Search

Due to their persistent widespread severe winds, derechos pose significant threats to human safety and property, and they are as hazardous and fatal as many tornadoes and hurricanes. Yet, automated detection of derechos remains challenging due to the absence of spatiotemporally continuous observations and the complex criteria employed to define the phenomenon. This study proposes a physically based definition of derechos that contains the key features of derechos described in the literature and allows their automated objective identification using either observations or model simulations. The automated detection is composed of three algorithms: the Flexible Object Tracker algorithm to track mesoscale convective systems (MCSs), a semantic segmentation convolutional neural network to identify bow echoes, and a comprehensive algorithm to classify MCSs as derechos or non-derecho events. Using the new approach, we develop a novel high-resolution (4 km and hourly) observational dataset of derechos over the United States east of the Rocky Mountains from 2004 to 2021. The dataset is analyzed to document the derecho climatology in the United States. Many more derechos (increased by ~50–400 %) are identified in the dataset (~31 events per year) than in previous estimations (~6–21 events per year), but the spatial distribution and seasonal variation patterns resemble earlier studies with a peak occurrence in the Great Plains and Midwest during the warm season. In addition, around 20 % of damaging gust (≥ 25.93 m s-1) reports are produced by derechos during the dataset period over the United States east of the Rocky Mountains. The dataset is available at https://doi.org/10.5281/zenodo.10884046 (Li et al., 2024). [ABSTRACT FROM AUTHOR]

Recently, the offline‐trained Siamese pipeline has drawn wide attention due to its outstanding tracking performance. However, the existing Siamese trackers utilise offline training to extract 'universal' features, which is insufficient to effectively distinguish between the target and fluctuating interference in embedding the information of the two branches, leading to inaccurate classification and localisation. In addition, the Siamese trackers employ a pre‐defined scale for cropping the search candidate region based on the previous frame's result, which might easily introduce redundant background noise (clutter, similar objects etc.), affecting the tracker's robustness. To solve these problems, the authors propose two novel sub‐network spatial employed to spatial feature embedding for robust object tracking. Specifically, the proposed spatial remapping (SRM) network enhances the feature discrepancy between target and distractor categories by online remapping, and improves the discriminant ability of the tracker on the embedding space. The MAML is used to optimise the SRM network to ensure its adaptability to complex tracking scenarios. Moreover, a temporal information proposal‐guided (TPG) network that utilises a GRU model to dynamically predict the search scale based on temporal motion states to reduce potential background interference is introduced. The proposed two network is integrated into two popular trackers, namely SiamFC++ and TransT, which achieve superior performance on six challenging benchmarks, including OTB100, VOT2019, UAV123, GOT10K, TrackingNet and LaSOT, TrackingNet and LaSOT denoting them as SiamSRMC and SiamSRMT, respectively. Moreover, the proposed trackers obtain competitive tracking performance compared with the state‐of‐the‐art trackers in the attribute of background clutter and similar object, validating the effectiveness of our method. [ABSTRACT FROM AUTHOR]

The large number of near-Earth asteroids (NEAs) has greatly impacted human space activities and Earth security. However, detecting NEAs in astronomical images with complex, varying backgrounds is still extremely challenging. In this paper, we propose a deep segmentation assisted asteroid tracking algorithm, termed DSAT, to construct a possible pipeline for faint NEA tracking in astronomical images. First, the single-frame object detection problem is converted to a segmentation problem, enabling robust extraction of faint potential moving objects. Then, a multiframe motion prior-based moving object tracking algorithm is proposed to find real NEAs. We further propose a distance tolerance criterion to help DSAT achieve effective tracking in practical situations when detection has partially failed. Finally, the pipeline is tested with both simulated and real astronomical images at different SNRs and in crowded fields. The results showed that our pipeline has the potential to detect and track faint NEAs in complex backgrounds. Our code is publicly available at https://github.com/zhenhongdu/DeepSegAsteroidTracker. [ABSTRACT FROM AUTHOR]

The Kalman Filter based on uniform assumption has been a crucial motion estimation module in trackers. However, it has limitations in non-uniform motion modeling and computational efficiency when applied to large-scale object tracking scenarios. To address these issues, we propose a novel Parallel Kalman Filter (PKF), which simplifies conventional state variables to reduces computational load and enable effective non-uniform modeling. Within PKF, we propose a non-uniform formulation which models non-uniform motion as uniform motion by transforming the time interval Δ t from a constant into a variable related to displacement, and incorporate a deceleration strategy into the control-input model of the formulation to tackle the escape problem in Multi-Object Tracking (MOT); an innovative parallel computation method is also proposed, which transposes the computation graph of PKF from the matrix to the quadratic form, significantly reducing the computational load and facilitating parallel computation between distinct tracklets via CUDA, thus making the time consumption of PKF independent of the input tracklet scale, i.e., O(1). Based on PKF, we introduce Fast, the first fully GPU-based tracker paradigm, which significantly enhances tracking efficiency in large-scale object tracking scenarios; and FastTrack, the MOT system composed of Fast and a general detector, offering high efficiency and generality. Within FastTrack, Fast only requires bounding boxes with scores and class ids for a single association during one iteration, and introduces innovative GPU-based tracking modules, such as an efficient GPU 2D-array data structure for tracklet management, a novel cost matrix implemented in CUDA for automatic association priority determination, a new association metric called HIoU, and the first implementation of the Auction Algorithm in CUDA for the asymmetric assignment problem. Experiments show that the average time per iteration of PKF on a GTX 1080Ti is only 0.2 ms; Fast can achieve a real-time efficiency of 250FPS on a GTX 1080Ti and 42FPS even on a Jetson AGX Xavier, outperforming conventional CPU-based trackers. Concurrently, FastTrack demonstrates state-of-the-art performance on four public benchmarks, specifically MOT17, MOT20, KITTI, and DanceTrack, and attains the highest speed in large-scale tracking scenarios of MOT20. [ABSTRACT FROM AUTHOR]

Multi-object tracking (MOT) is a thriving research field in computer vision. The tracklet-based MOT frameworks are frequently employed to generate long and stable trajectories in work scenes that involve long-term occlusion. However, most of these methods train tracklet feature encoders using complex loss functions, lacking an end-to-end paradigm guided by association results, which ultimately leads to limited MOT performance. To address this issue, a graph-based tracklet association framework that seamlessly integrates tracklet feature learning with tracklet association, thereby achieving tracklet association in an end-to-end manner. Specifically, we perform tracklet-based MOT in the graph domain and transform the tracklet association problem into an edge classification task. A message passing network (MPN) is used to update the tracklet features globally, which enhances the robustness of the tracklet features. Additionally, an attention-based feature update function is proposed to ensure the significance of current object. The effectiveness of the proposed framework is demonstrated using MOT17 and MOT20 benchmark datasets, and the experimental results show that the graph-based tracklet association network is a model-independent and plug-and-play component that could combine with different frame-based trackers to boost the MOT performance significantly. [ABSTRACT FROM AUTHOR]

Object tracking is an important topic in computer vision. Most existing trackers require an accurate initial position of the target. However, in the real application, the initial location for tracking may not be accurate, which may lead to tracking drift. To solve this problem, we propose a simple deep learning-based method called Refiner that can produce the accurate position of an object given its rough location. Specifically, we propose an end-to-end position refinement network that consists of a backbone network, a feature enhancement module, a feature fusion module, and a shape predictor; the shape predictor includes two branches: a bounding box prediction branch and a mask prediction branch. We improve the spatial robustness of existing trackers by correcting the inaccurate initial position. In addition, the proposed method can also be used in the tracking process to improve the accuracy of the subsequent tracking results. Lots of experiments on the object tracking benchmarks verify its effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

Coverage is one of the fundamental problems in directional sensor networks (DSNs). This problem is more complicated when we deal with heterogeneous DSNs (HDSNs). Prolonging the network lifetime is another important problem in this area. The problem of finding k disjoint cover sets known as the Set k-Cover problem can solve both the coverage and lifetime issues. In this paper a distributed algorithm is proposed for the Set k-Cover problem in HDSNs and then the method is applied for target k-tracking problem. In the Set k-Cover problem, directional sensors are partitioned into k disjoint sets where each set covers the entire area, and in object k-tracking problem, the object must be tracked by at least k sensors. The proposed algorithms are based on the notion of homology in algebraic topology. We consider the nerve complex corresponding to the HDSN and demonstrate how topological properties of the nerve complex of the network can be used to formulate the Set k-Cover problem as an integer linear programming problem. Then, we propose a distributed algorithm based on the subgradient method for this problem. After that, we propose a distributed algorithm for object k-tracking based on the solution of the Set k-Cover problem. Finally, we evaluate the performance of the proposed algorithms by conducting simulation experiments. [ABSTRACT FROM AUTHOR]

The increasing demand for artificially intelligent smartphone cradles has prompted the need for real-time moving object detection. Real-time moving object tracking requires the development of algorithms for instant tracking analysis without delays. In particular, developing a system for smartphones should consider different operating systems and software development environments. Issues in current real-time moving object tracking systems arise when small and large objects coexist, causing the algorithm to prioritize larger objects or struggle with consistent tracking across varying scales. Fast object motion further complicates accurate tracking and leads to potential errors and misidentification. To address these issues, we propose a deep learning-based real-time moving object tracking system which provides an accuracy priority mode and a speed priority mode. The accuracy priority mode achieves a balance between the high accuracy and speed required in the smartphone environment. The speed priority mode optimizes the speed of inference to track fast-moving objects. The accuracy priority mode incorporates CSPNet with ResNet to maintain high accuracy, whereas the speed priority mode simplifies the complexity of the convolutional layer while maintaining accuracy. In our experiments, we evaluated both modes in terms of accuracy and speed. [ABSTRACT FROM AUTHOR]

In this new era of intelligence and automation, it is important to develop intelligent software to analyse traffic data and detect abnormal activities occurring in the public. Information from GPS, Surveillance cameras, traffic management systems etc will be helpful for the researchers to develop such algorithms. In this research work, we propose a method to detect traffic accidents and used a deep convolutional neural network (D-CNN) and Centroid based vehicle tracking algorithm for vehicle detection. Overlapping bounding boxes and speed of the vehicle are considered for collision detection. The vehicle is tracked using a centroid tracking algorithm to find acceleration, speed and trajectory values of each vehicle in the continuous frames. The trajectory and angle change after the collision can be used to classify the accidents. The result shows a detection accuracy of 99% in such a way outperforms the other latest methods. The results from the proposed method can be used in several accident reconstruction softwares like PC crash, ARPro etc. [ABSTRACT FROM AUTHOR]

Object tracking tasks often face problems such as illumination and appearance changes due to various scenarios. To effectively solve such problems, single object tracking algorithms in recent years have improved the stability of the tracker by introducing online learning, reducing tracking drift, and tracking objects stably even with large deformations. However, on the one hand, employing online learning increases computation costs, causing the tracker to run slower. On the other hand, because it learns from the tracking results, the tracking template will only be correctly updated if the object is accurately tracked. Any incorrect update will result in tracking failure. In this paper, we propose a filtering algorithm for online learning sample processing in object tracking to remedy these issues. Our method can select valid training samples and discard unnecessary training samples in object tracking. Numerous experiments demonstrate that our algorithm improves the tracking accuracy, reduces the computation complexity, and improves the running speed and run‐time. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

The analysis of chicken movement on the farm has several applications in evaluating the well-being and health of birds. Low locomotion may be associated with locomotor problems, and undesirable bird movement patterns may be related to environmental discomfort or fear. Our objective was to test the BoT-SORT object tracking architecture embedded in Yolo v8 to monitor the movement of cage-free chickens and extract measures to classify running, exploring, and resting behaviors, the latter of which includes all other behaviors that do not involve displacement. We trained a new model with a dataset of 3623 images obtained with a camera installed on the ceiling (top images) from an experiment with layers raised cage-free in small-scale aviaries and housed in groups of 20 individuals. The model presented a mAP of 98.5%, being efficient in detecting and tracking the chickens in the video. From the tracking, it was possible to record the movements and directions of individual birds, and we later classified the movement. The results obtained for a group of 20 chickens demonstrated that approximately 84% of the time, the birds remained resting, 10% of the time exploring, and 6% of the time running. The BoT-SORT algorithm was efficient in maintaining the identification of the chickens, and our tracking algorithm was efficient in classifying the movement, allowing us to quantify the time of each movement class. Our algorithm and the measurements we extract to classify bird movements can be used to assess the welfare and health of chickens and contribute to establishing standards for comparisons between individuals and groups raised in different environmental conditions. [ABSTRACT FROM AUTHOR]

Pig husbandry constitutes a significant segment within the broader framework of livestock farming, with porcine well-being emerging as a paramount concern due to its direct implications on pig breeding and production. An easily observable proxy for assessing the health of pigs lies in their daily patterns of movement. The daily movement patterns of pigs can be used as an indicator of their health, in which more active pigs are usually healthier than those who are not active, providing farmers with knowledge of identifying pigs' health state before they become sick or their condition becomes life-threatening. However, the conventional means of estimating pig mobility largely rely on manual observations by farmers, which is impractical in the context of contemporary centralized and extensive pig farming operations. In response to these challenges, multi-object tracking and pig behavior methods are adopted to monitor pig health and welfare closely. Regrettably, these existing methods frequently fall short of providing precise and quantified measurements of movement distance, thereby yielding a rudimentary metric for assessing pig health. This paper proposes a novel approach that integrates optical flow and a multi-object tracking algorithm to more accurately gauge pig movement based on both qualitative and quantitative analyses of the shortcomings of solely relying on tracking algorithms. The optical flow records accurate movement between two consecutive frames and the multi-object tracking algorithm offers individual tracks for each pig. By combining optical flow and the tracking algorithm, our approach can accurately estimate each pig's movement. Moreover, the incorporation of optical flow affords the capacity to discern partial movements, such as instances where only the pig's head is in motion while the remainder of its body remains stationary. The experimental results show that the proposed method has superiority over the method of solely using tracking results, i.e., bounding boxes. The reason is that the movement calculated based on bounding boxes is easily affected by the size fluctuation while the optical flow data can avoid these drawbacks and even provide more fine-grained motion information. The virtues inherent in the proposed method culminate in the provision of more accurate and comprehensive information, thus enhancing the efficacy of decision-making and management processes within the realm of pig farming. [ABSTRACT FROM AUTHOR]

In order to reduce the clock offset caused by network delay jitter, network asymmetry, clock drift and clock frequency difference in the clock synchronization process of IEEE1588, a modified Strong Tracking Adaptive Kalman Filter (MSTAKF) algorithm was proposed. The Kalman filter model established by the algorithm is based on clock offset, network delay, frequency deviation and asymmetry ratio. The filter includes the modified strong tracking algorithm based on fuzzy logic and hypothesis test proposed in this paper, which can effectively eliminate the influence of outliers, and finally make the model tracking effect better. In addition, the modified Sage‐Husa algorithm based on Cholesky decomposition is also included, which makes the measurement noise more stable and ensures the positivity of the measurement noise covariance. The experimental results show that the proposed MSTAKF is smoother than Karman filter (KF), Sage‐Husa adaptive Kalman filter (AKF) and strong tracking Sage‐Husa adaptive Kalman filter (STAKF). Compared with these filter algorithms after testing in the LAN environment, the precision improvement and accuracy rate are significantly improved. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

Generally, the goal of a student activity involving analysis of video data is to obtain the ( I x i , I y i ) position of a particular object in as many frames of the video as possible. Recent years have seen an explosion of interest in analyzing the motion of objects in video data as a way for students to connect the concepts of physics to something tangible like a video recording of an experiment.[[1]] Inexpensive devices can now record videos at 120 or 240 frames/s, and, among instructors, there is an appetite for substantially rethinking the objectives of introductory labs (e.g., Refs. In this paper, we both include some general advice to help object-tracking algorithms locate an object and provide our own algorithms that are simpler and potentially more effective than the sophisticated image-processing algorithms that are currently being used. [Extracted from the article]

With the development of remote sensing earth observation technology, object tracking has gained attention for its broad application prospects in computer vision. However, object tracking is challenging owing to the background clutter, occlusion, and scale variation that often appear in remote sensing videos. Many existing trackers cannot accurately track the object for remote sensing videos with complex backgrounds. Several tracking methods can handle just one situation, such as occlusion. In this article, we propose a Siamese multi-scale adaptive search (SiamMAS) network framework to achieve object tracking for remote sensing videos. First, a multi-scale cross correlation is presented to obtain a more discriminative model and comprehensive feature representation, improving the performance of the model to handle complex backgrounds in remote sensing videos. Second, an adaptive search module is employed that augments the Kalman filter with a partition search strategy for object motion estimation. The Kalman filter is adopted to re-detect the object when the network cannot track the object in the current frame. Moreover, the partition search strategy can help the Kalman filter accomplish a more accurate region-proposal selection. Finally, extensive experiments on remote sensing videos taken from Jilin-1 commercial remote sensing satellites show that the proposed tracking algorithm achieves strong tracking performance with 0.913 precision while running at 37.528 frames per second (FPS), demonstrating its effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

The correlation filter tracking framework has attracted extensive attention from scholars in object tracking due to its efficient and accurate performance. However, this framework cannot adaptively perceive the target's effective spatio-temporal changes. Thus, the correlation filter is easily misled by the occlusion in the occluded scene, resulting in tracking failure. To solve the above limitation effectively, we propose a spatio-temporal self-learning object-tracking model based on an anti-occlusion mechanism. On the one hand, we design a spatially regularized factor based on the local response change vector to give the filter adaptive spatial adjustment ability. On the other hand, we construct a temporal regularized factor with adaptive adjustment capability based on a multimodal template pool to help the filter distinguish between effective appearance changes of the target and ineffective variations caused by occlusion. Then, the alternating direction multiplier method decomposes the proposed model into several simpler subproblems. A small number of iterative steps can obtain the optimal solution of the model. Finally, the OTB100 dataset is employed to compare the proposed method with several state-of-the-art tracking algorithms. Experiments show that the proposed method improves the tracking performance in occluded scenes. [ABSTRACT FROM AUTHOR]

In existing online multiple object tracking algorithms, schemes that combine object detection and re‐identification (ReID) tasks in a single model for simultaneous learning have drawn great attention due to their balanced speed and accuracy. However, different tasks require to focus different features. Learning two different tasks in the same model extracted features can lead to competition between the two tasks, making it difficult to achieve optimal performance. To reduce this competition, a task‐related attention network, which uses a self‐attention mechanism to allow each branch to learn on feature maps related to its task is proposed. Besides, a smooth gradient‐boosting loss function, which improves the quality of the extracted ReID features by gradually shifting the focus to the hard negative samples of each object during training is introduced. Extensive experiments on MOT16, MOT17, and MOT20 datasets demonstrate the effectiveness of the proposed method, which is also competitive in current mainstream algorithm. [ABSTRACT FROM AUTHOR]

An accurate and rapid identification of pig behavior can play a crucial role in the evaluation of health status and welfare. Individual detection and tracking of pigs can greatly enhance the recognition and statistical analysis of their behaviors in large-scale farming environments. This shift from group to precision management can enable better treatment and care, ultimately leading to the goal of welfare-friendly farming. Nevertheless, the occlusion and poses challenges often remain in the actual scenario of pig farming during object tracking, where the group-raised pigs often overlap and move irregularly with each other. Among them, the current mainstream optimization tracking algorithms (such as ByteTrack) can be used to effectively reduce the missed detection from the occlusion using low-score box matching. But it may increase the false detection. The occlusion recovery matching scheme of the OC-SORT algorithm can be used to significantly reduce the identity changes from the occlusion. However, both algorithms still lack the appearance feature matching, which can easily lead to the target loss in the low frame rate or stuttering videos. The appearance feature matching can be expected to solve occlusion and irregular motion. Unfortunately, the existing convolution-based identity recognition networks often struggle to obtain the appearance features with better discriminatory properties, due to the highly similar appearance of pigs. In this study, a PigsTrack tracker was proposed to further optimize the multi-object tracking for the group-raised pigs in actual scenarios. Firstly, the high-performance detector (YOLOX) was used in the tracker to reduce the false and missed detection rates of pigs in the occluded scenarios. Then, an identity recognition (Re-ID) network was constructed using the Transformer model to extract the appearance features with better discriminative properties. Secondly, the matching strategies of appearance feature, IoU, and occlusion recovery were integrated with the tracker for the accurate tracking of group-raised pigs using the idea of OCSORT. 20 videos were selected (a total of 36 000 frames) from the original PBVD (Pigs Behaviours Video) dataset. The obtained dataset was labeled using the DarkLabel tool. Furthermore, the main evaluation metrics included HOTA (higher order tracking accuracy), MOTA (multiple object tracking accuracy), and IDF1 (identification F1 score). The comparison showed that the PigsTrack tracker outperformed the rest in all indicators of tracking performance with the HOTA of 85.66%, MOTA of 98.59%, and IDF1 of 99.57 %, which were higher than those of DeepSORT, OC-SORT, ByteTrack and StrongSORT trackers, respectively. Additionally, the three-step matching also performed better stable tracking to reduce the number of false and missed detections, as well as the identity switches, with a total of only 13 identity switches. Overall, the PigsTrack can be suitable for the multi-object tracking of group-reared pigs in the breeding environment, particularly for the pig interactions, occlusions, and similarities in appearance. The visualization of the attention map confirmed that the precise tracking of pigs in long videos was achieved in the feature extractor using the Transformer model. Furthermore, more distinctive appearance features were extracted for more detailed information. The tracking results were compared with and without occlusion recovery matching. As such, the occlusion recovery matching greatly contributed to the tracking pigs in dense environments. Finally, the pig behavior in videos was analyzed using the PigsTrack with the Slowfast network. Experiments demonstrated that superior accuracy, effectiveness, and continuity were achieved to recognize and statistically analyze the daily behaviors of group-reared pigs. In addition, the ABVD dataset was utilized to create Dataset 2 for the supplementary experiments, and then to validate the generalization performance of the model in multi-target tracking tasks of group-raised pigs. The performance of the PigsTrack tracker on Dataset 2 (with HOTA, MOTA, and IDF1 scores of 69.14%, 94.82%, and 90.11%, respectively) outperformed the rest, indicating only a slight gap, compared with the PBVD dataset. Therefore, it is also necessary to further optimize the experimental environment and tracker for the engaged pigs in aggressive behavior. Overall, the PigsTrack can provide an important foundation for precision farming. [ABSTRACT FROM AUTHOR]

Summary: This article formulates and solves a problem of model following adaptive control (MFAC) of nodes in the complex dynamic network (CDN) with the dynamic matrix‐relation links. From the perspective of composite systems, the CDN with the dynamic matrix‐relation links is considered as an interconnected system composed of intercoupling node subsystem (NS) and link subsystem (LS). Based on this idea, employing the Riccati matrix differential equation as the dynamic equation of LS, and the vector differential equation with the second derivative term as the dynamic equation of NS. Then, to remove the restriction that the state of LS is unavailable, an adaptive control scheme is proposed for NS with the help of constructed dynamic auxiliary tracking target for LS, such that the problem of MFAC of nodes is solved. Besides, it is proved mathematically that the state of NS is asymptotically tracking the state of the given reference model when the state of LS is asymptotically tracking the constructed dynamic auxiliary tracking target. Finally, the numerical simulation is given to verify the effectiveness of the theoretical results in this article. [ABSTRACT FROM AUTHOR]

In marine surveillance, distinguishing between normal and anomalous vessel movement patterns is critical for identifying potential threats in a timely manner. Once detected, it is important to monitor and track these vessels until a necessary intervention occurs. To achieve this, track association algorithms are used, which take sequential observations comprising the geological and motion parameters of the vessels and associate them with respective vessels. The spatial and temporal variations inherent in these sequential observations make the association task challenging for traditional multi-object tracking algorithms. Additionally, the presence of overlapping tracks and missing data can further complicate the trajectory tracking process. To address these challenges, in this study, we approach this tracking task as a multivariate time series problem and introduce a 1D CNN-LSTM architecture-based framework for track association. This special neural network architecture can capture the spatial patterns as well as the long-term temporal relations that exist among the sequential observations. During the training process, it learns and builds the trajectory for each of these underlying vessels. Once trained, the proposed framework takes the marine vessel's location and motion data collected through the automatic identification system (AIS) as input and returns the most likely vessel track as output in real-time. To evaluate the performance of our approach, we utilize an AIS dataset containing observations from 327 vessels traveling in a specific geographic region. We measure the performance of our proposed framework using standard performance metrics such as accuracy, precision, recall, and F1 score. When compared with other competitive neural network architectures, our approach demonstrates a superior tracking performance. [ABSTRACT FROM AUTHOR]

Moving object tracking is one of the key issues in the domain of computer vision. A variety of challenges are posed while tracking the object in the real-world scenario. In this paper, we have proposed a particle filtering-based algorithm to track the moving object in a complex real-world environment having shadows, dynamic entities in the background, bad weather condition and illumination variation. Specifically, we have attempted to have an effective target and scene model by adhering to the notion of feature fusion. In this regard, a novel feature fusion strategy is proposed which is scene dependent. Besides, for an efficient target model, the template for tracking is partitioned with a view to capture the local attributes of the object. In every partition, two features, namely the local binary pattern (LBP) and the mean RGB color features, are fused in a probabilistic framework. The weights for the probabilistic fusion are computed based on the local scene dynamics. The feature that strongly favors the scene in a given template partition is assigned more weightage and vice versa in the fusion process. The combined feature over all the template partitions is used to model the target. The particles that represent the state of the object evolve through a dynamic state model to track the object. The proposed tracking algorithm is successfully tested on videos considered from LASIESTA, CDnet, and DAVIS 2016 data base and it showed improved tracking accuracy as compared to existing algorithms. [ABSTRACT FROM AUTHOR]

Due to the problem of low tracking accuracy and weak tracking stability of current multi-object pedestrian tracking algorithms in complex scenes for videos, a Detection-based Online Multi-Object Pedestrian Tracking (DOMOPT) network is proposed. First, a Multi-Level Feature Fusion (MLFF) pedestrian detection network is proposed based on the Center and Scale Prediction (CSP) algorithm. The pyramid convolutional neural network is used as the backbone to enhance the feature extraction capability for small objects. The shallow features and deep features at multiple levels are integrated to fully obtain the position and semantic information to further improve the detection performance for small objects. Then, on the basis of Joint Detection and Embedding (JDE) architecture, a Multi-Branch Pedestrian Appearance (MBPA) feature extraction network is proposed and added into the pedestrian detection network to extract the appearance feature vector corresponding to each pedestrian. The pedestrian appearance feature extraction is treated as a classification task jointly training with the pedestrian detection task, using the multi-task learning strategy. Experimental results show that the proposed network has better tracking accuracy and stability compared with state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

This study proposes a method for detecting and tracking traversable regions in off-road conditions for unmanned ground vehicles (UGVs). Off-road conditions, such as rough terrain or fields, present significant challenges for UGV navigation, and detecting and tracking traversable regions is essential to ensure safe and efficient operation. Using a 3D laser scanner and range-image-based approach, a method is proposed for detecting traversable regions under off-road conditions; this is followed by a Bayesian fusion algorithm for tracking the traversable regions in consecutive frames. Our range-image-based traversable-region-detection approach enables efficient processing of point cloud data from a 3D laser scanner, allowing the identification of traversable areas that are safe for the unmanned ground vehicle to drive on. The effectiveness of the proposed method was demonstrated using real-world data collected during UGV operations on rough terrain, highlighting its potential as a solution for improving UGV navigation capabilities in challenging environments. [ABSTRACT FROM AUTHOR]

Most existing Siamese Network trackers rely on a predefined anchor box to predict object position. However, they require complicated hyperparameter settings. The authors directly forecast the object boundary by using the fully convolutional network as the head of the tracking network to solve this issue and simplify the application. The end‐to‐end design avoids the setting of hyperparameters and candidate boxes. The authors also discovered that the validation loss decreased less than the training loss throughout Siamese training. The authors changed the normalisation layer from batch normalisation to group normalisation to solve this issue. It solved the problem that the loss function is difficult to decrease and increased training efficiency. Test experiments on the tracking dataset, including VOT2019 and GOT10k, show that the authors' network outperforms DaSiamRPN and SiamFC regarding precision under the same network size and runs at 24 FPS on an AMD 4800 CPU. It also runs at 306 FPS on a 3090 GPU. [ABSTRACT FROM AUTHOR]

Visual object tracking (VOT) is a fundamental and complex problem in computer vision field. In the past few years, the research focus has been shifted from template matching to deep learning models. Especially, the Siamese networks dominate tracking domain in recent years, which take the first frame as the reference and perform object detection and localization in the following frames. However, most of them could not capture target changes due to the lack of strong feature representation abilities. To address these issue, we propose an advanced tracking network in this paper based on recurrent historical localization information. Unlike traditional symmetric structures, we utilize two convolution layers to perform target classification that predicts the initial target center. Then, we apply a gated recurrent unit that fuses multi-resolution features with historical localization information to yield the final optimized target position. Extensive experiments have been conducted on six mainstream datasets: OTB100, GOT-10k, TrackingNet, LaSOT, VOT2018 and NFS, where our tracker exhibits state-of-the-art performances. [ABSTRACT FROM AUTHOR]

Frequent state monitoring and controller updates can enhance the precision of tracking control, while simultaneously overburdening the communication network transmission. In this paper, For the purpose of saving communication costs, we propose event-triggered control algorithms for the optimal tracking control problem. First, we reconstruct the discrete-time nonlinear system into a converted system. Then, the adaptive dynamic programming algorithm is employed to find the optimal controller off-line, and the event-triggered scheme is used to reduce the communication costs online. Novel triggering conditions with fewer assumptions are designed to implement the event-triggered scheme. Different from existing works, the event-triggered scheme can be introduced not only into the converted system but also into the actual system, which is more practical because the actual controller is what one can only access in practice. In addition, with the developed algorithms, the tracking error can be proved to be stable at the origin, in other words, the actual system can be guaranteed to track the desired trajectory. Algorithms developed in this paper are implemented by three neural networks, the model network, the action network and the critic network. Finally, examples are presented to verify the effectiveness and rationality of the algorithms. [ABSTRACT FROM AUTHOR]

Border tracking in binary images is an important operation in many computer vision applications. The problem consists in finding borders in a 2D binary image (where all of the pixels are either 0 or 1). There are several algorithms available for this problem, but most of them are sequential. In a former paper, a parallel border tracking algorithm was proposed. This algorithm was designed to run in Graphics Processing units, and it was based on the sequential algorithm known as the Suzuki algorithm. In this paper, we adapt the previously proposed GPU algorithm so that it can be executed in multicore computers. The resulting algorithm is evaluated against its GPU counterpart. The results show that the performance of the GPU algorithm worsens (or even fails) for very large images or images with many borders. On the other hand, the proposed multicore algorithm can efficiently cope with large images. [ABSTRACT FROM AUTHOR]

In object tracking, motion blur is a common challenge induced by rapid movement of target object or long time exposure of the camera, which leads to poor tracking performance. Traditional solutions usually perform image recovery operations before tracking object. However, most image recovery methods usually have higher computational cost, which decreases the tracking speed. In order to solve the above problems, we propose a deblurring Transformer-based tracking method embedding the conditional cross-attention. The proposed method integrates three important modules: (1) an image quality assessment (IQA) module to estimate image quality; (2) an image deblurring module based on lightweight adversarial network to improve image quality; and (3) a tracking module based on Transformer with conditional cross-attention to enhance the object localization ability. Experimental results on two UAV object tracking benchmarks show that the proposed trackers achieve competitive results compared to several state-of-the-art trackers. [ABSTRACT FROM AUTHOR]

Tracking droplets in microfluidics is a challenging task. The difficulty arises in choosing a tool to analyze general microfluidic videos to infer physical quantities. The state-of-the-art object detector algorithm You Only Look Once (YOLO) and the object tracking algorithm Simple Online and Realtime Tracking with a Deep Association Metric (DeepSORT) are customizable for droplet identification and tracking. The customization includes training YOLO and DeepSORT networks to identify and track the objects of interest. We trained several YOLOv5 and YOLOv7 models and the DeepSORT network for droplet identification and tracking from microfluidic experimental videos. We compare the performance of the droplet tracking applications with YOLOv5 and YOLOv7 in terms of training time and time to analyze a given video across various hardware configurations. Despite the latest YOLOv7 being 10% faster, the real-time tracking is only achieved by lighter YOLO models on RTX 3070 Ti GPU machine due to additional significant droplet tracking costs arising from the DeepSORT algorithm. This work is a benchmark study for the YOLOv5 and YOLOv7 networks with DeepSORT in terms of the training time and inference time for a custom dataset of microfluidic droplets. [ABSTRACT FROM AUTHOR]

Bridge deformation response data are the basis for calculating the dynamic parameters of the bridge, and it is of great significance to accurately measure the deformation response of the bridge during the load test and service conditions. A bridge deformation measurement method using an unmanned aerial system (UAS) with dual cameras and a deep learning-based object tracking method is proposed to measure the bridge deformation. The contributions are as follows: (1) To address the problem that the movement of the UAS brings error to the deformation measurement results, dual cameras with telephoto and wide-angle lenses are used to simultaneously capture the deformed points and stable points on the bridge, so as to simultaneously measure the deformation of the bridge and the displacement of the UAS, and then the displacement of UAS is eliminated by using the homography relationship between the two cameras. (2) To solve the problem that the traditional digital image correlation-based displacement measurement method is easily disturbed by factors such as light changes and occlusion, a displacement calculation method based on object detection network and target tracking algorithm is proposed to achieve the stable target displacement measurement. Finally, the proposed method was verified in a laboratory test and applied to the deformation measurement of an in-service bridge to verify the practicability of the proposed method. [ABSTRACT FROM AUTHOR]

This study proposes a visual tracking system that can detect and track multiple fast-moving appearance-varying targets simultaneously with 500 fps image processing. The system comprises a high-speed camera and a pan-tilt galvanometer system, which can rapidly generate large-scale high-definition images of the wide monitored area. We developed a CNN-based hybrid tracking algorithm that can robustly track multiple high-speed moving objects simultaneously. Experimental results demonstrate that our system can track up to three moving objects with velocities lower than 30 m per second simultaneously within an 8-m range. The effectiveness of our system was demonstrated through several experiments conducted on simultaneous zoom shooting of multiple moving objects (persons and bottles) in a natural outdoor scene. Moreover, our system demonstrates high robustness to target loss and crossing situations. [ABSTRACT FROM AUTHOR]

Object tracking using Hyperspectral Images (HSIs) obtains satisfactory result in distinguishing objects with similar colors. Yet, the tracking algorithm tends to fail when the target undergoes deformation. In this paper, a SiamRPN based hyperspectral tracker is proposed to deal with this problem. Firstly, a band selection method based on a genetic optimization method is designed for rapidly reducing the redundancy of information in HSIs. Specifically, three bands with highest joint entropy are selected. To solve the problem that the information of the template in the SiamRPN model decays over time, an update network is trained on the dataset from general objective tracking benchmark, which can obtain effective cumulative templates. The use of cumulative templates with spectral information makes it easier to track the deformed target. In addition, transfer learning of the pre-trained SiamRPN is designed to obtain a better model for HSIs. The experimental results show that the proposed tracker can obtain good tracking results over the entire public dataset, and that it is better than the other popular trackers when the target's deformation is qualitatively and quantitatively compared, achieving an overall success rate of 57.5% and a deformation challenge success rate of 70.8%. [ABSTRACT FROM AUTHOR]

Designing a robust aircraft control system used to achieve a good tracking performance and stable dynamic behavior against working disturbances problem has attracted attention of control engineers. In this paper, a pitch angle control system for aircraft is designed utilizing liner quadratic Gaussian (LQG) optimal controller technique with a numerical tuning algorithm method in the longitudinal plane through cruising stage. Main design approach of LQG controller includes obtaining best weighting matrices values using trial and error method that consumes effort and takes more time, in addition, there is no guarantees to obtain optimum values for weighting matrices elements. In this research, genetic algorithm (GA) is used to optimize the state and control weighting matrices and determine best values for their elements. The proposed traditional and optimized LQG pitch controller schemes are implemented utilizing Matlab simulation tool and their performance are presented and compared based on transient and steady state performance parameters. The simulation results reveal the ability of the optimized GA_LQG controller to reject the effect of the noises in the aircraft system dynamic and achieve a good and stable tracking performance compared with that of the conventional LQG pitch control system. [ABSTRACT FROM AUTHOR]

Multi-object tracking (MOT) is a prominent and important study in point cloud processing and computer vision. The main objective of MOT is to predict full tracklets of several objects in point cloud. Occlusion and similar objects are two common problems that reduce the algorithm's performance throughout the tracking phase. The tracking performance of current MOT techniques, which adopt the 'tracking-by-detection' paradigm, is degrading, as evidenced by increasing numbers of identification (ID) switch and tracking drifts because it is difficult to perfectly predict the location of objects in complex scenes that are unable to track. Since the occluded object may have been visible in former frames, we manipulated the speed and location position of the object in the previous frames in order to guess where the occluded object might have been. In this paper, we employed a unique intersection over union (IoU) method in three-dimension (3D) planes, namely a distance IoU non-maximum suppression (DIoU-NMS) to accurately detect objects, and consequently we use 3D-DIoU for an object association process in order to increase tracking robustness and speed. By using a hybrid 3D DIoU-NMS and 3D-DIoU method, the tracking speed improved significantly. Experimental findings on the Waymo Open Dataset and nuScenes dataset, demonstrate that our multistage data association and tracking technique has clear benefits over previously developed algorithms in terms of tracking accuracy. In comparison with other 3D MOT tracking methods, our proposed approach demonstrates significant enhancement in tracking performances. [ABSTRACT FROM AUTHOR]

Multi-object Tracking is an important issue that has been widely investigated in computer vision. However, in practical applications, moving targets are often occluded due to complex changes in the background, which leads to frequent pedestrian ID switches in multi-object tracking. To solve the problem, we present a multi-object tracking algorithm based on FairMOT and Circle Loss. In this paper, HRNet is adopted as the baseline. Then, Polarized Self-Attention is added to HRNet-w32 to obtain weights of helpful information based on its modeling advantages. Moreover, the re-identification branch is optimized, and the Circle Loss is selected as the loss function to acquire more discriminative pedestrian features and to distinguish different pedestrians. The method proposed is assessed on the public MOT17 datasets. The experimental results show that the MOTA score achieves 69.5%, IDF1 reaches 70.0%, and the number of ID switches (IDs) decreases 636 times compared to the TraDes algorithm. [ABSTRACT FROM AUTHOR]

Daphnia magna is an important organism in ecotoxicity studies because it is sensitive to toxic substances and easy to culture in laboratory conditions. Its locomotory responses as a biomarker are highlighted in many studies. Over the last several years, multiple high-throughput video tracking systems have been developed to measure the locomotory responses of Daphnia magna. These high-throughput systems, used for high-speed analysis of multiple organisms, are essential for efficiently testing ecotoxicity. However, existing systems are lacking in speed and accuracy. Specifically, speed is affected in the biomarker detection stage. This study aimed to develop a faster and better high-throughput video tracking system using machine learning methods. The video tracking system consisted of a constant temperature module, natural pseudo-light, multi-flow cell, and an imaging camera for recording videos. To measure Daphnia magna movements, we developed a tracking algorithm for automatic background subtraction using k-means clustering, Daphnia classification using machine learning methods (random forest and support vector machine), and tracking each Daphnia magna location using the simple online real-time tracking algorithm. The proposed tracking system with random forest performed the best in terms of identification (ID) precision, ID recall, ID F1 measure, and ID switches, with scores of 79.64%, 80.63%, 78.73%, and 16, respectively. Moreover, it was faster than existing tracking systems such as Lolitrack and Ctrax. We conducted an experiment to observe the impact of toxicants on behavioral responses. Toxicity was measured manually in the laboratory and automatically using the high-throughput video tracking system. The median effective concentration of Potassium dichromate measured in the laboratory and using the device was 1.519 and 1.414, respectively. Both measurements conformed to the guideline provided by the Environmental Protection Agency of the United States; therefore, our method can be used for water quality monitoring. Finally, we observed Daphnia magna behavioral responses in different concentrations after 0, 12, 18, and 24 h and found that there was a difference in movement according to the concentration at all hours. [ABSTRACT FROM AUTHOR]

A convolutional neural network-based multiobject detection and tracking algorithm can be applied to vehicle detection and traffic flow statistics, thus enabling smart transportation. Aiming at the problems of the high computational complexity of multiobject detection and tracking algorithms, a large number of model parameters, and difficulty in achieving high throughput with a low power consumption in edge devices, we design and implement a low-power, low-latency, high-precision, and configurable vehicle detector based on a field programmable gate array (FPGA) with YOLOv3 (You-Only-Look-Once-version3), YOLOv3-tiny CNNs (Convolutional Neural Networks), and the Deepsort algorithm. First, we use a dynamic threshold structured pruning method based on a scaling factor to significantly compress the detection model size on the premise that the accuracy does not decrease. Second, a dynamic 16-bit fixed-point quantization algorithm is used to quantify the network parameters to reduce the memory occupation of the network model. Furthermore, we generate a reidentification (RE-ID) dataset from the UA-DETRAC dataset and train the appearance feature extraction network on the Deepsort algorithm to improve the vehicles' tracking performance. Finally, we implement hardware optimization techniques such as memory interlayer multiplexing, parameter rearrangement, ping-pong buffering, multichannel transfer, pipelining, Im2col+GEMM, and Winograd algorithms to improve resource utilization and computational efficiency. The experimental results demonstrate that the compressed YOLOv3 and YOLOv3-tiny network models decrease in size by 85.7% and 98.2%, respectively. The dual-module parallel acceleration meets the demand of the 6-way parallel video stream vehicle detection with the peak throughput at 168.72 fps. [ABSTRACT FROM AUTHOR]

In recent years, the object‐tracking algorithm based on Siamese network has gradually become the mainstream algorithm in the field of object tracking due to its characteristics of balancing speed and accuracy. The majority of Siamese‐based trackers only use the first frame extraction template for subsequent tracking in order to prevent the introduction of noise. However, merely with a single initial template employed, it is difficult to achieve the best performance of the tracker in the face of complex tracking environments such as occlusion, motion blur, and non‐rigid deformation. Therefore, the present paper proposes a new multi‐template fusion module based on graph attention network (G‐M module), which consists of two parts: a graph‐attention‐network‐based feature‐embedding module (G module) and a multi‐template fusion module (M module). It can greatly reduce the background noise introduced by template updating while improving the tracker's ability to adapt to changes in object appearance. In addition, in order to maximize the value of G‐M module, the present paper also puts forward a two‐stage template update threshold judgment mechanism. The Pearson correlation coefficient (PCCs) is introduced and combined with APCE and the maximum response value (F‐max) to filter out reliable templates for updating. In this paper, the proposed method is applied to the SiamFC and SiamFC++ trackers. Extensive experiments on mainstream data sets, such as OTB2015, VOT2016, and GOT‐10 k, show that the proposed method can effectively update the tracking template and improve the tracker performance. © 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

The object tracking algorithm based on Siamese network often extracts the deep feature of the target to be tracked from the first frame of the video sequence as a template, and uses the template for the whole tracking process. Because the manually annotated target in the first frame of video sequence is more accurate, these algorithms often have stable performance. However, it is difficult to adapt to the changing target features only using the target template extracted from the first frame. Inspired by the feature fusion network based on a transformer, this paper proposes a template update module called multi‐template temporary information fusion module (MTFM), which can be trained offline. By fusing multiple target template features on time series, the template can always adapt to the changes of target appearance in the tracking process. In order to train the MTFM, this paper proposes a training method using time series data and Mean Square Error (MSE) as the loss function. This paper uses the MTFM on SiamFC++ tracker, and obtains good experimental results in three challenging datasets, including VOT2016, OTB100 and GOT‐10k. The running speed of the algorithm on graphics processing unit (GPU) is maintained at about 200fps, which exhibits good real‐time performance. [ABSTRACT FROM AUTHOR]

Visual tracking is one of the most challenging problems in computer vision. Most state-of-the-art visual trackers suffer from three challenging problems: nondiverse discriminate feature representation, coarse object locator, and limited quantities of positive samples. In this paper, a multi-view multi-expert region proposal prediction algorithm for tracking is proposed to solve the above problems concurrently in one framework. The proposed algorithm integrates multiple views and exploits powerful multiple sources of information, which can solve nondiverse discriminate feature representation problem effectively. It builds multiple SVM classifier models on the expanded bounding boxes and adds the regional suggestion network module to accurately optimize it to predict optimal object location, which naturally alleviates the coarse object locator and limited quantities of positive samples problems at the same time. A comprehensive evaluation of the proposed approach on various benchmark sequences has been performed. The evaluation results demonstrate our method can significantly improve the tracking performance by combining the advantages of lightweight region proposal network predictive learning model and multi-view expert groups. The experimental results demonstrate the proposed approach outperforms other state-of-the-art visual trackers. [ABSTRACT FROM AUTHOR]

Object detection based on deep learning is one of the most important and fundamental tasks of computer vision. High-performance detection algorithms have been widely used in many practical fields. For the management of workers wearing helmets in construction scenarios, this paper proposes a framework model based on the YOLOv5 detection algorithm, combined with multi-object tracking algorithms, to monitor and track whether workers wear safety helmets in real-time video. The improved StrongSORT tracking algorithm of DeepSORT is selected to reduce the loss of the tracked object caused by the occlusion, trajectory blur, and motion scale of the object. The safety helmet dataset is trained with YOLOv5s, and the best result of training is used as the weight model in the StrongSORT tracking algorithm. The experimental results show that the mAP@0.5 of all classes in the YOLOv5s model can reach 95.1% in the validation dataset, mAP@0.5:0.95 is 62.1%, and the precision of wearing helmet is 95.7%. After the box regression loss function was changed from CIOU to Focal-EIOU, the mAP@0.5 increased to 95.4%, mAP@0.5:0.95 increased to 62.9%, and the precision of wearing helmet increased to 96.5%, which were increased by 0.3%, 0.8% and 0.8%, respectively. StrongSORT can update object trajectories in video frames at a speed of 0.05 s per frame. Based on the improved YOLOv5s combined with the StrongSORT tracking algorithm, the helmet-wearing tracking detection can achieve better performance. [ABSTRACT FROM AUTHOR]

The understanding of human-object interactions is fundamental in First Person Vision (FPV). Visual tracking algorithms which follow the objects manipulated by the camera wearer can provide useful information to effectively model such interactions. In the last years, the computer vision community has significantly improved the performance of tracking algorithms for a large variety of target objects and scenarios. Despite a few previous attempts to exploit trackers in the FPV domain, a methodical analysis of the performance of state-of-the-art trackers is still missing. This research gap raises the question of whether current solutions can be used "off-the-shelf" or more domain-specific investigations should be carried out. This paper aims to provide answers to such questions. We present the first systematic investigation of single object tracking in FPV. Our study extensively analyses the performance of 42 algorithms including generic object trackers and baseline FPV-specific trackers. The analysis is carried out by focusing on different aspects of the FPV setting, introducing new performance measures, and in relation to FPV-specific tasks. The study is made possible through the introduction of TREK-150, a novel benchmark dataset composed of 150 densely annotated video sequences. Our results show that object tracking in FPV poses new challenges to current visual trackers. We highlight the factors causing such behavior and point out possible research directions. Despite their difficulties, we prove that trackers bring benefits to FPV downstream tasks requiring short-term object tracking. We expect that generic object tracking will gain popularity in FPV as new and FPV-specific methodologies are investigated. [ABSTRACT FROM AUTHOR]

Target tracking algorithms based on deep learning have achieved good results in public datasets. Among them, the network tracking algorithm based on Siamese tracking has a high accuracy and fast speed, which has attracted significant attention. However, the Siamese tracker uses the AlexNet network as its backbone and the network layers are relatively shallow, so it does not make full use of the ability of the deep neural network. If only the backbones of target tracking are replaced, there will be no obvious improvement, such as in the cases of ResNet and Inception. Therefore, this paper designs a wider and deeper network structure. At a wider level, a mechanism that can adaptively adjust the receptive field (RF) size is designed. Firstly, multiple branches are divided by the split operator, and each branch has a different size of kernel corresponding to a different size of RF; then, the fuse operator is used to fuse the information of each branch to obtain the selection weights; and finally, according to the selection, the aggregation feature map is weighted. At a deeper level, a new kind of residual models is designed. The channel is simplified by pruning in order to improve the tracking speed. According to the above, a wider and deeper Siamese network was proposed in this paper. The experimental results show that the structure proposed in this paper achieves a good tracking effect and real-time performance on six kinds of datasets. The proposed tracker achieves an SUC and Prec of LaSOT of 0.569 and 0.571, respectively. [ABSTRACT FROM AUTHOR]

Current surveillance systems frequently use fixed-angle cameras and record a feed from those cameras. There are several disadvantages to such systems, including a low resolution for far away objects, a limited frame range and wasted disk space. This paper presents a novel algorithm for automatically detecting, tracking and zooming in on active targets. The object tracking system is connected to a camera that has a 360° horizontal and 90° vertical movement range. The combination of tracking, movement identification and zoom means that the system is able to effectively improve the resolution of small or distant objects. The object detection system allows for the disk space to be conserved as the system ceases recording when no valid targets are detected. Using an adaptive object segmentation algorithm, it is possible to detect the shape of moving objects efficiently. When processing multiple targets, each target is assigned a color and is treated separately. The tracking algorithm is able to adapt to targets moving at different speeds and is able to control the camera according to a predictive formula to prevent the loss of image quality due to camera trail. In the test environment, the zoom can sufficiently lock onto the head of a moving human; however, simultaneous tracking and zooming occasionally results in a failure to track. If this system is deployed with a facial recognition algorithm, the recognition accuracy can be effectively improved. [ABSTRACT FROM AUTHOR]

Drones, or general UAVs, equipped with cameras have been fast deployed with a wide range of applications, including agriculture, aerial photography, and surveillance. Consequently, automatic understanding of visual data collected from drones becomes highly demanding, bringing computer vision and drones more and more closely. To promote and track the developments of object detection and tracking algorithms, we have organized three challenge workshops in conjunction with ECCV 2018, ICCV 2019 and ECCV 2020, attracting more than 100 teams around the world. We provide a large-scale drone captured dataset, VisDrone, which includes four tracks, i.e., (1) image object detection, (2) video object detection, (3) single object tracking, and (4) multi-object tracking. In this paper, we first present a thorough review of object detection and tracking datasets and benchmarks, and discuss the challenges of collecting large-scale drone-based object detection and tracking datasets with fully manual annotations. After that, we describe our VisDrone dataset, which is captured over various urban/suburban areas of 14 different cities across China from North to South. Being the largest such dataset ever published, VisDrone enables extensive evaluation and investigation of visual analysis algorithms for the drone platform. We provide a detailed analysis of the current state of the field of large-scale object detection and tracking on drones, and conclude the challenge as well as propose future directions. We expect the benchmark largely boost the research and development in video analysis on drone platforms. All the datasets and experimental results can be downloaded from https://github.com/VisDrone/VisDrone-Dataset. [ABSTRACT FROM AUTHOR]

The detection and tracking of small targets under low signal-to-clutter ratio (SCR) has been a challenging task for infrared search and track (IRST) systems. Track-before-detect (TBD) is a widely-known algorithm which can solve this problem. However, huge computation costs and storage requirements limit its application. To address these issues, a dynamic programming (DP) and multiple hypothesis testing (MHT)-based infrared dim point target detection algorithm (DP–MHT–TBD) is proposed. It consists of three parts. (1) For each pixel in current frame, the second power optimal merit function-based DP is designed and performed in eight search areas to find the target search area that contains the real target trajectory. (2) In the target search area, the parallel MHT model is designed to save the tree-structured trajectory space, and a two-stage strategy is designed to mitigate the contradiction between the redundant trajectories and the requirements of more trajectories under low SCR. After constant false alarm segmentation of the energy accumulation map, the preliminary candidate points can be obtained. (3) The target tracking method is designed to eliminate false alarms. In this work, an efficient second power optimal merit function-based DP is designed to find the target search area for each pixel, which greatly reduces the trajectory search space. A two-stage MHT model, in which pruning for the tree-structured trajectory space is avoided and all trajectories can be processed in parallel, is designed to further reduce the hypothesis space exponentially. This model greatly reduces computational complexity and saves storage space, improving the engineering application of the TBD method. The DP–MHT–TBD not only takes advantage of the small computation amount of DP and high accuracy of an exhaustive search but also utilizes a novel structure. It can detect a single infrared point target when the SCR is 1.5 with detection probability above 90% and a false alarm rate below 0.01%. [ABSTRACT FROM AUTHOR]

Automatic monitoring tools can be useful for assessing the health and welfare status of animals. Specifically, a computer-vision-based tracking tool could be helpful to remotely and automatically monitoring the behaviour of individual animals. However, animals housed in partly covered pens present a particular challenge for animal tracking due to the possibility for animals to disappear from and reappear in the field of view (FOV). The aim of this study was to develop a tracking method for weaner pigs housed in partly covered pens, with the particular aim to re-identify individuals when they reappear in the FOV. In this study a one-shot tracker in which the detection and re-identification (re-ID) branches were jointly trained was adopted for tracking pigs. Three associations based on re-ID features and intersection over union (IoU) were used for matching the correct ID, especially re-identifying individuals reappearing in the FOV. Two sets of short videos were selected to test the model, with a first set of two short videos (mean ± SD: 1m50s ± 20) and a second set of three short videos (mean ± SD: 10m08s ± 3m52s). The model reached the performance of 91.41% and 88.33% in MOTA and IDF1 on the first set of videos, and 81.17% in mean tracking percentage per individual on the second set. The test on one long video (from the same pen, length: 85 m) achieved a tracking percentage of 16.78% per individual. The suggested method improved automatic individual behaviour analysis in complex environments where animals can leave the FOV. • A long-term tracking algorithm was developed for group-housed pigs. • Re-ID features and IoU were used to associate and identify pigs entering the field of view. • An average tracking percentage of 16.78% was obtained on a 85-min-long video. [ABSTRACT FROM AUTHOR]

Pedestrian detection and tracking based on computer vision has gradually become an international pattern recognition, which is one of the most active research topics in the field of computer vision and artificial intelligence. Using the theoretical results in the field of pattern recognition and computer vision technology, we are committed to detect and track pedestrians from video sequences. In addition to computer vision-based passer-by detection and tracking technology as the key, in the advanced computer vision action and analysis, it has a direct impact on the accuracy and robustness of its understanding. We analyzed various targets, such as subsequent recognition motion and pedestrian motion, and described them as high-level application processing, such as action understanding. In addition, because of the unique texture of human clothes compared with the surrounding natural landscape, they are highly "prominent" from the perspective of human visual system, and they are particularly prominent in the peripheral part of human contact with the background. In this paper, a binary function based on importance is proposed. As the space representation of image itself is not sensitive to noise and local signal, space representation is used. In addition, as an observation model, it can reduce the adverse effects of background noise and local noise on the tracking algorithm. Through the function block tracking, the pedestrian's body can be tracked in detail. At the same time, the color band learning method is used to update the target template online to deal with the changes of target appearance caused by sunshine, pedestrian posture, and other factors. According to the experimental results, even if the appearance and posture of pedestrians change greatly, it has a stable tracking effect. [ABSTRACT FROM AUTHOR]

In-situ irradiation transmission electron microscopy (TEM) offers unique insights into the millisecond-timescale post-cascade process, such as the lifetime and thermal stability of defect clusters, vital to the mechanistic understanding of irradiation damage in nuclear materials. Converting in-situ irradiation TEM video data into meaningful information on defect cluster dynamic properties (e.g., lifetime) has become the major technical bottleneck. Here, we present a solution called the DefectTrack, the first dedicated deep learning-based one-shot multi-object tracking (MOT) model capable of tracking cascade-induced defect clusters in in-situ TEM videos in real-time. DefectTrack has achieved a Multi-Object Tracking Accuracy (MOTA) of 66.43% and a Mostly Tracked (MT) of 67.81% on the test set, which are comparable to state-of-the-art MOT algorithms. We discuss the MOT framework, model selection, training, and evaluation strategies for in-situ TEM applications. Further, we compare the DefectTrack with four human experts in quantifying defect cluster lifetime distributions using statistical tests and discuss the relationship between the material science domain metrics and MOT metrics. Our statistical evaluations on the defect lifetime distribution suggest that the DefectTrack outperforms human experts in accuracy and speed. [ABSTRACT FROM AUTHOR]