Object Tracking Method Research Articles

SiamFCA: A new fish single object tracking method based on siamese network with coordinate attention in aquaculture

In aquaculture, tracking fish helps to obtain information about their growth, behavior, location and health, etc. However, there are some problems in the process of fish tracking. Firstly, to solve the noise problem of underwater images, we use the contrast limited adaptive histogram equalization (CLAHE) to carry out local histogram equalization in each small block, which increases the image contrast while retaining more details. This operation enhances the contrast of the image while suppressing noise to solve the problem of noisy images. Then, to solve the problems of similarity interference, occlusion and scale variation among fish, we propose a new method for fish single object tracking in aquaculture environments, called SiamFCA, by combining the coordinated attention mechanism (CA). The method is based on the siamese network framework and uses a modified AlexNet to extract the depth features of image pairs. And through the coordinate attention mechanism, the model can learn the relationship between different objects more accurately, which can improve the accuracy and robustness of the model when dealing with complex image scenes. Finally, the correlation operation between classification and regression is performed based on the region proposal network (RPN). By classifying, region suggestions can be labeled as positive or negative samples, and regression can calibrate the position and size of the region proposal to match targets more accurately. The experimental results show that the precision and success rate of the algorithm are 84.2 % and 62.4 %, respectively, which are better than the tracking methods such as SiamRPN and DaSiamRPN, etc. The SiamFCA runs at 83 fps, which meets the demand of real-time fish tracking in aquaculture.

Segmentation and tracking of vegetable plants by exploiting vegetable shape feature for precision spray of agricultural robots

AbstractFor robotic precision spray application in vegetable farms, simultaneous accurate instance segmentation and robust tracking of plants are of great importance and a prerequisite for the following spray action. With onboard cameras, agricultural robots can apply Multiple Object Tracking and Segmentation (MOTS) methods, for instance, segmentation and tracking of plants. By assigning a unique identification for each vegetable, it ensures the robot to spray each vegetable exactly once, while traversing along the farm rows. Conventional MOTS methods, which are mostly designed for tracking pedestrians or vehicles, usually extract their color and texture features for associating different targets in consecutive images. However, vegetable plants of the same species normally show similar color and texture, which leads to degraded performance when conventional MOTS methods are used. To solve the challenging problem of associating vegetables with similar color and texture in consecutive images, in this paper, a novel MOTS method that exploits contour and blob features is proposed, for instance, segmentation and tracking of multiple vegetable plants. The method takes advantage of the fact that different plants normally possess different shape contours and blob properties. With images captured on top of them, these features of the same plant show little difference in consecutively captured images. Comprehensive experiments including ablation studies are conducted, which prove its superior performance over two state‐of‐the‐art MOTS methods. Compared with the conventional MOTS methods, the proposed method is able to re‐identify objects which have gone out of the camera field of view and re‐appear again using the proposed data association strategy, which is important to ensure each vegetable be sprayed only once when the robot travels back and forth. Although the method is tested on lettuce farm, it can be applied to other similar vegetables, such as broccoli and canola. Both the code and the dataset of this paper are publicly released for the benefit of the community: https://github.com/NanH5837/LettuceMOTS.

Moving scene object tracking method based on deep convolutional neural network

The effect of target tracking is not ideal when facing various complex tracking scenarios such as non-rigid deformation of target, frequent occlusion, clutter of target background and interference of similar objects. In this paper, the feature based on deep convolutional neural network is used for target tracking in moving scenes, and a sliding window target segmentation method is proposed to study the impact of data normalization and data set expansion on the final result. In order to select more distinguishing features, principal component analysis is used to process the features of Deep Convolution Neural Network (DCNN), and the features of different network layers of DCNN are compared. The feature coding algorithm is studied, and the extracted DCNN features are encoded by Fisher Vectors algorithm, and compared with the locality-constrained linear encoding technique. Experiments show that the feature based on deep convolutional neural network in this paper can obtain higher accuracy than the traditional feature fusion method. According to the result analysis, the tracking accuracy of deep convolutional neural network algorithm is improved under the condition of illumination variation. In the case of local occlusion, the tracking accuracy is also improved.

Evaluating and modeling the effects of brightness on visual attention using multiple object tracking method

This article systematically evaluates and models how brightness affects performance in Multiple Object Tracking (MOT) within screen-based environments. MOT proficiency is essential across various fields, and comprehending the elements that impact MOT performance holds significance for occupational health and safety. While previous studies have scrutinized the influence of brightness on object recognition, its repercussions on MOT performance in screen-based environments remain comparatively less comprehended. This research aims to bridge this gap by delving into the distinct and combined impacts of brightness-related factors on MOT performance. Additionally, it seeks to construct a computational model that can forecast MOT performance across diverse brightness conditions. The outcomes of this study will offer valuable insights into core psychological processes, thereby steering the development of more efficient visual displays to enhance occupational health and safety. Our findings revealed a significant correlation between brightness levels and MOT performance, with optimal tracking observed at medium brightness levels. Additionally, complex object motion patterns were found to exacerbate the challenges of tracking in low brightness settings. These insights have direct implications for screen-based interfaces, suggesting the need for adaptive brightness settings based on the content's complexity and the user's task.

QDTrack: Quasi-Dense Similarity Learning for Appearance-Only Multiple Object Tracking.

Similarity learning has been recognized as a crucial step for object tracking. However, existing multiple object tracking methods only use sparse ground truth matching as the training objective, while ignoring the majority of the informative regions in images. In this paper, we present Quasi-Dense Similarity Learning, which densely samples hundreds of object regions on a pair of images for contrastive learning. We combine this similarity learning with multiple existing object detectors to build Quasi-Dense Tracking (QDTrack), which does not require displacement regression or motion priors. We find that the resulting distinctive feature space admits a simple nearest neighbor search at inference time for object association. In addition, we show that our similarity learning scheme is not limited to video data, but can learn effective instance similarity even from static input, enabling a competitive tracking performance without training on videos or using tracking supervision. We conduct extensive experiments on a wide variety of popular MOT benchmarks. We find that, despite its simplicity, QDTrack rivals the performance of state-of-the-art tracking methods on all benchmarks and sets a new state-of-the-art on the large-scale BDD100K MOT benchmark, while introducing negligible computational overhead to the detector.

Spatio-temporal interactive fusion based visual object tracking method

Visual object tracking tasks often struggle with utilizing inter-frame correlation information and handling challenges like local occlusion, deformations, and background interference. To address these issues, this paper proposes a spatio-temporal interactive fusion (STIF) based visual object tracking method. The goal is to fully utilize spatio-temporal background information, enhance feature representation for object recognition, improve tracking accuracy, adapt to object changes, and reduce model drift. The proposed method incorporates feature-enhanced networks in both temporal and spatial dimensions. It leverages spatio-temporal background information to extract salient features that contribute to improved object recognition and tracking accuracy. Additionally, the model’s adaptability to object changes is enhanced, and model drift is minimized. A spatio-temporal interactive fusion network is employed to learn a similarity metric between the memory frame and the query frame by utilizing feature enhancement. This fusion network effectively filters out stronger feature representations through the interactive fusion of information. The proposed tracking method is evaluated on four challenging public datasets. The results demonstrate that the method achieves state-of-the-art (SOTA) performance and significantly improves tracking accuracy in complex scenarios affected by local occlusion, deformations, and background interference. Finally, the method achieves a remarkable success rate of 78.8% on TrackingNet, a large-scale tracking dataset.

Comparing Multiple Extended Object Tracking with Point Based Multi Object Tracking for LiDAR in a Maritime Context

In this paper, we compare a recently developed multiple extended object tracking method with a point-based tracker for maritime applications by evaluating the methods on both real and simulated LiDAR data. Being able to track other vessels is a key part of maritime situational awareness, and multi-object tracking is well suited to this task. Traditionally, multi-object tracking uses the point approximation to simplify the target tracking problem, meaning that an object is assumed to only generate a single measurement. With recent advances in sensor technology, this approximation is no longer valid. Multiple extended object tracking instead makes use of methods that estimates the extent of a target using all available measurements. However, the target models used in extended object tracking are by necessity more complex. We compare these two approaches on real and simulated LiDAR data, which was gathered by tracking smaller vessels. We find that the extended object method has greater performance on simulated data and experimental data with comparatively little wake clutter. However, on experimental data with more wake clutter, the point-based method outperforms the extended object method since it is less sensitive to the disturbance caused by wake clutter. We can conclude that multiple extended object tracking methods have the potential to be more accurate for LiDAR data, but since the models are more complex, measurement sources need to be modeled correctly.

Graph Attention Networks and Track Management for Multiple Object Tracking

Multiple object tracking (MOT) constitutes a critical research area within the field of computer vision. The creation of robust and efficient systems, which can approximate the mechanisms of human vision, is essential to enhance the efficacy of multiple object-tracking techniques. However, obstacles such as repetitive target appearances and frequent occlusions cause considerable inaccuracies or omissions in detection. Following the updating of these inaccurate observations into the tracklet, the effectiveness of the tracking model, employing appearance features, declines significantly. This paper introduces a novel method of multiple object tracking, employing graph attention networks and track management (GATM). Utilizing a graph attention network, an attention mechanism is employed to capture the relationships of nodes within the graph as well as node-to-node correlations across graphs. This mechanism allows selective focus on the features of advantageous nodes and enhances discriminability between node features, subsequently improving the performance and robustness of multiple object tracking. Simultaneously, we categorize distinct tracklet states and introduce an efficient track management method, which employs varying processing techniques for tracklets in diverse states. This method can manage occluded tracks in crowded scenes and improves tracking accuracy. Experiments conducted on three challenging public datasets (MOT16, MOT17, and MOT20) demonstrate that our method could deliver competitive performance.

Learning Template-Constraint Real-Time Siamese Tracker for Drone AI Devices via Concatenation

Significant progress has been made in object tracking tasks thanks to the application of deep learning. However, current deep neural network-based object tracking methods often rely on stacking sub-modules and introducing complex structures to improve tracking accuracy. Unfortunately, these approaches are inefficient and limit the feasibility of deploying efficient trackers on drone AI devices. To address these challenges, this paper introduces ConcatTrk, a high-speed object tracking method designed specifically for drone AI devices. ConcatTrk utilizes a lightweight network architecture, enabling real-time tracking on edge devices. Specifically, the proposed method primarily uses the concatenation operation to construct its core tracking steps, including multi-scale feature fusion, intra-frame feature matching, and dynamic template updating, which aim to reduce the computational overhead of the tracker. To ensure tracking performance in UAV tracking scenarios, ConcatTrk implements a learnable feature matching operator along with a simple and efficient template constraint branch, which enables accurate tracking by discriminatively matching features and incorporating periodic template updates. Results of comprehensive experiments on popular benchmarks, including UAV123, OTB100, and LaSOT, show that ConcatTrk has achieved promising accuracy and attained a tracking speed of 41 FPS on an edge AI device, Nvidia AGX Xavier. ConcatTrk runs 8× faster than the SOTA tracker TransT while using 4.9× fewer FLOPs. Real-world tests on the drone platform have strongly validated its practicability, including real-time tracking speed, reliable accuracy, and low power consumption.

Analisis Metode Kalman Filter, Particle Filter dan Correlation Filter Untuk Pelacakan Objek

Object tracking is a challenging in computer vision. Object tracking is divided into two, which can be one object or several objects, depending on the object being observed. The process of tracking an object in the form of one object is to estimate the target in the next sequence based on information from the first frame given. In object tracking in the form of single object tracking, there are five steps that are often used in discriminatory methods, including motion models, feature extraction, observation models, model updates and integration methods. Although various algorithms of object tracking are proposed, there are still failures in the object tracking process caused by occlusion, non-rigid target deformation, and other factors. This study proposes the implementation of the Kalman filter, particle filter, and correlation filter methods for object tracking in video data. The results of the implementation of the three methods can track objects in traffic video data and the script circuit video. In object tracking calculations and method analysis, the kalman filter gets 96.89% where the kalman method is better in terms of accuracy compared to other methods. Meanwhile, in the average performance of computation time, the correlation method gets 26.69 FPS, where the correlation method is superior compared to other competitor methods.
 Keywords – Kalman Filter; Particle Filter; Correlation Filter; Object Tracking; Object Tracking in Video

Use of threshold parameter variation for tropical cyclone tracking

Abstract. Assessing the capacity of numerical models to produce viable tropical cyclones, as well as assessing the climatological behavior of simulated tropical cyclones, requires an objective tracking method. These make use of parameter thresholds to determine whether a detected feature, such as a vorticity maximum or a warm core, is strong enough to indicate a tropical cyclone. The choice of parameter thresholds is generally subjective. This study proposes and assesses the parallel use of many threshold parameter combinations, combining a number of weaker and stronger values. The tracking algorithm succeeds in tracking tropical cyclones within the model data, beginning at their aggregation stage or shortly thereafter and ending when they interact strongly with extratropical flow and transition into extratropical cyclones or when their warm core decays. The sensitivity of accumulated cyclone energy to tracking errors is assessed. Tracking errors include the faulty initial detection and termination of valid tropical cyclones and systems falsely identified as tropical cyclones. They are found to not significantly impact the accumulated cyclone energy. Thus, the tracking algorithm produces an adequate estimate of the accumulated cyclone energy within the underlying data.

Deep features fusion for KCF-based moving object tracking

Real-time object tracking and occlusion handling are critical research areas in computer vision and machine learning. Developing an efficient and accurate object-tracking method that can operate in real-time while handling occlusion is essential for various applications, including surveillance, autonomous driving, and robotics. However, relying solely on a single hand-crafted feature results in less robust tracking. As a hand-crafted feature extraction technique, HOG effectively detects edges and contours, which is essential in localizing objects in images. However, it does not capture fine details in object appearance and is sensitive to changes in lighting conditions. On the other hand, the grayscale feature has computational efficiency and robustness to changes in lighting conditions. The deep feature can extract features that express the image in more detail and discriminate between different objects. By fusing different features, the tracking method can overcome the limitations of individual features and capture a complete representation of the object. The deep features can be generated with transfer learning networks. However, selecting the right network is difficult, even in real-time applications. This study integrated the deep feature architecture and hand-crafted features HOG and grayscale in the KCF method to solve this problem. The object images were obtained through at least three convolution blocks of transfer learning architecture, such as Xception, DenseNet, VGG16, and MobileNet. Once the deep feature was extracted, the HOG and grayscale features were computed and combined into a single stack. In the KCF method, the stacked features acquired the actual object location by conveying a maximum response. The result shows that this proposed method, especially in the combination of Xception, grayscale, and HOG features, can be implemented in real-time applications with a small center location error.

A fast matching method for skip displacement recognition of hoisting system

Determining if there is any coal left in the coal mine skip unloading is a challenging problem. We propose a new object tracking method to provide new ideas to solve the problem of whether there is residual coal in the skip unloading. In this study, a fast vision-based measurement framework about new diamond search normalized cross-correlation (NDS-NCC), including a movement prediction part and a template tracking part, is used. Firstly, a diamond search method with fast adaptive coarse localization is proposed to overcome the time-consuming problem of NCC template matching process. Secondly, we incorporate a template updating strategies that allows targets to be tracked consistently. The performance of the NDS-NCC algorithm was tested on the bench and at the coal mine site for balancing cylinders and wire rope tracking. The results show that the NDS-NCC algorithm has better efficiency (7 ms/frame) and accuracy (0.4512 px) in the coal mine field application.

PPE: Point position embedding for single object tracking in point clouds

AbstractExisting 3D single object tracking methods primarily extract features from the global coordinates of point clouds, overlooking the potential exploitation of their positional information. However, due to the unordered, sparse, and irregular nature of point clouds, effectively exploring their positional information presents a significant challenge. In this letter, the network is explicitly reformulated by introducing a point position embedding module in conjunction with a self‐attention coding module, replacing the use of global coordinate inputs. The proposed reformulation is further integrated into a top‐notch model M2‐Track, called Point Position Embedding (PPE) in this letter. Comprehensive empirical analysis are performed on the KITTI and NuScenes datasets. Experimental results show that the PPE surpasses M2‐Track by a large margin in overall performance. Especially for the challenging NuScenes dataset, the method attains the highest precision and success in all classes compared to state‐of‐the‐art methods. The code is available at https://github.com/GZHU‐DVL/PPE.

Real-time Tracking of Handheld Object from Color or Depth Images.

Object tracking during rehabilitation could help a therapist to evaluate a patient's movement and progress. Hence, we present an image-based method for real-time tracking of handheld objects due to its ease of use and availability of color or depth cameras. We use an efficient projective point correspondence method and generalize the use of precomputed spare viewpoint information to allow real-time tracking of a rigid object. The method runs at more than 30 fps on a CPU while achieving submillimeter accuracy on synthetic datasets and robust tracking on a semi-synthetic dataset.Clinical relevance Real-time, accurate, and robust tracking of an object using an image-based method is a promising tool for rehabilitation applications as it is practical for clinical settings.

GLT-T: Global-Local Transformer Voting for 3D Single Object Tracking in Point Clouds

Current 3D single object tracking methods are typically based on VoteNet, a 3D region proposal network. Despite the success, using a single seed point feature as the cue for offset learning in VoteNet prevents high-quality 3D proposals from being generated. Moreover, seed points with different importance are treated equally in the voting process, aggravating this defect. To address these issues, we propose a novel global-local transformer voting scheme to provide more informative cues and guide the model pay more attention on potential seed points, promoting the generation of high-quality 3D proposals. Technically, a global-local transformer (GLT) module is employed to integrate object- and patch-aware prior into seed point features to effectively form strong feature representation for geometric positions of the seed points, thus providing more robust and accurate cues for offset learning. Subsequently, a simple yet effective training strategy is designed to train the GLT module. We develop an importance prediction branch to learn the potential importance of the seed points and treat the output weights vector as a training constraint term. By incorporating the above components together, we exhibit a superior tracking method GLT-T. Extensive experiments on challenging KITTI and NuScenes benchmarks demonstrate that GLT-T achieves state-of-the-art performance in the 3D single object tracking task. Besides, further ablation studies show the advantages of the proposed global-local transformer voting scheme over the original VoteNet. Code and models will be available at https://github.com/haooozi/GLT-T.

Exploring the potential of Siamese network for RGBT object tracking

Siamese tracking is one of the most promising object tracking methods today due to its balance of performance and speed. However, it still performs poorly when faced with some challenges such as low light or extreme weather. This is caused by the inherent limitations of visible images, and a common way to cope with it is to introduce infrared data as an aid to improve the robustness of tracking. However, most of the existing RGBT trackers are variants of MDNet (Hyeonseob Nam and Bohyung Han, Learning multi-domain convolutional neural networks for visual tracking, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 4293–4302.), which have significant limitations in terms of operational efficiency. On the contrary, the potential of Siamese tracking in the field of RGBT tracking has not been effectively exploited due to the reliance on large-scale training data. To solve this dilemma, in this paper, we propose an end-to-end Siamese RGBT tracking framework that is based on cross-modal feature enhancement and self-attention (SiamFEA). We draw on the idea of migration learning and employ local fine-tuning to reduce the dependence on large-scale RGBT data and verify the feasibility of this approach, and then we propose a reliable fusion approach to efficiently fuse the features of different modalities. Specifically, we first propose a cross-modal feature enhancement module to exploit the complementary properties of dual-modality, followed by capturing non-local attention in channel and spatial dimensions for adaptive weighted fusion, respectively. Our network was trained end-to-end on the LasHeR (Chenglong Li, Wanlin Xue, Yaqing Jia, Zhichen Qu, Bin Luo, Jin Tang, LasHeR: A Large-scale High-diversity Benchmark for RGBT Tracking, CoRR abs/2104.13202, 2021) training set and reached new SOTAs on GTOT (C. Li, H. Cheng, S. Hu, X. Liu, J. Tang, L. Lin, Learning collaborative sparse representation for grayscale-thermal tracking, IEEE Trans. Image Process, 25 (12) (2016) 5743–5756.), RGBT234 (C. Li, X. Liang, Y. Lu, N. Zhao, and J. Tang, “Rgb-t object tracking: Benchmark and baseline,” Pattern Recognition, vol. 96, p. 106977, 2019.), and LasHeR (Chenglong Li, Wanlin Xue, Yaqing Jia, Zhichen Qu, Bin Luo, Jin Tang, LasHeR: A Large-scale High-diversity Benchmark for RGBT Tracking, CoRR abs/2104.13202, 2021) while running in real-time.

Coastal Ship Tracking with Memory-Guided Perceptual Network

Coastal ship tracking is used in many applications, such as autonomous navigation, maritime rescue, and environmental monitoring. Many general object-tracking methods based on deep learning have been explored for ship tracking, but they often fail to accurately track ships in challenging scenarios, such as occlusion, scale variation, and motion blur. We propose a memory-guided perception network (MGPN) to address these issues. MGPN has two main innovative improvements. The dynamic memory mechanism (DMM) in the proposed method stores past features of the tracked target to enhance the model’s feature fusion capability in the temporal dimension. Meanwhile, the hierarchical context-aware module (HCAM) enables the interaction of different scales, global and local information, to address the scale discrepancy of targets and improve the feature fusion capability in the spatial dimension. These innovations enhance the robustness of tracking and reduce inaccuracies in the bounding boxes. We conducted an in-depth ablation study to demonstrate the effectiveness of DMM and HCAM. Finally, influenced by the above two points, MGPN has achieved state-of-the-art performance on a large offshore ship tracking dataset, which contains challenging scenarios such as complex backgrounds, ship occlusion, and varying scales.

Experimentelle und numerische Untersuchung von Fluid-Partikel-Interaktionen in Wasser

ZusammenfassungFür die Entwicklung verbesserter Modelle zur Beschreibung des Sedimenttransports in Fließgewässern ist das grundlegende Verständnis der Interaktion zwischen den festen Partikeln und dem bewegten Fluid (Wasser) bedeutend. In diesem Beitrag werden aktuelle Entwicklungen im Bereich der Fluid-Partikel-Interaktion anhand von zwei konkreten Forschungsarbeiten von Gold et al. (2023) und Worf et al. (2022) vorgestellt. Die experimentelle Forschungsarbeit von Gold et al. (2023) untersucht mithilfe von Messmethoden auf dem neuesten Stand der Technik die Strömung rund um Kugeln verschiedener Dichte, die in anfangs ruhendem Wasser oszillieren. Über den gesamten Dichtebereich (m^{*}=rho _{S}/rho _{F}= 1,14; 14,95; Dichteverhältnis von Solid und Fluid) konnte für das Kugelpendel eine vergleichbare Charakteristik der Wirbelablösung festgestellt werden. Ebenso wird eine neue Methode der Objektverfolgung (DOT) vorgestellt, welche eine zeitlich und räumlich aufgelöste Analyse der Strömungsstrukturen im Fluidfeld ermöglicht. Gold et al. (2023) zeigen, dass es während der ersten Periode zur Wirbelablösung kommt. Dieser Wirbel pflanzt sich in Richtung Beckenboden fort und dissipiert letztendlich. Weiters wurde ein Dämpfungsoptimum des Kugelpendels im Bereich von m^{*}=2{,}50 beobachtet. Außerdem wurde ein Experiment von einem Zylinderpendel mit m^{*}=4{,}98 mittels Immersed-Boundary-Methode numerisch untersucht. Dabei wurde der Vorgang der Bildung und Ablösung bis hin zum Zerfall eines Wirbelrings beschrieben. Zusätzlich wurde bei dieser Untersuchung von Worf et al. (2022) die Bildung von Randwirbeln beobachtet, beschrieben und mit der anfänglichen Entwicklung der dreidimensionalen Strömung und „zusätzlichen Masse“ in Verbindung gebracht.

Target recognition and tracking system based on UAV platform

Machine learning can extract the features of the target from the known image dataset and detect the target quickly by using the trained deep neural network. Due to UAVs’ fast flight speed and long detection distance, real-time and accurate target detection requirements are relatively high. For such scenarios,a ground object recognition and tracking method that integrates the YOLOv5 target detection algorithm and KCF target tracking algorithm are proposed. The combination of the two algorithms can meet the requirements of recognition speed, recognition rate, and tracking accuracy in the current scene. Two algorithms are run simultaneously in the recognition process, the detection frame with the highest confidence is selected as the final detection result, and the detected target is tracked by the KCF target tracking algorithm based on correlation filtering. The experimental results show that this method not only meets the requirements of recognition distance, recognition accuracy, and recognition speed of ground target but also can track a ground target in real-time, with strong robustness and real-time performance.