Siamese network target tracking algorithm based on dynamic template updating

被引：0
|
作者：
Cai H. [1 ,2 ]
Wang X.-W. [1 ]
Fu Q. [3 ]
Wang W.-G. [4 ]
Li Y.-C. [3 ]
机构：
[1] School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun
[2] Changchun China Optical Science and Technology Museum, Changchun
[3] School of Opto-Electronic Engineering, Changchun University of Science and Technology, Changchun
[4] No.2 Department of Urology, The First Hospital of Jilin University, Changchun
来源：
Jilin Daxue Xuebao (Gongxueban)/Journal of Jilin University (Engineering and Technology Edition) | 2022年 / 52卷 / 05期
关键词：
Anchor-free; Computer vision; Dynamic update template; Siamese network; Target tracking;
D O I：
10.13229/j.cnki.jdxbgxb20200962
中图分类号：
学科分类号：
摘要：
In the complex scene, especially when the appearance of the target changes dramatically, the tracking accuracy will be seriously affected. Here, a siamese network target tracking algorithm based on dynamic update template is proposed. The algorithm uses Resnet-50 as the backbone network to extract the depth feature of the image, which enhances the feature extraction ability of the network. For the response map obtained by depth feature cross-correlation operation, the position and scale of the target object are predicted directly by anchor free method. For the template update, the template dynamic update strategy is used to determine whether the template is updated. If it needs to be updated, the template update subnetwork is used to estimate the best template for the next frame tracking. The experimental results show that the proposed algorithm has strong robustness against the drastic changes of the target appearance, and effectively improves the tracking accuracy and success rate. At the same time, the average rate of the algorithm reaches 32 frames per second, which meets the real-time requirements. © 2022, Jilin University Press. All right reserved.
引用
收藏
页码：1106 / 1116
页数：10
相关论文
共 27 条
  • [1] Tang Yi-ming, Liu Yu-fei, Huang Hong, Survey of single-target visual tracking algorithms, Measurement and Control Technology, 39, 8, pp. 21-34, (2020)
  • [2] Xu Y, Wang Z, Li Z, Et al., SiamFC++: towards robust and accurate visual tracking with target estimation guidelines, Proceedings of Thirty-Fourth the AAAI Conference on Artificial Intelligence, pp. 12549-12556, (2020)
  • [3] Zhang Z, Peng H., Deeper and wider siamese networks for real-time visual tracking, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4591-4600, (2019)
  • [4] Henriques J F, Caseiro R, Martins P, Et al., High-speed tracking with kernelized correlation filters, IEEE Transactions on Pattern Analysis and Machine Intel-ligence, 37, 3, pp. 583-596, (2015)
  • [5] Yang Xin, Xia Si-jun, Liu Dong-xue, Et al., Target tracking based on improved accelerated gradient under tracking-learning-detection framework, Journal of Jilin University(Engineering and Technology Edition), 48, 2, pp. 533-538, (2018)
  • [6] Cai Hua, Chen Guang-qiu, Liu Guang-wen, Et al., Novelty fragments-based target tracking with multiple instance learning under occlusions, Journal of Jilin University(Engineering and Technology Edition), 47, 1, pp. 281-287, (2017)
  • [7] Zhang Yao, Cai Hua, Li Xin-da, Et al., Spatio-temporal context face tracking algorithm based on Adaboost first frame detection, Journal of Jilin University(Science Edition), 58, 2, pp. 314-320, (2020)
  • [8] Danelljan M, Hager G, Khan F S, Et al., Convolutional features for correlation filter based visual tracking, IEEE International Conference on Computer Vision Workshop, pp. 621-629, (2015)
  • [9] Danelljan M, Bhat G, Shahbaz K, Et al., Eco: effificient convolution operators for tracking, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 6638-6646, (2017)
  • [10] Nam H, Han B., Learning multi-domain convolutional neural networks for visual tracking, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 4293-4302, (2015)
123