Research

20 pages, 16707 KiB

Open AccessArticle

Moving Vehicle Tracking with a Moving Drone Based on Track Association

by Seokwon Yeom and Don-Ho Nam

Appl. Sci. 2021, 11(9), 4046; https://doi.org/10.3390/app11094046 - 29 Apr 2021

Cited by 11 | Viewed by 4455

The drone has played an important role in security and surveillance. However, due to the limited computing power and energy resources, more efficient systems are required for surveillance tasks. In this paper, we address detection and tracking of moving vehicles with a small [...] Read more.

The drone has played an important role in security and surveillance. However, due to the limited computing power and energy resources, more efficient systems are required for surveillance tasks. In this paper, we address detection and tracking of moving vehicles with a small drone. A moving object detection scheme has been developed based on frame registration and subtraction followed by morphological filtering and false alarm removing. The center position of the detected object area is the input to the tracking target as a measurement. The Kalman filter estimates the position and velocity of the target based on the measurement nearest to the state prediction. We propose a new data association scheme for multiple measurements on a single target. This track association method consists of the hypothesis testing between two tracks and track fusion through track selection and termination. We reduce redundant tracks on the same target and maintain the track with the least estimation error. In the experiment, drones flying at an altitude of 150 m captured two videos in an urban environment. There are a total of 9 and 23 moving vehicles in each video; the detection rates are 92% and 89%, respectively. The number of valid tracks is significantly reduced from 13 to 10 and 56 to 26 in the first and the second video, respectively. In the first video, the average position RMSE of two merged tracks are improved by 83.6% when only the fused states are considered. In the second video, the average position and velocity RMSE are 1.21 m and 1.97 m/s, showing the robustness of the proposed system. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

16 pages, 10489 KiB

Open AccessArticle

3D Snow Sculpture Reconstruction Based on Structured-Light 3D Vision Measurement

by Wancun Liu, Liguo Zhang, Xiaolin Zhang and Lianfu Han

Appl. Sci. 2021, 11(8), 3324; https://doi.org/10.3390/app11083324 - 07 Apr 2021

Cited by 6 | Viewed by 1849

Abstract

Structured-light technique is an effective method for indoor 3D measurement, but it is hard to obtain ideal results outdoors because of complex illumination interference on sensors. This paper presents a 3D vision measurement method based on digital image processing to improve resistance to [...] Read more.

Structured-light technique is an effective method for indoor 3D measurement, but it is hard to obtain ideal results outdoors because of complex illumination interference on sensors. This paper presents a 3D vision measurement method based on digital image processing to improve resistance to noise of measuring systems, which ensuresnormal operation of a structured-light sensor in the wild without changing its components, and the method is applied in 3D reconstruction of snow sculpture. During image preprocessing, an optimal weight function is designed based on noise classification and minimum entropy, and the color images are transformed into monochromatic value images to eliminate most environmental noise. Then a Decision Tree Model (DTM) in a spatial-temporal context of video sequence is used to extract and track stripe. The model is insensitive to stubborn noise and reflection in the images, and the result of the model after coordinate transformation is a 3D point cloud of the corresponding snow sculpture. In experimental results, the root mean square (RMS) error and mean error are less than 0.722 mm and 0.574 mm respectively, showing that the method can realize real-time, robust and accurate measurement under a complex illumination environment, and can therefore provide technical support for snow sculpture 3D measurement. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

17 pages, 3565 KiB

Open AccessArticle

Classifying Upper Arm Gym-Workouts via Convolutional Neural Network by Imputing a Biopotential-Kinematic Relationship

by Ji-Hyeon Yoo, Ho-Jin Jung, Yi-Sue Jung, Yoon-Bee Kim, Chang-Jae Lee, Sung-Tae Shin and Han-Ul Yoon

Appl. Sci. 2021, 11(6), 2845; https://doi.org/10.3390/app11062845 - 22 Mar 2021

Cited by 1 | Viewed by 1927

Abstract

This paper proposes a systemic approach to upper arm gym-workout classification according to spatio-temporal features depicted by biopotential as well as joint kinematics. The key idea of the proposed approach is to impute a biopotential-kinematic relationship by merging the joint kinematic data into [...] Read more.

This paper proposes a systemic approach to upper arm gym-workout classification according to spatio-temporal features depicted by biopotential as well as joint kinematics. The key idea of the proposed approach is to impute a biopotential-kinematic relationship by merging the joint kinematic data into a multichannel electromyography signal and visualizing the merged biopotential-kinematic data as an image. Under this approach, the biopotential-kinematic relationship can be imputed by counting on the functionality of a convolutional neural network: an automatic feature extractor followed by a classifier. First, while a professional trainer is demonstrating upper arm gym-workouts, electromyography and joint kinematic data are measured by an armband-type surface electromyography (sEMG) sensor and a RGB-d camera, respectively. Next, the measured data are augmented by adopting the amplitude adjusted Fourier Transform. Then, the augmented electromyography and joint kinematic data are visualized as one image by merging and calculating pixel components in three different ways. Lastly, for each visualized image type, upper arm gym-workout classification is performed via the convolutional neural network. To analyze classification accuracy, two-way rANOVA is performed with two factors: the level of data augmentation and visualized image type. The classification result substantiates that a biopotential-kinematic relationship can be successfully imputed by merging joint kinematic data in-between biceps- and triceps-electromyography channels and visualizing as a time-series heatmap image. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

16 pages, 8096 KiB

Open AccessArticle

Multi-Scale Ensemble Learning for Thermal Image Enhancement

by Yuseok Ban and Kyungjae Lee

Appl. Sci. 2021, 11(6), 2810; https://doi.org/10.3390/app11062810 - 22 Mar 2021

Cited by 5 | Viewed by 1812

Abstract

In this study, we propose a multi-scale ensemble learning method for thermal image enhancement in different image scale conditions based on convolutional neural networks. Incorporating the multiple scales of thermal images has been a tricky task so that methods have been individually trained [...] Read more.

In this study, we propose a multi-scale ensemble learning method for thermal image enhancement in different image scale conditions based on convolutional neural networks. Incorporating the multiple scales of thermal images has been a tricky task so that methods have been individually trained and evaluated for each scale. However, this leads to the limitation that a network properly operates on a specific scale. To address this issue, a novel parallel architecture leveraging the confidence maps of multiple scales have been introduced to train a network that operates well in varying scale conditions. The experimental results show that our proposed method outperforms the conventional thermal image enhancement methods. The evaluation is presented both quantitatively and qualitatively. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

11 pages, 1209 KiB

Open AccessArticle

An Adaptive Optimization Method Based on Learning Rate Schedule for Neural Networks

by Dokkyun Yi, Sangmin Ji and Jieun Park

Appl. Sci. 2021, 11(2), 850; https://doi.org/10.3390/app11020850 - 18 Jan 2021

Cited by 2 | Viewed by 1584

Abstract

Artificial intelligence (AI) is achieved by optimizing the cost function constructed from learning data. Changing the parameters in the cost function is an AI learning process (or AI learning for convenience). If AI learning is well performed, then the value of the cost [...] Read more.

Artificial intelligence (AI) is achieved by optimizing the cost function constructed from learning data. Changing the parameters in the cost function is an AI learning process (or AI learning for convenience). If AI learning is well performed, then the value of the cost function is the global minimum. In order to obtain the well-learned AI learning, the parameter should be no change in the value of the cost function at the global minimum. One useful optimization method is the momentum method; however, the momentum method has difficulty stopping the parameter when the value of the cost function satisfies the global minimum (non-stop problem). The proposed method is based on the momentum method. In order to solve the non-stop problem of the momentum method, we use the value of the cost function to our method. Therefore, as the learning method processes, the mechanism in our method reduces the amount of change in the parameter by the effect of the value of the cost function. We verified the method through proof of convergence and numerical experiments with existing methods to ensure that the learning works well. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

27 pages, 8237 KiB

Open AccessArticle

The Influence of Image Processing and Layer-to-Background Contrast on the Reliability of Flatbed Scanner-Based Characterisation of Additively Manufactured Layer Contours

by David Blanco, Pedro Fernández, Alejandro Fernández, Braulio J. Alvarez and José Carlos Rico

Appl. Sci. 2021, 11(1), 178; https://doi.org/10.3390/app11010178 - 27 Dec 2020

Cited by 5 | Viewed by 1831

Abstract

Flatbed scanners (FBSs) provide non-contact scanning capabilities that could be used for the on-machine verification of layer contours in additive manufacturing (AM) processes. Layer-wise contour deviation assessment could be critical for dimensional and geometrical quality improvement of AM parts, because it would allow [...] Read more.

Flatbed scanners (FBSs) provide non-contact scanning capabilities that could be used for the on-machine verification of layer contours in additive manufacturing (AM) processes. Layer-wise contour deviation assessment could be critical for dimensional and geometrical quality improvement of AM parts, because it would allow for close-loop error compensation strategies. Nevertheless, contour characterisation feasibility faces many challenges, such as image distortion compensation or edge detection quality. The present work evaluates the influence of image processing and layer-to-background contrast characteristics upon contour reconstruction quality, under a metrological perspective. Considered factors include noise filtering, edge detection algorithms, and threshold levels, whereas the distance between the target layer and the background is used to generate different contrast scenarios. Completeness of contour reconstruction is evaluated by means of a coverage factor, whereas its accuracy is determined by comparison with a reference contour digitised in a coordinate measuring machine. Results show that a reliable contour characterisation can be achieved by means of a precise adjustment of image processing parameters under low layer-to-background contrast variability. Conversely, under anisotropic contrast conditions, the quality of contour reconstruction severely drops, and the compromise between coverage and accuracy becomes unbalanced. These findings indicate that FBS-based characterisation of AM layers will demand developing strategies that minimise the influence of anisotropy in layer-to-background contrast. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

21 pages, 8987 KiB

Open AccessArticle

A Study on Tensile Strain Distribution and Fracture Coordinate of Nanofiber Mat by Digital Image Correlation System

by Nak Gyu Park, Kyung Min Hong and Kyu Hyeung Kwon

Appl. Sci. 2020, 10(17), 5992; https://doi.org/10.3390/app10175992 - 29 Aug 2020

Cited by 2 | Viewed by 2332

Abstract

Strain gauges are commonly used for tension tests to obtain the strain of a metal test specimen. They make contact, however, so the gauges are not applicable to every type of test specimen. That is the reason why a non-contact type measurement system [...] Read more.

Strain gauges are commonly used for tension tests to obtain the strain of a metal test specimen. They make contact, however, so the gauges are not applicable to every type of test specimen. That is the reason why a non-contact type measurement system is required. Nanofibrous mats, manufactured by electrospinning, have different structures and thicknesses. Displacement and strain distributions for all ranges of the specimen have never been demonstrated for nanofiber mats so far. Wrinkled nanofibrous mats of polyurethane were made and then tension-tested. The Digital Image Correlation (DIC) method was employed to measure displacement, then to calculate strain for all areas of the specimen. The DIC system consisted of a CMOS camera, control PC and operating software with a DIC algorithm: then, the Center of Gravity (COG) algorithm was used for this system. A cross-head speed of 3 mm/min was set for the tension test. The image record speed was one frame a second. In total, 400 image frames were obtained from the start, and then displacement and strain distributions were acquired for a 400 second tension test. The strain distribution from DIC system showed good agreement with the test result by a universal testing machine. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

15 pages, 6109 KiB

Open AccessArticle

Automatic Tortuosity Estimation of Nerve Fibers and Retinal Vessels in Ophthalmic Images

by Honghan Chen, Bang Chen, Dan Zhang, Jiong Zhang, Jiang Liu and Yitian Zhao

Appl. Sci. 2020, 10(14), 4788; https://doi.org/10.3390/app10144788 - 12 Jul 2020

Cited by 1 | Viewed by 2828

Abstract

The tortuosity changes of curvilinear anatomical organs such as nerve fibers or vessels have a close relationship with a number of diseases. Therefore, the automatic estimation and representation of the tortuosity is desired in medical image for such organs. In this paper, an [...] Read more.

The tortuosity changes of curvilinear anatomical organs such as nerve fibers or vessels have a close relationship with a number of diseases. Therefore, the automatic estimation and representation of the tortuosity is desired in medical image for such organs. In this paper, an automated framework for tortuosity estimation is proposed for corneal nerve and retinal vessel images. First, the weighted local phase tensor-based enhancement method is employed and the curvilinear structure is extracted from raw image. For each curvilinear structure with a different position and orientation, the curvature is measured by the exponential curvature estimation in the 3D space. Then, the tortuosity of an image is calculated as the weighted average of all the curvilinear structures. Our proposed framework has been evaluated on two corneal nerve fiber datasets and one retinal vessel dataset. Experiments on three curvilinear organ datasets demonstrate that our proposed tortuosity estimation method achieves a promising performance compared with other state-of-the-art methods in terms of accuracy and generality. In our nerve fiber dataset, the method achieved overall accuray of 0.820, and 0.734, 0.881 for sensitivity and specificity, respectively. The proposed method also achieved Spearman correlation scores 0.945 and 0.868 correlated with tortuosity grading ground truth for arteries and veins in the retinal vessel dataset. Furthermore, the manual labeled 403 corneal nerve fiber images with different levels of tortuosity, and all of them are also released for public access for further research. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

13 pages, 14056 KiB

Open AccessArticle

Deep Learning-Based Wrapped Phase Denoising Method for Application in Digital Holographic Speckle Pattern Interferometry

by Ketao Yan, Lin Chang, Michalis Andrianakis, Vivi Tornari and Yingjie Yu

Appl. Sci. 2020, 10(11), 4044; https://doi.org/10.3390/app10114044 - 12 Jun 2020

Cited by 29 | Viewed by 3659

Abstract

This paper presents a new processing method for denoising interferograms obtained by digital holographic speckle pattern interferometry (DHSPI) to serve in the structural diagnosis of artworks. DHSPI is a non-destructive and non-contact imaging method that has been successfully applied to the structural diagnosis [...] Read more.

This paper presents a new processing method for denoising interferograms obtained by digital holographic speckle pattern interferometry (DHSPI) to serve in the structural diagnosis of artworks. DHSPI is a non-destructive and non-contact imaging method that has been successfully applied to the structural diagnosis of artworks by detecting hidden subsurface defects and quantifying the deformation directly from the surface illuminated by coherent light. The spatial information of structural defects is mostly delivered as local distortions interrupting the smooth distribution of intensity during the phase-shifted formation of fringe patterns. Distortions in fringe patterns are recorded and observed from the estimated wrapped phase map, but the inevitable electronic speckle noise directly affects the quality of the image and consequently the assessment of defects. An effective method for denoising DHSPI wrapped phase based on deep learning is presented in this paper. Although a related method applied to interferometry for reducing Gaussian noise has been introduced, it is not suitable for application in DHSPI to reduce speckle noise. Thus, the paper proposes a new method to remove speckle noise in the wrapped phase. Simulated data and experimental captured data from samples prove that the proposed method can effectively reduce the speckle noise of the DHSPI wrapped phase to extract the desired information. The proposed method is helpful for accurately detecting defects in complex defect topography maps and may help to accelerate defect detection and characterization procedures. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

18 pages, 3652 KiB

Open AccessArticle

Renyi’s Entropy Based Multilevel Thresholding Using a Novel Meta-Heuristics Algorithm

by Wei Liu, Yongkun Huang, Zhiwei Ye, Wencheng Cai, Shuai Yang, Xiaochun Cheng and Ibrahim Frank

Appl. Sci. 2020, 10(9), 3225; https://doi.org/10.3390/app10093225 - 06 May 2020

Cited by 31 | Viewed by 3051

Abstract

Multi-level image thresholding is the most direct and effective method for image segmentation, which is a key step for image analysis and computer vision, however, as the number of threshold values increases, exhaustive search does not work efficiently and effectively and evolutionary algorithms [...] Read more.

Multi-level image thresholding is the most direct and effective method for image segmentation, which is a key step for image analysis and computer vision, however, as the number of threshold values increases, exhaustive search does not work efficiently and effectively and evolutionary algorithms often fall into a local optimal solution. In the paper, a meta-heuristics algorithm based on the breeding mechanism of Chinese hybrid rice is proposed to seek the optimal multi-level thresholds for image segmentation and Renyi’s entropy is utilized as the fitness function. Experiments have been run on four scanning electron microscope images of cement and four standard images, moreover, it is compared with other six classical and novel evolutionary algorithms: genetic algorithm, particle swarm optimization algorithm, differential evolution algorithm, ant lion optimization algorithm, whale optimization algorithm, and salp swarm algorithm. Meanwhile, some indicators, including the average fitness values, standard deviation, peak signal to noise ratio, and structural similarity index are used as evaluation criteria in the experiments. The experimental results show that the proposed method prevails over the other algorithms involved in the paper on most indicators and it can segment cement scanning electron microscope image effectively. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

17 pages, 11593 KiB

Open AccessArticle

Fine-Grained Action Recognition by Motion Saliency and Mid-Level Patches

by Fang Liu, Liang Zhao, Xiaochun Cheng, Qin Dai, Xiangbin Shi and Jianzhong Qiao

Appl. Sci. 2020, 10(8), 2811; https://doi.org/10.3390/app10082811 - 18 Apr 2020

Cited by 9 | Viewed by 2156

Abstract

Effective extraction of human body parts and operated objects participating in action is the key issue of fine-grained action recognition. However, most of the existing methods require intensive manual annotation to train the detectors of these interaction components. In this paper, we represent [...] Read more.

Effective extraction of human body parts and operated objects participating in action is the key issue of fine-grained action recognition. However, most of the existing methods require intensive manual annotation to train the detectors of these interaction components. In this paper, we represent videos by mid-level patches to avoid the manual annotation, where each patch corresponds to an action-related interaction component. In order to capture mid-level patches more exactly and rapidly, candidate motion regions are extracted by motion saliency. Firstly, the motion regions containing interaction components are segmented by a threshold adaptively calculated according to the saliency histogram of the motion saliency map. Secondly, we introduce a mid-level patch mining algorithm for interaction component detection, with object proposal generation and mid-level patch detection. The object proposal generation algorithm is used to obtain multi-granularity object proposals inspired by the idea of the Huffman algorithm. Based on these object proposals, the mid-level patch detectors are trained by K-means clustering and SVM. Finally, we build a fine-grained action recognition model using a graph structure to describe relationships between the mid-level patches. To recognize actions, the proposed model calculates the appearance and motion features of mid-level patches and the binary motion cooperation relationships between adjacent patches in the graph. Extensive experiments on the MPII cooking database demonstrate that the proposed method gains better results on fine-grained action recognition. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

15 pages, 7391 KiB

Open AccessArticle

Compact and Accurate Scene Text Detector

by Minjun Jeon and Young-Seob Jeong

Appl. Sci. 2020, 10(6), 2096; https://doi.org/10.3390/app10062096 - 20 Mar 2020

Cited by 25 | Viewed by 2959

Abstract

Scene text detection is the task of detecting word boxes in given images. The accuracy of text detection has been greatly elevated using deep learning models, especially convolutional neural networks. Previous studies commonly aimed at developing more accurate models, but their models became [...] Read more.

Scene text detection is the task of detecting word boxes in given images. The accuracy of text detection has been greatly elevated using deep learning models, especially convolutional neural networks. Previous studies commonly aimed at developing more accurate models, but their models became computationally heavy and worse in efficiency. In this paper, we propose a new efficient model for text detection. The proposed model, namely Compact and Accurate Scene Text detector (CAST), consists of MobileNetV2 as a backbone and balanced decoder. Unlike previous studies that used standard convolutional layers as a decoder, we carefully design a balanced decoder. Through experiments with three well-known datasets, we then demonstrated that the balanced decoder and the proposed CAST are efficient and effective. The CAST was about 1.1x worse in terms of the F1 score, but 30∼115x better in terms of floating-point operations per second (FLOPS). Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

17 pages, 6047 KiB

Open AccessArticle

Improved Cost Computation and Adaptive Shape Guided Filter for Local Stereo Matching of Low Texture Stereo Images

by Hua Liu, Rui Wang, Yuanping Xia and Xiaoming Zhang

Appl. Sci. 2020, 10(5), 1869; https://doi.org/10.3390/app10051869 - 09 Mar 2020

Cited by 24 | Viewed by 3247

Abstract

Dense stereo matching has been widely used in photogrammetry and computer vision applications. Even though it has a long research history, dense stereo matching is still challenging for occluded, textureless and discontinuous regions. This paper proposed an efficient and effective matching cost measurement [...] Read more.

Dense stereo matching has been widely used in photogrammetry and computer vision applications. Even though it has a long research history, dense stereo matching is still challenging for occluded, textureless and discontinuous regions. This paper proposed an efficient and effective matching cost measurement and an adaptive shape guided filter-based matching cost aggregation method to improve the stereo matching performance for large textureless regions. At first, an efficient matching cost function combining enhanced image gradient-based matching cost and improved census transform-based matching cost is introduced. This proposed matching cost function is robust against radiometric variations and textureless regions. Following this, an adaptive shape cross-based window is constructed for each pixel and a modified guided filter based on this adaptive shape window is implemented for cost aggregation. The final disparity map is obtained after disparity selection and multiple steps disparity refinement. Experiments were conducted on the Middlebury benchmark dataset to evaluate the effectiveness of the proposed cost measurement and cost aggregation strategy. The experimental results demonstrated that the average matching error rate on Middlebury standard image pairs is 9.40%. Compared with the traditional guided filter-based stereo matching method, the proposed method achieved a better matching result in textureless regions. Full article

(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume II)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Intelligent Processing on Image and Optical Information, Volume II

Share This Special Issue

Special Issue Editor

Special Issue Information

Keywords

Published Papers (13 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI