such as drop height, moisture content, impact surfaces, and temperatures. The results revealed that seeds dropped onto metal surfaces incurred a higher total damage (15.52%) compared to concrete (12.86%) and seed-to-seed abrasion (6.29%). Greater total damage to seeds was observed at an ambient temperature of −10 °C (13.66%) than at 20 °C (9.46%). Increased drop height increased seeds’ mass flow velocity and correspondingly caused increases in both physical and physiological damage to seeds. On the other hand, increased moisture levels caused a decreasing trend in the physical damage but increased physiological damage to the seeds. The limitations of the developed models were thoroughly discussed, providing important insights for future studies. The results of this study promise to deliver substantial benefits to the seed/grain handling industry, especially in minimizing impact-induced damage. Full article

This paper explores the potential of infrared thermography and geostatistics in animal production and presents the results of the application of the combination of these techniques, contributing significantly to efforts to obtain animals’ responses to the environments in which they are located and thereby ensuring improvements in productivity and animal welfare. The objective was to verify the variability in surface temperature in pigs submitted to different climate control systems using geostatistics. Three growing animals per stall were selected. Dry bulb temperature (Tbd, °C), relative humidity (RH, %) and thermal images were recorded at 08:00 and 12:00 h. To analyze the data, semivariograms were made, the theoretical model was validated and kriging maps were constructed. The mean temperature of the pigs in the pen with adiabatic evaporative cooling (AEC) ranged from 32.40 to 36.25 °C; for the pigs in the forced ventilation (FV) pen, the range of variation was from 32.51 to 36.81 °C. In the control group (Con), with natural ventilation, the average temperature was 37.51 to 38.45 °C. The geostatistical analysis provided a mathematical model capable of illustrating the variation in temperature in the caudal–dorsal regions of the pigs according to the environments to which the animals were subjected. Full article

Soil water content (θ) is a crucial soil parameter that is determined in many studies involving engineering, geology, and soil and environmental sciences. For instance, evaluating the soil strength, groundwater recharge, hydraulic conductivity, and soil aeration status depends on θ. The measurement of θ is fundamental for monitoring and controlling several soil processes. The gamma-ray attenuation (GRA) technique is a fast and non-destructive way of evaluating θ in soils with very contrasting compositions. Although, GRA is rarely explored in lab physics classes. The proposal of an experiment using a teaching GRA apparatus for measuring θ is presented. The experimental setup consisted of a ¹³⁷Cs radioactive source, a Geiger-Müller detector, and a radiation counter. Soil samples with four distinct granulometric compositions were analyzed. Strong linear correlations were found between the transmitted gamma-ray photon intensity and θ (correlation coefficients varying from −0.95 to −0.98). The soil porosity, measured by the conventional and GRA methods, presented differences that varied from c. 7.8% to c. 18.2%. In addition, strong linear relationships (correlation coefficients from 0.90 to 0.98) were observed between the GRA and the traditional (gravimetric) method of θ measurement. It was verified that the teaching GRA apparatus is useful for measuring θ. In addition, the apparatus allows the introduction of some important aspects related to the study of modern physics for undergraduate students of many fields of knowledge. Full article

Glyphosate is the most widely used herbicide for weed control in citrus orchards in Brazil; therefore, it is likely that several species have gained resistance to this herbicide and that more than one resistant species can be found in the same orchard. The objective was to identify weeds resistant to glyphosate in citrus orchards from different regions of the São Paulo State (SP) and determine how many resistant species are present within the same orchard. Seeds of Amaranthus deflexus, A. hybridus, Bidens pilosa, Chloris elata, Conyza bonariensis, Digitaria insularis, Solanum Americanum, and Tridax procumbens, which, as reported by growers, are suspected to be resistant to glyphosate, were collected from plants that survived the last application of this herbicide (>720 g of acid equivalent [ae] ha^–1) in sweet orange and Tahiti acid lime orchards. Based on dose–response and shikimic acid accumulation assays, all populations of A. deflexus, A. hybridus, B. pilosa, and T. procumbens were sensitive to glyphosate. However, populations of B. pilosa from the Olimpia region (R-NS, R-PT and R-OdA) showed signs of resistance based on plant mortality rates by 50% within a population (LD₅₀ = 355–460 g ae ha⁻¹). All populations of C. bonariensis, C. elata, and D. insularis were resistant to glyphosate, presenting resistance ratios from 1.9 to 27.6 and low shikimate accumulation rates. Solanum americanum also showed resistance, with resistance ratios ranging from 4.3 to 25.4. Most of the citrus orchards sampled presented the occurrence of more than one species resistant to glyphosate: Nossa Senhora—one species; Olhos D’agua and Passatempo—two species; Araras—four species; and Cordeiropolis and Mogi-Mirim—up to five species. The results reported in this paper provide evidence of multiple species in citrus orchards from São Paulo that have exhibited resistance to glyphosate. This underscores the difficulties in managing glyphosate-resistant weeds which are prevalent throughout the country, such as C. bonariensis and D. insularis. The presence of these resistant species further complicates the control of susceptible species that may also develop resistance. In addition, the glyphosate resistance of S. americanum was identified for the first time. Full article

Accurate estimates of pasture biomass in grazing lands are currently a time-consuming and resource-intensive task. The process generally includes physically cutting, bagging, labelling, drying, and weighing grass samples using multiple “quadrats” placed on the ground. Quadrats vary in size but are typically in the order of 0.25 m² (i.e., 0.5 m × 0.5 m) up to 1.0 m². Measurements from a number of harvested quadrats are then averaged to get a site estimate. This study investigated the use of photographs and ‘machine learning’ to reduce the time factor and difficulty in taking pasture biomass measurements to potentially make the estimations more accessible through the use of mobile phone cameras. A dataset was created from a pre-existing archive of quadrat photos and corresponding hand-cut pasture biomass measurements taken from a diverse range of field monitoring sites. Sites were clustered and one was held back per model for testing. The models were based on DenseNet121. Individual quadrat errors were large but more promising results were achieved when estimating the site mean pasture biomass. Another two smaller additional datasets were created post-training which were used to further assess the ensemble; they provided similar absolute errors to the original dataset, but significantly larger relative errors. The first was made from harvested quadrats, and the second was made using a pasture height meter in conjunction with a mobile phone camera. The models performed well across a variety of situations and locations but underperformed when assessed on some sites with very different vegetation. More data and refinement of the approach outlined in the paper will reduce the number of models needed and help to correct errors. These models provide a promising start, but further investigation, refinement, and data are needed before becoming a usable application. Full article

Throughout history, the use of animals for agricultural and forestry work has been closely associated with human societies, with multiple references to animal power being utilized for various tasks since the Neolithic period. However, the advent of industrialization has fundamentally transformed the reality of society, leading to a significant shift towards the mechanization of processes. Despite this, animal traction continues to play an important role as a workforce in many developing countries and developed nations, where there is a renewed interest in the use of animal traction, particularly for tasks intended to have a reduced environmental impact and a smaller carbon footprint. The present study conducted a SWOT analysis to examine the potential of animal traction as an alternative for the recovery processes of forest residual woody biomass, particularly when the use of mechanical equipment is not feasible. This can contribute to the creation of value chains for residual products, which can be harnessed for energy recovery. The utilization of modern animal traction can promote the sustainable development of projects at the local and regional level, with efficient utilization of endogenous resources and the creation of value for residual forest woody biomass. This approach can thus facilitate the optimization of supply chains, from biomass to energy. Full article

Poultry farming plays a significant role in ensuring food security and economic growth in many countries. However, various factors such as feeding management practices, environmental conditions, and diseases lead to poultry mortality (dead birds). Therefore, regular monitoring of flocks and timely veterinary assistance is crucial for maintaining poultry health, well-being, and the success of poultry farming operations. However, the current monitoring method relies on manual inspection by farm workers, which is time-consuming. Therefore, developing an automatic early mortality detection (MD) model with higher accuracy is necessary to prevent the spread of infectious diseases in poultry. This study aimed to develop, evaluate, and test the performance of YOLOv5-MD and YOLOv6-MD models in detecting poultry mortality under various cage-free (CF) housing settings, including camera height, litter condition, and feather coverage. The results demonstrated that the YOLOv5s-MD model performed exceptionally well, achieving a high mAP@0.50 score of 99.5%, a high FPS of 55.6, low GPU usage of 1.04 GB, and a fast-processing time of 0.4 h. Furthermore, this study also evaluated the models’ performances under different CF housing settings, including different levels of feather coverage, litter coverage, and camera height. The YOLOv5s-MD model with 0% feathered covering achieved the best overall performance in object detection, with the highest mAP@0.50 score of 99.4% and a high precision rate of 98.4%. However, 80% litter covering resulted in higher MD. Additionally, the model achieved 100% precision and recall in detecting hens’ mortality at the camera height of 0.5 m but faced challenges at greater heights such as 2 m. These findings suggest that YOLOv5s-MD can detect poultry mortality more accurately than other models, and its performance can be optimized by adjusting various CF housing settings. Therefore, the developed model can assist farmers in promptly responding to mortality events by isolating affected birds, implementing disease prevention measures, and seeking veterinary assistance, thereby helping to reduce the impact of poultry mortality on the industry, ensuring the well-being of poultry and the overall success of poultry farming operations. Full article

This paper presents a machine learning approach to automatically classifying post-harvest vegetal species. Color images of vegetal species were applied to convolutional neural networks (CNNs) and support vector machine (SVM) classifiers. We focused on okra as the target vegetal species and classified it into two quality types. However, our approach could also be applied to other species. The machine learning solution consists of several components, and each design process and its combinations are essential for classification quality. Therefore, we carefully investigated their effects on classification accuracy. Through our experimental evaluation, we confirmed the following: (1) in color space selection, HLG (hue, lightness, and green) and HSL (hue, saturation, and lightness) are essential for vegetal species; (2) suitable preprocessing techniques are required owing to the complexity of the data and noise load; and (3) the diversity extension of learning image data by mixing different datasets obtained under different conditions is quite effective in reducing the overfitting possibility. The results of this study will assist AI practitioners in the design and development of post-harvest classifications based on machine learning. Full article

Hydrothermal treatment of rice, called “Parboiling”, is an ancient traditional process in Asian countries. It consists of soaking rough rice in water and steaming it, and it both reduces the level of grain breakage and increases head yield of rice during milling. However, parboiling of rice is associated with some drawbacks regarding consumer preferences: the loss of its aroma, reduced rice-kernel whiteness and increased kernel hardness. This research study was carried out to develop a mild hydrothermal treatment that could be applied to basmati paddy by controlling hydrothermal treatment, i.e., soaking water temperature, steaming pressure and time. The Basmati 370 paddy variety was used for this study. The results revealed that, by soaking the paddy in hot water (70 ± 2 °C) for 75 min and steaming the soaked paddy for 20 min with non-pressurized steam at atmospheric pressure, and soaking the paddy for 120 min in hot water (70 ± 2 °C) and steaming the soaked paddy for 4 min with pressurized steam (4 kg/cm²), the optimum treatments are achieved. These optimum hydrothermal treatments were able to produce high head rice yield and preserve the basmati aroma, colour, hardness and palatability characteristics similar to non-parboiled basmati rice. Further, nutritional values such as vitamin B and protein content were also significantly preserved by these mild hydrothermal treatments. These optimized treatment combinations achieved minimized grain breakage while increasing head rice yield during milling and, at the same time, preserved basmati aroma, kernel whiteness, cooking and palatability characteristics similar to non-parboiled rice. Full article

Makeup water constitutes a key component in the water management of microalgal cultivation systems. However, the effect of makeup water addition on microalgal growth remains largely unexplored. This study compared two deionized water addition intervals (1 day and 4 days) for their effect on the growth of Scenedesmus dimorphus (S. dimorphus hereafter) in 2000 mL Pyrex bottles under controlled conditions. Cell counts and dry algal biomass (DAB) were measured to characterize the microalgal growth rate. Water addition intervals impacted algal cell counts but had little effect on DAB. Adding makeup water every day resulted in a higher growth rate (8.80 ± 1.46 × 10⁵ cells mL⁻¹ day⁻¹; p = 0.22, though) and an earlier occurrence of the peak cell count (day 9) than adding it every 4 days (6.95 ± 1.68 × 10⁵ cells mL⁻¹ day⁻¹ and day 12, respectively). It is speculated that water loss over an extended period and the following makeup water addition posed stress on S. dimorphus. Surpassing the peak cell count, S. dimorphus continued to grow in DAB, resulting in an increased cell weight as a response to nutrient starvation. Optical density at 670 nm (OD₆₇₀) was also measured. Its correlation with DAB was found to be affected by water addition intervals (R² = 0.955 for 1 day and 0.794 for 4 days), possibly due to a water loss-induced change in chlorophyll a content. This study is expected to facilitate the makeup water management of photobioreactor and open pond cultivation systems. Full article

This work presents a remote sensing application to estimate the leaf area index (LAI) in two rice (Oryza sativa L.) varieties (IDIAP 52-05 and IDIAP FL 137-11), as a proxy for crop performance. In-field, homogeneous spectroradiometric measurements (350–1050 nm) were carried in two campaigns (June–November 2017 and January–March 2018), on a private farm, TESKO, located in Juan Hombrón, Coclé Province, Panama. The spectral fingerprint of IDIAP 52-05 plants was collected in four dates (47, 67, 82 and 116 days after sowing), according to known phenological stages of rice plant growth. Moreover, true LAI or green leaf area was measured from representative plants and compared to LAI calculated from normalized PlanetScope multi-spectral satellite images (selected according to dates close to the in-field collection). Two distinct estimation models were used to establish the relationships of measured LAI and two vegetational spectral indices (NDVI and MTVI2). The results show that the MTVI2 based model has a slightly higher predictive ability of true LAI ($R^2$ = 0.92, RMSE = 2.20), than the NDVI model. Furthermore, the satellite images collected were corrected and satellite LAI was contrasted with true LAI, achieving in average 18% for Model 2 for MTVI2, with the NDVI (Model 1) corrected model having a smaller error around 13%. This work provides an important advance in precision agriculture, specifically in the monitoring of total crop growth via LAI for rice crops in the Republic of Panama. Full article

The energy in agricultural systems is two-fold: transformation and utilization. The assessment and proper use of energy in agricultural systems is important to achieve economic benefits and overall sustainability. Therefore, this study was conducted to evaluate the energy balance of crop and livestock production, net energy ratio (NER), and water use efficiency (WUE) of crops of a selected farm in Sri Lanka using the life cycle assessment (LCA) approach. In order to assess the diversification, 18 crops and 5 livestock types were used. The data were obtained from farm records, personal contacts, and previously published literature. Accordingly, the energy balance in crop production and livestock production was −316.87 GJ ha⁻¹ Year⁻¹ and 758.73 GJ Year⁻¹, respectively. The energy related WUE of crop production was 31.35 MJ m⁻³. The total energy balance of the farm was 736.2 GJ Year⁻¹. The results show a negative energy balance in crop production indicating an efficient production system, while a comparatively higher energy loss was shown from the livestock sector. The procedure followed in this study can be used to assess the energy balance of diversified agricultural systems, which is important for agricultural sustainability. This can be further developed to assess the carbon footprint in agricultural systems. Full article

Supplying high-quality fresh sweetpotato roots to the consumer requires sorting the roots by quality and removing culls deemed unsuitable for fresh markets at packing facilities. The sorting operation is traditionally performed by manual labor. This study surveyed the sorting lines of seven commercial sweetpotato packinghouses in Mississippi during the packing season of 2021. Sorting for defects entirely relied on labor, which accounted for up to 50% of the total labor in packinghouses. A cost–benefit analysis was conducted to determine the cost-effectiveness of implementing automated sorting technology as an alternative to manual sorting. The net benefits of automated sorting depended on labor savings and equipment costs. Machines at or less than USD 100,000 were economically beneficial with payback periods of less than three years when four or more workers could be replaced, while machines of USD 350,000 and higher would be not justifiable when quick economical returns were sought. Automated sorting promises to increase the profitability and competitiveness of fresh market sweetpotato packing industries. Full article

Agrometeorological models are great tools for predicting yields and improving decision-making. High-quality climatic data are essential for using these models. However, most developing countries have low-quality data with low frequency and spatial coverage. In this case, two main options are available: gathering more data in situ, which is expensive, or using gridded data, obtained from several sources. The main objective here was to evaluate the quality of two gridded climatic databases for filling gaps of real weather stations in the context of developing agrometeorological models. Therefore, a comparative analysis of gridded database and INMET data (precipitation and air temperature) was conducted using an agrometeorological model for sweet orange yield estimation. Both gridded databases had high determination and concordance coefficients for maximum and minimum temperatures. However, higher errors and lower confidence coefficients were observed for precipitation data due to their high dispersion. BR-DWGD indicated more accurate results and correlations in all scenarios evaluated in relation to NasaPower, pointing out that BR-DWGD may be better at filling gaps and providing inputs to simulate attainable yield in the Brazilian citrus belt. Nevertheless, due to the BR-DWGD database’s geographical and temporal limitations, NasaPower is still an alternative in some cases. Additionally, when using NasaPower, it is recommended to use a measured precipitation source to improve prediction quality. Full article

Piling behavior (PB) is a common issue that causes negative impacts on the health, welfare, and productivity of the flock in poultry houses (e.g., cage-free layer, breeder, and broiler). Birds pile on top of each other, and the weight of the birds can cause physical injuries, such as bruising or suffocation, and may even result in death. In addition, PB can cause stress and anxiety in the birds, leading to reduced immune function and increased susceptibility to disease. Therefore, piling has been reported as one of the most concerning production issues in cage-free layer houses. Several strategies (e.g., adequate space, environmental enrichments, and genetic selection) have been proposed to prevent or mitigate PB in laying hens, but less scientific information is available to control it so far. The current study aimed to develop and test the performance of a novel deep-learning model for detecting PB and evaluate its effectiveness in four CF laying hen facilities. To achieve this goal, the study utilized different versions of the YOLOv6 models (e.g., YOLOv6t, YOLOv6n, YOLOv6s, YOLOv6m, YOLOv6l, and YOLOv6l relu). The objectives of this study were to develop a reliable and efficient tool for detecting PB in commercial egg-laying facilities based on deep learning and test the performance of new models in research cage-free facilities. The study used a dataset comprising 9000 images (e.g., 6300 for training, 1800 for validation, and 900 for testing). The results show that the YOLOv6l relu-PB models perform exceptionally well with high average recall (70.6%), mAP@0.50 (98.9%), and mAP@0.50:0.95 (63.7%) compared to other models. In addition, detection performance increases when the camera is placed close to the PB areas. Thus, the newly developed YOLOv6l relu-PB model demonstrated superior performance in detecting PB in the given dataset compared to other tested models. Full article

Field imagery is an effective way to capture the state of the entire field; yet, current field inspection approaches, when accounting for image resolution and processing speed, using existent imaging systems, do not always enable real-time field inspection. This project involves the innovation of novel technologies by using an FPGA-based image processing (FIP) device that eliminates the technical limitations of the current agricultural imaging services available in the market and will lead to the development of a market-ready service solution. The FIP prototype developed in this study was tested in both a laboratory and outdoor environment by using a digital single-lens reflex (DSLR) camera and web camera, respectively, as the reference system. The FIP system had a high accuracy with a Lin’s concordance correlation coefficient of 0.99 and 0.91 for the DLSR and web camera reference system, respectively. The proposed technology has the potential to provide on-the-spot decisions, which in turn, will improve the compatibility and sustainability of different land-based systems. Full article

The aim of this note is to provide a quick overview of the installation and adjustment of an exclusively mechanical standalone automatic device that self-adjusts to weather changes to control the frequency and duration of the irrigation. The “hydraulic evapotranspiration multisensor” (HEM) is composed of a reduced evaporation pan with water, a magnet with a floater floating in the pan, a hydraulic device operated by a magnetic hydraulic valve that has the ability to adjust the frequency of irrigation, and a hydraulic system that returns water to the pan during each irrigation event through an adjustable dripper to replace the water lost due to the fact of evaporation. This note is particularly relevant for arid–semi-arid regions where agricultural production is fully dependent on irrigation. Full article

Bruise damage is one of the mechanical injuries that fresh produce can sustain during the postharvest supply chain. The study investigated the effect of drop impact levels, storage temperatures, and the storage period on the quality changes of cucumbers. A known mass ball was used to damage cucumbers once from three different drop heights (30, 60, and 90 cm) before they were stored for 24 days at 5 °C, 10 °C, and 22 °C. The data showed that the bruise area (BA), bruise susceptibility (BS), yellowness, and chroma* increased with the increase in the drop height and storage temperature. The study found that the bruise area (BA) and bruise susceptibility (BS) of the damaged cucumbers increased substantially (p < 0.05) with the increase in storage temperature and drop height. Due to the increment in drop height, storage temperature, and the storage period, the weight loss (Wl)% significantly increased after 24 days of storage. The storage period affects the firmness of damaged cucumbers stored in all storage conditions. The highest value of lightness (L*) was observed for the cucumbers bruised from the 60 cm drop height and stored at 22 °C with a value of 43.08 on day 24 of storage. Hue*, redness (a*), and total soluble solids (TSS) were all unaffected by the drop height. This study can serve as a resource for horticultural researchers and experts involved in the fresh fruit and vegetable supply chain. The study pays attention to the importance of postharvest supply chain activities, such as handling and storage to maintain the quality and prolong the shelf life of perishable produce, such as cucumbers. Full article

Remote sensing technology applied to agricultural crops has emerged as an efficient tool to speed up the data acquisition process in decision-making. In this study, we aimed to evaluate the performance of the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Red Edge (NDRE) in estimating biomass accumulation in common bean crops. The research was conducted at the Federal University of Lavras, where the ANFC 9 cultivar was used in an area of approximately seven hectares, in a second crop, in 2022. A total of 31 georeferenced points spaced at 50 m were chosen to evaluate height, width and green biomass, with collections on days 15, 27, 36, 58, 62 and 76 of the crop cycle. The images used in the study were obtained from the PlanetScope CubeSat satellite, with a spatial resolution of 3 m. The data obtained were subjected to a Pearson correlation (R) test and multiple linear regression analysis. The green biomass variable was significantly correlated with plant height and width. The NDVI performed better than the NDRE, with higher values observed at 62 Days After Sowing (DAS). The model that integrates the parameters of height, width and NDVI was the one that presented the best estimate for green biomass in the common bean crop. The M1 model showed the best performance to estimate green biomass during the initial stage of the crop, at 15, 27 and 36 DAS (R² = 0.93). These results suggest that remote sensing technology can be effectively applied to assess biomass accumulation in common bean crops and provide accurate data for decision-makers. Full article

Including the correct combination of factors for the application technology of pesticides can improve their distribution on their targets. The aim of this work was to use multivariate analysis to study the effect size and the order of influence of three factors that interfere with pesticide application technology in corn crops. A 2 × 2 × 3 factorial experiment was conducted with two droplet size classes (fine and coarse), two application rates (80 and 150 L ha⁻¹), and the presence of adjuvants (mineral oil one and two, and no adjuvant). A knapsack boom sprayer was used for the applications. Droplet deposition on the corn leaves was evaluated by detecting a tracer added to the spray via spectrophotometry and the droplet spectrum by analyzing water-sensitive papers. Univariate and multivariate statistical analyses were performed to integrate the variables analyzed. Droplet size has proven to be the most important factor in spraying planning, and the second factor is the application rate. With the association between fine droplets and higher application rates, a better performance was obtained in coverage, droplet density, and droplet deposition. Full article

Machine Learning (ML) has enabled many image-based object detection and recognition-based solutions in various fields and is the state-of-the-art method for these tasks currently. Therefore, it is of interest to apply this technique to different questions. In this paper, we explore whether it is possible to classify apple cultivars based on fruits using ML methods and images of the apple in question. The goal is to develop a tool that is able to classify the cultivar based on images that could be used in the field. This helps to draw attention to the variety and diversity in fruit growing and to contribute to its preservation. Classifying apple cultivars is a certain challenge in itself, as all apples are similar, while the variety within one class can be high. At the same time, there are potentially thousands of cultivars indicating that the task becomes more challenging when more cultivars are added to the dataset. Therefore, the first question is whether a ML approach can extract enough information to correctly classify the apples. In this paper, we focus on the technical requirements and prerequisites to verify whether ML approaches are able to fulfill this task with a limited number of cultivars as proof of concept. We apply transfer learning on popular image processing convolutional neural networks (CNNs) by retraining them on a custom apple dataset. Afterward, we analyze the classification results as well as possible problems. Our results show that apple cultivars can be classified correctly, but the system design requires some extra considerations. Full article

The Green Revolution led to an increased use of synthetic pesticides, causing environmental pollution. As an alternative, biopesticides made from entomopathogenic agents such as fungi have been sought. This study aimed to design and evaluate the performance of a harvester machine for efficiently collecting entomopathogenic spores of Metarhizium anisopliae and Beauveria bassiana grown on rice and corn substrates. The spore yield was estimated, and a spore count and a colony-forming unit (CFU) count were performed. Statistical analysis was conducted to compare the mean values of spores obtained from different combinations of solid substrate and fungi. The Corn-Metarhizium combination produced 34.15 g of spores per kg of substrate and 1.51 × 10⁹ CFUs mL⁻¹. Similarly, the Rice-Metarhizium combination produced 57.35 g per kg and 1.59 × 10⁹ CFUs mL⁻¹. Meanwhile, the Corn-Beauveria combination yielded 35.47 g per kg and 1.00 × 10⁹ CFUs mL⁻¹, while the Rice-Beauveria combination had a yield of 38.26 g per kg and 4.50 × 10⁸ CFUs mL⁻¹. Based on the reported results, the Rice-Metarhizium combination appears to be the most effective, yielding the highest number of harvested spores per kg of substrate. The study estimated a total cost of approximately $409.31 for manufacturing the harvester, considering only the cost of the materials. These results could potentially increase the availability and affordability of entomopathogenic fungi in integrated pest management. Full article

Developing different robotic platforms for farm operations is vital to addressing the increasing world population. A harvesting robot significantly increases a farm’s productivity while farmers focus on other relevant farm operations. From the literature, it could be summarized that the design concepts of the harvesting mechanisms were categorized as grasping and cutting, vacuum suction plucking systems, twisting and plucking mechanisms, and shaking and catching. Meanwhile, robotic system components include the mobile platform, manipulators, and end effectors, sensing and localization, and path planning and navigation. The robotic system must be cost-effective and safe. The findings of this research could contribute to the design process of developing a harvesting robot or developing a harvesting module that can be retrofitted to a commercially available mobile platform. This paper provides an overview of the most recent harvesting robots’ different concept designs and system components. In particular, this paper will highlight different agricultural ground mobile platforms and their associated mechanical design, principles, challenges, and limitations to characterize the crop environment relevant to robotic harvesting and to formulate directions for future research and development for cotton harvesting platforms. Full article

Deferred grazing (DG) consists in adapting the number of animals and the number of days grazed to the availability of pasture. Compared to continuous grazing (CG), which is based on a permanent and low stocking rate, DG is a management strategy that aims at optimizing the use of the resources available in the Mediterranean Montado ecosystem. This study with sheep grazing, carried out between 2019 and 2021 on a 4 ha pasture in Alentejo region of the Southern of Portugal, assesses the impact of these two grazing management systems on soil compaction as a result of animal trampling. This area of native natural grassland (a dryland pasture, mixture of grasses, legumes, and composite species) was divided into four grazing parks of 1 ha each, two under DG management and two under CG management. At the end of the study, the cone index (CI, in kPa) was measured in the topsoil layer (0–30 cm) with an electronic cone penetrometer at 48 georeferenced areas (12 in each park). The results of CI measurement showed no significant differences between treatments in all depths measured (0–10, 10–20, and 20–30 cm). These findings are encouraging from the point of view of soil conservation and sustainability, revealing good prospects for the intensification of extensive livestock production. Future work should evaluate the long-term impact and consider, at the same time, other ecosystem services and system productivity indicators. Full article

Single-frame circulation aquaculture belongs to the important category of sustainable agriculture development. In light of the visual-detection problem related to survival rate of Portunus in single-frame three-dimensional aquaculture, a fusion recognition algorithm based on YOLOV5, RCN (RefineContourNet) image recognition of residual bait ratio, centroid moving distance, and rotation angle was put forward. Based on three-parameter identification and LWLR (Local Weighted Linear Regression), the survival rate model of each parameter of Portunus was established, respectively. Then, the softmax algorithm was used to obtain the classification and judgment fusion model of Portunus’ survival rate. In recognition of the YOLOV5 residual bait and Portunus centroid, the EIOU (Efficient IOU) loss function was used to improve the recognition accuracy of residual bait in target detection. In RCN, Portunus edge detection and recognition, the optimized binary cross-entropy loss function based on double thresholds successfully improved the edge clarity of the Portunus contour. The results showed that after optimization, the mAP (mean Average Precision) of YOLOV5 was improved, while the precision and mAP (threshold 0.5:0.95:0.05) of recognition between the residual bait and Portunus centroid were improved by 2% and 1.8%, respectively. The loss of the optimized RCN training set was reduced by 4%, and the rotation angle of Portunus was obtained using contour. The experiment shows that the recognition accuracy of the survival rate model was 0.920, 0.840, and 0.955 under the single parameters of centroid moving distance, residual bait ratio, and rotation angle, respectively; and the recognition accuracy of the survival rate model after multi-feature parameter fusion was 0.960. The accuracy of multi-parameter fusion was 5.5% higher than that of single-parameter (average accuracy). The fusion of multi-parameter relative to the single-parameter (average) accuracy was a higher percentage. Full article

Drought has been known to be a natural hazard reflecting geographic and climatic characteristics. Satellite technology advancements have benefited drought assessment and monitoring to formulate plans for dealing with this slow-onset disaster. However, combining satellite remote sensing (RS) and meteorological data for drought monitoring is lacking in the literature. This study uses satellite RS and meteorological-based drought indicators to assess drought risk in the Ilocos Norte, Philippines. Data analysis included the retrieval of vegetation conditions using Sentinel-1 and Sentinel-2 data. The standardized precipitation index (SPI) and Keetch–Byram drought index (KBDI) were calculated to account for climatic variabilities. Results revealed that the Sentinel-1 backscatter coefficient decreased by −2 dB in the cropland area, indicating crop growth irregularities compared to grassland areas. These irregularities were supported by Sentinel-2 normalized difference vegetation index (NDVI) strong fluctuations during the two-year observation period. A significant coefficient of determination (R² > 0.60) between the Sentinel-1 backscatter coefficient and Sentinel-2 NDVI was observed for the study area. On the one hand, only KBDI significantly correlated (R² > 0.60) with the cropland area’s RS data-derived drought indicators. These results revealed RS data variability for drought risk management but are still valuable for developing an early warning system. Full article

Currently, Brazil is the leading producer of sugarcane in the world, with self-sufficiency in the use of ethanol as a biofuel, as well as being one of the largest suppliers of sugar to the world. This study aimed to develop a predictive model for sugarcane production based on data extracted from aerial imagery obtained from drones or satellites, allowing the precise tracking of plant development in the field. A model based on a semiparametric approach associated with the inverse Gaussian distribution applied to vegetation indices (VIs), such as the Normalized Difference Vegetation Index (NDVI) and Visible Atmospherically Resistant Index (VARI), was developed with data from drone images obtained from two field experiments with randomized replications and four sugarcane varieties. These experiments were performed under conditions identical to those applied by sugarcane farmers. Further, the model validation was carried out by scaling up the analyses with data extracted from Sentinel-2 images of several commercial sugarcane fields. Very often, in countries such as Brazil, sugarcane crops occupy extensive areas. Consequently, the development of tools capable of being operated remotely automatically benefits the management of this crop in the field by avoiding laborious and time-consuming sampling and by promoting the reduction of operation costs. The results of the model application in both sources of data, i.e., data from field experiments as well as the data from commercial fields, showed a suitable level of overlap between the data of predicted yield using VIs generated from drone and satellite images with the data of verified yield obtained by measuring the production of experiments and commercial fields, indicating that the model is reliable for forecasting productivity months before the harvest time. Full article

The difference in the matrix present in soil samples from different areas limits the performance of nutrient analysis via XRF sensors, and only a few strategies to mitigate this effect to ensure an accurate analysis have been proposed so far. In this context, this research aimed to compare the performance of different predictive models, including a simple linear regression (RS), multiple linear regression (MLR), partial least-squares regression (PLS), and random forest (RF) models for the prediction of Ca and K in agricultural soils. RS models were evaluated on XRF data without (RS1) and with (RS2) Compton normalization. In addition, it was assessed whether using soil texture information and/or vis–NIR spectra as auxiliary variables would optimize the predictive performance of the models. The results showed that all strategies allowed the mitigation of the matrix effect to some degree, enabling the determination of their Ca and K contents with excellent predictive performance (R² ≥ 0.84). The best performance was obtained using RS2 for the Ca prediction (R² = 0.92, RSME = 48.25 mg kg⁻¹ and relative improvement (RI) of 52.3% compared to RS1) and using an RF for the K prediction (R² = 0.84, RSME = 17.43 mg kg⁻¹ and RI of 24.3% compared to RS1). The results indicated that sophisticated models did not always perform better than linear models. Furthermore, using texture data and vis–NIR spectra as auxiliary data was promising only for the K prediction, which showed an error reduction in the order of 10%, contrasting with the Ca prediction, which did not reduce the prediction error by more than 1%. The best modeling approach in our study proved to be attribute-specific. These results give further insight into the development of intelligence modeling strategies for sensor-based soil analysis. Full article

Recently, Panama wilt disease that attacks banana leaves has caused enormous economic losses to farmers. Early detection of this disease and necessary preventive measures can avoid economic damage. This paper proposes an improved method to predict Panama wilt disease based on symptoms using an agro deep learning algorithm. The proposed deep learning model for detecting Panama wilts disease is essential because it can help accurately identify infected plants in a timely manner. It can be instrumental in large-scale agricultural operations where Panama wilts disease could spread quickly and cause significant crop loss. Additionally, deep learning models can be used to monitor the effectiveness of treatments and help farmers make informed decisions about how to manage the disease best. This method is designed to predict the severity of the disease and its consequences based on the arrangement of color and shape changes in banana leaves. The present proposed method is compared with its previous methods, and it achieved 91.56% accuracy, 91.61% precision, 88.56% recall and 81.56% F1-score. Full article