Xinyue Zhang
ABSTRACT
A single feature cannot adapt to the dynamic changes of the scene during video target tracking. This paper, to address this issue, first studies the tracking algorithm of multi-feature fusion, which uses the complementarity between different features to better adapt to the scene changes. On this basis, the APCE anti-occlusion criterion is added to enable the algorithm to resist the influence of target occlusion on tracking to a certain extent. The experimental results show that the average tracking accuracy of the proposed algorithm is about 0.779, which is about 2% higher than that of the SAMF algorithm, and the tracking success rate can be as high as 72%.
- Bolme, D.S., Beveridge,et al . Visual object tracking using adaptive correlation filters[P].// Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on,2010.Google Scholar
- Henriques J F, Caseiro R, Exploiting the circulant structure of tracking-by- detection with kernels [C]// ECCV, 2012:702-715.Google Scholar
- Henriques J F, Caseiro R, High-Speed Tracking with Kernelized Correlation Filters.[J].IEEE transactions on pattern analysis and machine intelligence, 2015,37(3).Google ScholarDigital Library
- Danelljan M,Khan, F S,et al. Adaptive Color Attributes for Real-Time Visual Tracking[P].Computer Vision and Pattern Recognition (CVPR), 2014 IEEE Conference on,2014.Google Scholar
- LI Y,ZHU J.A scale adaptive kernel correlation filter tracker with feather integration[C]// Procedings of European Conference on Computer Vision.Germany:Springer,2014:254-265.Google Scholar
- M. Wang, Y. Liu and Z. Huang,et al.Large Margin Object Tracking with Circulant Feature Maps[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), USA:Honolulu, HI, 2017: 4800-4808.Google Scholar
- KHALID S.KHALIL T,NASREEEN S.A survey of feature selection and feature extraction techniques in machine learning[C].Science and Information Conference (SAI),2014:372-378.Google Scholar
- Struck: Structured Output Tracking with Kernels.[J].Hare Sam,Golodetz Stuart,Saffari Amir,Vineet Vibhav,Cheng Ming-Ming,Hicks Stephen L,Torr Philip H S. IEEE transactions on pattern analysis and machine intelligence . 2016(10).Google Scholar
- Xiao-Qin Zhang,Run-Hua Jiang,Chen-Xiang Fan,Tian-Yu Tong,Tao Wang,Peng-Cheng Huang.Advances in Deep Learning Methods for Visual Tracking:Literature Review and Fundamentals[J].International Journal of Automation and Computing,2021,18(03):311-333.Google Scholar
- Zirui. Zhang and Jun. Cheng, "Multi-Camera Tracking Helmet System," Journal of Image and Graphics, Vol. 1, No. 2, pp. 76-79, June 2013. doi: 10.12720/joig.1.2.76-79.Google ScholarCross Ref
- Jorge Henrique Busatto Casagrande and Marcelo Ricardo Stemmer, "Abnormal Motion Analysis for Tracking-Based Approaches Using Region-Based Method with Mobile Grid," Journal of Image and Graphics, Vol. 2, No. 1, pp. 22-27, June 2014. doi: 10.12720/joig.2.1.22-27.Google ScholarCross Ref
- Karthik Dinesh and Sumana Gupta, "Video Stabilization, Camera Motion Pattern Recognition and Motion Tracking Using Spatiotemporal Regularity Flow," Journal of Image and Graphics, Vol. 2, No. 1, pp. 33-40, June 2014. doi: 10.12720/joig.2.1.33-40.Google ScholarCross Ref
- Saad A. Yaseen and Sreela Sasi, "Robust Algorithm for Object Detection and Tracking in a Dynamic Scene," Journal of Image and Graphics, Vol. 2, No. 1, pp. 41-45, June 2014. doi: 10.12720/joig.2.1.41-45.Google ScholarCross Ref
- Chin-Shiuh Shieh, Yong-Shixa Jhan, Yuan-Li Liu, Mong-Fong Horng, and Tsair-Fwu Lee, "Video Object Tracking with Heuristic Optimization Methods," Journal of Image and Graphics, Vol. 6, No. 2, pp. 95-99, December 2018. doi: 10.18178/joig.6.2.95-99.Google ScholarCross Ref
Index Terms
- Research and Implementation of Multi-feature Tracking Algorithms
Recommendations
Long Time Target Tracking Algorithm Based on Multi Feature Fusion and Correlation Filtering
AIPR '21: Proceedings of the 2021 4th International Conference on Artificial Intelligence and Pattern RecognitionThis paper considers the problem of long-term target tracking in complex scenes when tracking failures are unavoidable due to illumination change, target deformation, scale change, motion blur, and other factors. More specifically, we propose a target ...
The Research on Multi-angle Face Tracking Based on Multi-feature Fusion and KCF
ICCPR '19: Proceedings of the 2019 8th International Conference on Computing and Pattern RecognitionA method for multi-angle face tracking based on multi-feature fusion and KCF algorithm is proposed in this paper. Firstly, Haar-like, MB-LBP and HOG features are combined to detect the face region, which is regarded as the initial tracking search ...
Stereo Vision Tracking System
ICFCC '09: Proceedings of the 2009 International Conference on Future Computer and CommunicationTarget tracking has become an area of interest in recent years. Target tracking is able to widen the perspective and view of the static camera to track a target, providing basic artificial intelligence features for robots, and serve as the platform for ...
Comments