<title>Robust object tracking using camshift with multicue integration</title>

This paper presents a new methodology for improving the tracking of multiple targets in complex scenes. The new method,Motion Parameter Sharing, incorporates social motion information into tracking predictions. This is achieved by allowing a tracker to share motion estimates within groups of targets which have previously been moving in a coordinated fashion. The method is intuitive and, as well as aiding the prediction estimates, allows the implicit formation of 'social groups' of targets as a side effect of the process. The underlying reasoning and method are presented, as well as a description of how the method fits into the framework of a typical Bayesian tracking system. This is followed by some preliminary results which suggest the method is more accurate and robust than algorithms which do not incorporate the social information available in multiple target scenarios.

Particle filter (PF) based object tracking methods have been widely used in computer vision. However, traditional particle filter trackers cannot effectively distinguish the target from the background in complex scenes since they only exploit appearance information of observation to determine the target region. In this paper, we present an adaptive particle filter based on energy field (EPF), which makes good use of moving information of previous frames adaptively to track the target. Besides, we present the mechanism of result rectification to ensure the target region is accurate. Experiment results on several challenging video sequences have verified that the adaptive EPF method is compared very robust and effective with the traditional particle filter in many complicated scenes.KeywordsTrackingparticle filterprobabilistic approximationimage sequence analysisdynamic scenes

In this paper we propose a pixel-wise visual tracking method using a novel tri-model representation. The newly proposed tri-model is composed of three models, which each model learns the target object, the background, and other non-target moving objects online. The proposed method performs tracking by simultaneous estimation of the holistic position of the target object and the pixel-wise labels. By utilizing the information in the background and the foreground models as well as the target model, our method obtains robust results even under background clutters and partial occlusions in complex scenes. Furthermore, our method is able to give pixel-wise results, and uses them in the learning process to prevent drifting. The method is extensively tested against seven representative trackers both quantitatively and qualitatively showing promising results.

Target detection and tracking important in many applications including intelligent monitoring system, defense system and terminal guidance system. Aiming to solve the problem of simulated target tracking, this paper proposes an adaptive algorithm which uses the fusion of the spectral and morphological features of multispectral image to realize the target tracking based on the Particle Filter. Firstly, the target area is manually initialized in the multispectral image and the spectral and texture features of the target are extracted. Secondly, we build the adaptive tracking model of multiple features under the framework of Particle Filter. We validate the effectiveness of the proposed approach on the MATLAB platform. The results show that the proposed approach achieves accurate and stable multispectral target tracking in complex scenes by improving the efficiency of particles usage under defective tracking conditions, which is of great theoretical and practical values for the application of multispectral target tracking technology.

Planar object tracking is an important research area of visual target tracking. In order to improve the speed and accuracy of plane object tracking in complex scenes, this paper proposes a coarse-to-fine plane object tracking algorithm based on fully convolutional dense network. First, the estimated bounding box of the target object in the input image is calculated according to the DiMP50 algorithm; then SIFT is used to locate the key points on the image, and the feature descriptors corresponding to the key points are extracted from the feature map generated based on the cascaded thick and thin dense network. The nearest neighbor algorithm matches the feature descriptors to obtain the paired keypoint set; then the roughly positioned bounding box is used as the a priori box to screen the matching keypoints from coarse to fine, and the keypoints in the a priori box are retained, using RANSAC algorithm estimates the optimal homography matrix parameters for the key points; finally, the real value is calculated by the homography matrix to obtain the final prediction frame of the target. Validated on the POT dataset, the results show that the performance of the planar object tracking algorithm proposed in this paper is better than the classical algorithm, and the average tracking accuracy on all image sequences reaches 64.76%.

The widespread application of Unmanned Aerial Vehicles (UAVs) has raised serious public safety and privacy concerns, making UAV perception crucial for anti-UAV tasks. However, existing UAV tracking datasets pre-dominantly feature conspicuous objects and lack diversity in scene complexity and attribute representation, limiting their applicability to real-world scenarios. To overcome these limitations, we present the CST Anti-UAV, a new thermal infrared dataset specifically designed for Single Object Tracking (SOT) in Complex Scenes with Tiny UAVs (CST). It contains 220 video sequences with over 240k high-quality bounding box annotations, highlighting two key properties: a significant number of tinysized UAV targets and the diverse and complex scenes. To the best of our knowledge, CST Anti-UAV is the first dataset to incorporate complete manual frame-level attribute annotations, enabling precise evaluations under varied challenges. To conduct an in-depth performance analysis for CST Anti-UAV, we evaluate 20 existing SOT methods on the proposed dataset. Experimental results demonstrate that tracking tiny UAVs in complex environments remains a challenge, as the state-of-the-art method achieves only 35.92% state accuracy, much lower than the 67.69% observed on the Anti-UAV410 dataset. These findings underscore the limitations of existing benchmarks and the need for further advancements in UAV tracking research. The CST Anti-UAV benchmark is about to be publicly released, which not only fosters the development of more robust SOT methods but also drives innovation in anti-UAV systems.

We present an automatic video object tracking algorithm capable of dealing with multiple simultaneous objects. The tracking is based on interactions between high-level and low-level image analysis results. The high-level result is a partition defining video objects, and the low-level result is a partition formed by homogeneous regions. For each region, a set of characteristic descriptors is produced. These region descriptors, and not regions themselves, are used to track the regions (and thus the objects) along time. Track management issues such as appearance and disappearance of objects, splitting and partial occlusions are resolved through interactions between regions and objects. Defining the tracking based on the parts of objects, identified by region segmentation, has led to a flexible technique that exploits the nature of the video object tracking problem. Experimental results show that the proposed method is able to track multiple rigid and deformable objects in indoor and outdoor scenes.

In this paper, by analyzing the characteristics of infrared moving targets, a Symmetric Frame Differencing Target Detection algorithm based on local clustering segmentation is proposed. In consideration of the high real-time performance and accuracy of traditional symmetric differencing, this novel algorithm uses local grayscale clustering to accomplish target detection after carrying out symmetric frame differencing to locate the regions of change. In addition, the mean shift tracking algorithm is also improved to solve the problem of missed targets caused by error convergence. As a result, a kernel-based mean shift target tracking algorithm based on detection updates is also proposed. This tracking algorithm makes use of the interaction between detection and tracking to correct the tracking errors in real time and to realize robust target tracking in complex scenes. In addition, the validity, robustness and stability of the proposed algorithms are all verified by experiments on mid-infrared aerial sequences with vehicles as targets.

Foreground detection is one of the key steps in computer vision applications. Many foreground and background models have been proposed and achieved promising performance in static scenes. However, due to challenges such as dynamic background, irregular movement, and noise, most algorithms degrade sharply in complex scenes. To address the problem, we propose a motion-aware structured matrix factorization approach (MSMF), which integrates the structural and spatiotemporal motion information into a unified sparse-low-rank matrix factorization framework. Technologically, it has three main contributions: First, a variant of structured sparsity-inducing norm is proposed to constrain both structure and sparsity of foreground. The model is robust to the statistical variability of the underlying foreground pixels in complex scenes. Second, to capture the ambiguous pixels, a spatiotemporal cube-based motion trajectory is extracted for assisting matrix factorization. Finally, to solve the optimization problem of structured matrix factorization, we develop an augmented Lagrange multiplier method with the alternating direction strategy and Douglas-Rachford monotone operator splitting algorithm. Experiments demonstrate that the proposed approach achieves impressive performance in separating irregular moving foreground while suppressing the dynamic background and the noise, and outperforms some state-of-the-art algorithms.

To realize the real-time tracking of the shoot point from light pen, we choose Camshift tracking algorithm, but it is difficult to track the point with dynamic and complicated background and in case that there is similar color with the shoot point from light pen. So, an improved method is proposed. Firstly, a red filter which weakens the effect of background, and makes the object prominent is introduced to improve the effect of color extraction based on Camshift algorithm. The red filter makes the algorithm overcome disturbance of the background color, and makes the automatic tracking of shoot point with high real-time performance and strong robustness. Then the automatic tracking of the shoot point from light pen is realized by changing the manner of selecting region. Finally, we realize the real-time tracking of the shoot point from light pen using OpenCV. The experimental results indicate that color object is efficiently tracked even with dynamic and complicated background by the new method which is presented in the paper.

The vehicle detection and tracking in driving assistance system are ordinarily achieved by the optical or radar technology. In this work, we explore video processing for driving assistance system. An object’s detection and tracking system based on the Faster R-CNN and Camshift algorithm is proposed, and Kalman filtering algorithm is used to predict the position of objects. Our system differs with other object-tracking algorithms based on deep learning, which directly detects each frame and has no tracking part, such as YOLO and SSD. We also introduce an evaluation criterion for driving assistance system according to the dangerous level of the surrounding cars. Through the experiments on the videos recorded by drive recorder, we show that our approach can achieve better performance in complex scenes than YOLO or SSD and can satisfy the real-time requirement by processing 30 frames per second.

In dynamic settings such as security, autonomous driving, and robotics, effective motion detection and classification are crucial for accurate tracking amidst target and background movements. Traditional approaches, typically designed for static environments, face challenges in complex scenes with multiple types of motion. This research presents a robust algorithm for motion detection in fully dynamic scenarios, utilizing the macro block technique to generate motion vectors, followed by motion vector analysis to classify distinct types of motion. These include camera motion, object motion, background motion, and complex motion, where both background and foreground move simultaneously. By segmenting and categorizing these motion types, the proposed approach improves detection precision in cluttered, real-world environments. Furthermore, the algorithm adapts to lighting variations and is independent of specific sensor setups. Moreover, the high agreement with human judgment, achieving a 90% accuracy rate, underscores the model's robustness and potential applicability in real-world scenarios where dynamic backgrounds are prevalent. This establishes a framework for future research in dynamic motion detection and classification.

This paper introduces robust tracking algorithm for fast and erratic moving object. CAMSHIFT algorithm has less computation and efficient performance for object tracking. However, the method fails to track a object if it moves out of search window by fast velocity and/or large movement. The size of the search window in CAMSHIFT algorithm should be selected manually also. To solve these problems, we propose an efficient prediction technique for fast movement of object using Kalman Filter with automatic initial setting and variable configuration technique for search window. The proposed method is compared to the traditional CAMSHIFT algorithm for searching and tracking performance of objects on test image frames.

Robust hand tracking via novel multi-cue integration

Background modeling is a fundamental yet challenging issue in video surveillance. Traditional methods usually adopt single feature type to solve the problem, while the performance is usually unsatisfactory when handling complex scenes. In this paper, we propose a multi-scale framework, which combines both texture and intensity feature, to achieve a robust and accurate solution. Our contributions are three folds: first, we provide a multi-scale analysis for the issue; second, for texture feature we propose a novel texture operator named Scale-invariant Center-symmetric Local Ternary Pattern, and a corresponding Pattern Adaptive Kernel Density Estimation technique for its probability estimation; third, we design a Simplified Gaussian Mixture Models for intensity feature. Our method is tested on several complex real world videos with illumination variation, soft shadows and dynamic backgrounds. The experimental results clearly demonstrate that our method is superior to the previous methods.