Proposed Tracking Algorithm Research Articles

UnifiedTT: Visual tracking with unified transformer

Target tracking is an important research task in computer vision. Existing tracking algorithms based on Siamese networks often suffer from the problem of information redundancy between adjacent frames and a lack of ability to capture global dependencies. When similar backgrounds appear around the target, the tracking performance usually significantly decreases. Although target tracking algorithms based on deep convolution and the Transformer have partially addressed these issues, achieving a good balance between the two remains a challenge. In this work, we propose a unified convolution and self-attention Siamese network for target tracking. By utilizing a feature extraction backbone network based on integrated convolution and self-attention styles, we are able to capture globally important regions and key frames while greatly reducing local redundant computations, thereby improving tracking performance. We apply this approach to the task of target tracking and enhance the feature extraction capabilities of both the target template and search template. Experimental results show that our proposed tracking algorithm outperforms some recent classical tracking algorithms, especially achieving improvements of 10.7 % on the high-diversity dataset GOT-10K and 24.7 % on the large-scale and high-quality LaSOT dataset.

Siamese Multi-Scale Adaptive Search Network for Remote Sensing Single-Object Tracking

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.

A Hybrid Excitation Model Based Lightweight Siamese Network for Underwater Vehicle Object Tracking Missions

Performing object tracking tasks and efficiently perceiving the underwater environment in real time for underwater vehicles is a challenging task due to the complex nature of the underwater environment. A hybrid excitation model based lightweight Siamese network is proposed to solve the mismatch between underwater objects with limited characteristics and complex deep learning models. The lightweight neural network is applied to the residual network in the Siamese network to reduce the computational complexity and cost of the model while constructing a deeper network. In addition, to deal with the changeable complex underwater environment and consider the timing of video tracking, the global excitation model (HE module) is introduced. The model adopts the excitation methods of space, channel, and motion to improve the accuracy of the algorithm. Based on the designed underwater vehicle, the underwater target tracking and target grabbing experiments are carried out, and the experimental results show that the proposed tracking algorithm has a high tracking success rate.

Global Context Attention for Robust Visual Tracking.

Although there have been recent advances in Siamese-network-based visual tracking methods where they show high performance metrics on numerous large-scale visual tracking benchmarks, persistent challenges regarding the distractor objects with similar appearances to the target object still remain. To address these aforementioned issues, we propose a novel global context attention module for visual tracking, where the proposed module can extract and summarize the holistic global scene information to modulate the target embedding for improved discriminability and robustness. Our global context attention module receives a global feature correlation map to elicit the contextual information from a given scene and generates the channel and spatial attention weights to modulate the target embedding to focus on the relevant feature channels and spatial parts of the target object. Our proposed tracking algorithm is tested on large-scale visual tracking datasets, where we show improved performance compared to the baseline tracking algorithm while achieving competitive performance with real-time speed. Additional ablation experiments also validate the effectiveness of the proposed module, where our tracking algorithm shows improvements in various challenging attributes of visual tracking.

Fish Tracking and Continual Behavioral Pattern Clustering Using Novel Sillago Sihama Vid (SSVid)

Aquaculture provides food security to many developing countries and enhances the socio-economic conditions of the fishermen. To enhance the productivity of the aquaculture, it is necessary to maintain stress free controlled eco-system for the fishes. For recognising the stress in fishes, behaviour analysis of fishes via tracking is imperative. Early detection of stress in fish facilitates fishermen to take precautionary measures promptly. Computer vision-based fish behaviour analysis of economically important fish species is challenging due to the lack of datasets, occlusions, rapid changes in swim directions etc. The present study proposes a multiple fish video dataset of an economically important species in a controlled environment, namely Sillago Sihama-Vid with accurate annotations. The study emulates the natural environment of Sillago Sihama in a large aquarium. This work proposes a novel fish tracking algorithm that incorporates swim direction information in addition to temporal, appearance, and spatial information. The inclusion of swim direction information reduces the number of identity switches. Comparative performance analysis of the proposed tracking algorithm with the conventional methods on the developed dataset highlights the performance efficiency. The proposed method has a clear performance improvement in MOTA, MOTP, IDSW and MT with respect to the other compared methods. The study also presents a novel unsupervised continual behaviour modelling strategy to model the evolving behaviours of the fishes. Further, interpretation of fish behaviour from the proposed behaviour modelling is performed to highlight the reliability of the proposed method. The significance of the proposed method is that, it is independent of training and labelled data. In addition, the method represents an innovative alternative to capture all the non observable behaviours of the fishes. The proposed tracking and behaviour modelling strategy act as a benchmark for developing algorithms to study fish behaviour via tracking. Finally, the dataset provides an opportunity for developing computer vision-based models to analyse the different behaviours of fish Sillago Sihama.

SGAT: Shuffle and graph attention based Siamese networks for visual tracking.

Siamese-based trackers have achieved excellent performance and attracted extensive attention, which regard the tracking task as a similarity learning between the target template and search regions. However, most Siamese-based trackers do not effectively exploit correlations of the spatial and channel-wise information to represent targets. Meanwhile, the cross-correlation is a linear matching method and neglects the structured and part-level information. In this paper, we propose a novel tracking algorithm for feature extraction of target templates and search region images. Based on convolutional neural networks and shuffle attention, the tracking algorithm computes the similarity between the template and a search region through a graph attention matching. The proposed tracking algorithm exploits the correlations between the spatial and channel-wise information to highlight the target region. Moreover, the graph matching can greatly alleviate the influences of appearance variations such as partial occlusions. Extensive experiments demonstrate that the proposed tracking algorithm achieves excellent tracking results on multiple challenging benchmarks. Compared with other state-of-the-art methods, the proposed tracking algorithm achieves excellent tracking performance.

Landmark tracking in 4D ultrasound using generalized representation learning

PurposeIn this study, we present and validate a novel concept for target tracking in 4D ultrasound. The key idea is to replace image patch similarity metrics by distances in a latent representation. For this, 3D ultrasound patches are mapped into a representation space using sliced-Wasserstein autoencoders.MethodsA novel target tracking method for 4D ultrasound is presented that performs tracking in a representation space instead of in images space. Sliced-Wasserstein autoencoders are trained in an unsupervised manner which are used to map 3D ultrasound patches into a representation space. The tracking procedure is based on a greedy algorithm approach and measuring distances between representation vectors to relocate the target . The proposed algorithm is validated on an in vivo data set of liver images. Furthermore, three different concepts for training the autoencoder are presented to provide cross-patient generalizability, aiming at minimal training time on data of the individual patient.ResultsEight annotated 4D ultrasound sequences are used to test the tracking method. Tracking could be performed in all sequences using all autoencoder training approaches. A mean tracking error of 3.23 mm could be achieved using generalized fine-tuned autoencoders. It is shown that using generalized autoencoders and fine-tuning them achieves better tracking results than training subject individual autoencoders.ConclusionIt could be shown that distances between encoded image patches in a representation space can serve as a meaningful measure of the image patch similarity, even under realistic deformations of the anatomical structure. Based on that, we could validate the proposed tracking algorithm in an in vivo setting. Furthermore, our results indicate that using generalized autoencoders, fine-tuning on only a small number of patches from the individual patient provides promising results.

Deep Feature Based Siamese Network for Visual Object Tracking

One of the most important and challenging research subjects in computer vision is visual object tracking. The information obtained from the first frame consists of limited and insufficient information to represent an object. If prior information about robust representation that can represent an object well is not sufficient, object tracking fails when not robustly responding to changes in features of the target object according to various factors, namely shape, illumination variation, and scene distortion. In this paper, a real-time single object tracking algorithm is proposed based on a Siamese network to solve this problem. For the object feature extraction, we designed a fully convolutional neural network that removes a fully connected layer and configured a convolution block consisting of a bottleneck structure that preserves the information in a previous layer. This network was designed as a Siamese network, while a regional proposal network was combined at the end of the network for object tracking. The ImageNet Large-Scale Visual Recognition Challenge 2017 dataset was used to train the network in the pre-training phase. Then, in the experimental phase, the object tracking benchmark dataset was used to quantitatively evaluate the network. The experimental results revealed that the proposed tracking algorithm produced more competitive results compared to other tracking algorithms.

Object Recognition and Tracking in Moving Videos for Maritime Autonomous Surface Ships

In autonomous driving technologies, a camera is necessary for establishing a path and detecting an object. Object recognition based on images from several cameras is required to detect impediments in autonomous ships. Furthermore, in order to avoid ship collisions, it is important to follow the movements of recognized ships. In this paper, we use the Singapore Maritime Dataset (SMD) and crawling image for model training. Then, we present four YOLO-based object recognition models and evaluate their performance in the maritime environment. Then, we propose a tracking algorithm to track the identified objects. Specially, in evaluation with high-motion video, the proposed tracking algorithm outperforms deep simple online and real-time tracking (DeepSORT) in terms of object tracking accuracy.

Real-Time Occlusion-Robust Deformable Linear Object Tracking With Model-Based Gaussian Mixture Model.

Tracking and manipulating deformable linear objects (DLOs) has great potential in the industrial world. However, estimating the object's state is crucial and challenging, especially when dealing with heavy occlusion situations and physical properties of different objects. To address these problems, we introduce a novel tracking algorithm to observe and estimate the states of DLO. The proposed tracking algorithm is based on the Coherent Point Drift (CPD), which registers the observed point cloud, and the finite element method (FEM) model encodes physical properties. The Gaussian mixture model with CPD regularization generates constraints to deform a given FEM model into desired shapes. The FEM model encodes the local structure, the global topology, and the material property to better approximate the deformation process in the real world without using simulation software. A series of simulations and real data tracking experiments have been conducted on deformable objects, such as rope and iron wire, to demonstrate the robustness and accuracy of our method in the presence of occlusion.

Multifeature Fusion Tracking Algorithm Based on Self-Associative Memory Learning Mechanism

A multi-feature fusion tracking algorithm updated with a self-associative memory learning mechanism is proposed to address the problems of short-time disappearance, re-emergence of the target and instability of single features in the kernelized correlation filtering algorithm. When extracting features, directional gradient histogram features, color features, and scale invariant features are fused instead of single features to collect more features of the target and increase the feature robustness. In the detection stage, the bimodal detection is proposed to judge whether the target model needs updating. Bimodal detection is used to judge the maximum target response in the search domain and predict the location of the target in the next frame. The self-associative memory learning mechanism was added into the updating template, and the original algorithm framework was improved to cope with the change of target model. The new algorithm update is biogenic, can recover fragment information, deal with complex and changeable tracking situation. Simulation experiments were conducted on the OTB50, OTB100, and UAV123 video datasets for the classical and new algorithms. The simulation verified that the proposed tracking algorithm has a high success and accuracy rate, which has research value. The tracking success rate improved by 23.6% and the accuracy rate improved by 18.8%.

Tracking and Track Management of Extended Targets in Range-Doppler Using Range-Compressed Airborne Radar Data

Ship tracking facilitates a comprehensive insight into maritime traffic situations and ensures its safety and security. However, with the current operational surveillance systems, detecting several maritime threats is still a major challenge. In this article, we propose a supportive ship tracking concept using an airborne-based radar sensor. The proposed tracking algorithm is suitable for dense multitarget scenarios. Tracking is performed in the range-Doppler domain. The primary advantage of using the range-Doppler domain is that ships even with low radar cross section moving with certain line-of-sight velocity appear out of the clutter region, thus improving their detectability. In addition, a powerful track management system is also developed to handle false targets. The simulated and real experimental results from the DLR’s airborne radar sensors, F-SAR and DBFSAR, are presented to prove the concept.

Efficient joint model learning, segmentation and model updating for visual tracking

The Tracking-by-segmentation framework is widely used in visual tracking to handle severe appearance change such as deformation and occlusion. Tracking-by-segmentation methods first segment the target object from the background, then use the segmentation result to estimate the target state. In existing methods, target segmentation is formulated as a superpixel labeling problem constrained by a target likelihood constraint, a spatial smoothness constraint and a temporal consistency constraint. The target likelihood is calculated by a discriminative part model trained independently from the superpixel labeling framework and updated online using historical tracking results as pseudo-labels. Due to the lack of spatial and temporal constraints and inaccurate pseudo-labels, the discriminative model is unreliable and may lead to tracking failure. This paper addresses the aforementioned problems by integrating the objective function of model training into the target segmentation optimization framework. Thus, during the optimization process, the discriminative model can be constrained by spatial and temporal constraints and provides more accurate target likelihoods for part labeling, and the results produce more reliable pseudo-labels for model learning. Moreover, we also propose a supervision switch mechanism to detect erroneous pseudo-labels caused by a severe change in data distribution and switch the classifier to a semi-supervised setting in such a case. Evaluation results on OTB2013, OTB2015 and TC-128 benchmarks demonstrate the effectiveness of the proposed tracking algorithm.

Object Tracking in Satellite Videos: Correlation Particle Filter Tracking Method With Motion Estimation by Kalman Filter

Object tracking in satellite videos faces various challenges such as target occlusion, target rotation, and background clutter. This study proposes a correlation particle filter algorithm with motion estimation for object tracking in satellite videos. The tracker, called CPKF, combines the strengths of the correlation, particle, and Kalman filters. Compared with existing tracking methods based on correlation filters, the proposed tracker has three major advantages: (1) Particle sampling, and motion estimation build robustness against partial and complete occlusion. (2) Color histogram model makes it robust to target rotation. (3) Fusion of multiple feature response maps effectively handle background clutter and low contrast. The experimental results demonstrate that the proposed tracking algorithm performs better than state-of-the-art methods.

FPGA implementation of multi-dimensional Kalman filter for object tracking and motion detection

Object tracking and motion detection are the major challenges in the real-time image and video processing applications. There are several tracking and prediction algorithms available to estimate and predict the state of a system. Kalman filter is the most widely used prediction algorithm as it is very simple, efficient and easy to implement for linear measurements. However, these types of filter algorithms are customized on hardware platforms such as Field-Programmable Gate Arrays (FPGAs) and Graphic Processing Units (GPUs) to achieve design requirements for embedded applications. In this work, a multi-dimensional Kalman filter (MDKF) algorithm is proposed for object tracking and motion detection. The numerical analysis of proposed tracking algorithm achieves competitive tracking performance in contrast with state-of-the-art tracking algorithms trained on standard benchmarks. Furthermore, MDKF is implemented on Xilinx Zynq™-7000 System-on-a chip (SoC). The implementation of MDKF on SoC performs 2× times tracking speed than that of software approach. The experimental results provide resource utilization of about 61.43% of Block RAMs (BRAMs), 90.09% of DSPs, 83.27% of Look-up tables (LUTs) and 82.35% of logic cells operating at 140 MHz with power consumption of 780 mW which outperforms previous related methods.

Improved P&O MPPT algorithm with efficient open-circuit voltage estimation for two-stage grid-integrated PV system under realistic solar radiation

Considering the vast improvement of photovoltaics (PVs) efficiency, this paper proposes a robust modified perturb and observe (MPO) maximum power point tracking (MPPT) algorithm. The control strategy of the proposed MPO-MPPT algorithm is based on dividing the solar module/cell P-V curve into four operating areas depending on the open-circuit voltage estimation method. The two areas, which are located far from the maximum power point (MPP), utilize large voltage fixed step-size that improves the tracking speed of the PV system. Otherwise, a small step-size P&O MPPT algorithm is used to minimize the steady-state oscillations for the other two areas due to the close locations of these areas to the MPP. As well, the proposed tracking algorithm eliminates the necessity of wide small step-size iterations by concentrating small step-size search area to 15% of the whole P-V operation region. To prove the effectiveness of the proposed MPO algorithm compared with the conventional MPPT techniques; sinusoidal, ramp, and one-day (10 hr.) irradiance profiles are applied to the solar PV system. The proposed MPO-MPPT algorithm based solar PV system has been built using MATLAB/SIMULINK software. Obviously, the system results confirm the theoretical analysis of the proposed algorithm, which boosts the PV system tracking efficiency to 99.7%.

Timestamp-Free Clock Parameters Tracking Using Extended Kalman Filtering in Wireless Sensor Networks

Clock synchronization is crucial for applications in wireless sensor networks, such as event scheduling and data fusion. In practical wireless networks, environmental changes cause the oscillator to be imperfect. Thus, it is necessary to track the nonlinear varying clock dynamically. In this paper, we propose a timestamp-free clock skew (frequency difference) and offset (time difference) joint tracking algorithm based on extended Kalman filter (EKF), which can be embedded into the general network data flow to achieve long-term synchronization without additional communication overhead. To further improve energy efficiency of clock synchronization, this paper develops an EKF clock skew tracking algorithm for silent nodes in receiver-only synchronization. For the situation that silent node overhears synchronization information from multiple active nodes, we present multi-information timestamp-free synchronization and multi-observation clock skew fusion scheme for silent nodes. Simulation results show the effectiveness of the proposed tracking algorithms.

A comprehensive approach to predict a rocket's impact with stochastic estimators and artificial neural networks

One of the current ways to continue space research is to launch ballistic rockets that carry scientific payloads. To improve the accuracy of the instantaneous evolution of the payload impact on the Earths surface, it is necessary to estimate indirect measures more efficiently. This study proposes a comprehensive approach to determine the impact point prediction of ballistic rocket payloads. This approach combines tracking algorithms that are based on stochastic estimators with artificial neural network (ANN) models. Four existing stochastic estimators, namely a recursive Kalman filter (RKF), an extended Kalman filter (EKF), an unscented Kalman filter (UKF), and a particle filter (PF) are compared with four proposed stochastic estimators. These include a recursive Kalman filter aided by an ANN (RKFN), an extended Kalman filter aided by an ANN (EKFN), an unscented Kalman filter aided by an ANN (UKFN), and a particle filter aided by an ANN (PFN). This study shows that the results that are obtained through the proposed tracking algorithms RKFN, EKFN, UKFN, and PFN are better than those of the existing tracking algorithms RKF, EKF, UKF, and PF. The proposed estimators can be efficient low-cost tools to mitigate inaccuracies during tracking up to the payload's impact.

Adaptive Gaussian Incremental Expectation Stadium Parameter Estimation Algorithm for Sports Video Analysis

In this paper, we propose an adaptive Gaussian incremental expectation stadium parameter estimation algorithm for sports video analysis and prediction through the study and analysis of sports videos. The features with more discriminative power are selected from the set of positive and negative templates using a feature selection mechanism, and a sparse discriminative model is constructed by combining a confidence value metric strategy. The sparse generative model is constructed by combining L1 regularization and subspace representation, which retains sufficient representational power while dealing with outliers. To overcome the shortcomings of the traditional multiplicative fusion mechanism, this paper proposes an adaptive selection mechanism based on Euclidean distance, which aims to detect deteriorating models in time during the dynamic tracking process and adopt corresponding strategies to construct more reasonable likelihood functions. Based on the Bayesian citation framework, the adaptive selection mechanism is used to combine the sparse discriminative model and the sparse generative model. Also, different online updating strategies are adopted for the template set and Principal Component Analysis (PCA) subspace to alleviate the drift problem while ensuring that the algorithm can adapt to the changes of target appearance in the dynamic tracking environment. Through quantitative and qualitative evaluation of the experimental results, it is verified that the algorithm proposed in this paper has stronger robustness compared with other classical algorithms. Our proposed visual object tracking algorithm not only outperforms existing visual object tracking algorithms in terms of accuracy, success rate, accuracy, and robustness but also achieve the performance required for real-time tracking in terms of execution speed on the central processing unit (CPU). This paper provides an in-depth analysis and discussion of the adaptive Gaussian incremental expectation stadium parameter estimation algorithm for sports video analysis. Using a variety of county-level algorithms for analysis and multiple solutions to improve the accuracy of the results, we obtain a more efficient and accurate algorithm.

A 500-Fps Pan-Tilt Tracking System With Deep-Learning-Based Object Detection

In this letter, we propose a fast mirror-drive pan-tilt target tracking system that can robustly track an object whose appearance varies in a complex background at 500 fps. By assuming a small image displacement between frames, which is a property of high-frame rate vision, we develop an fast object tracking algorithm by hybridizing the convolutional-neural-network (CNN) based object detection with template-matching (TM) based tracking operating at hundreds of frames per second (fps). For object tracking with high-speed visual feedback, the proposed tracking algorithm can remarkably reduce dozens-of-milliseconds-latency in the CNN-based object detection by simultaneously executing TM-based tracking for several images at consecutive frames within a few milliseconds. In the proposed pan-tilt tracking system, when the current tracked objects are occluded or out of the camera view, it can recognize objects to be newly tracked with CNN-based object detection at the rate of 33 fps with acceleration using graphic processing units (GPUs). Controlling the pan-tilt tracking system via visual feedback at 500 Hz, fast moving objects can be robustly tracked at the center of the camera view. The effectiveness of our method was experimentally demonstrated via several results when fast-moving pre-learned objects, such as toy cars were tracked in complex backgrounds.