Author:
Yuan Yue,Chu Jun,Leng Lu,Miao Jun,Kim Byung-Gyu
Abstract
AbstractThe methods combining correlation filters (CFs) with the features of convolutional neural network (CNN) are good at object tracking. However, the high-level features of a typical CNN without residual structure suffer from the shortage of fine-grained information, it is easily affected by similar objects or background noise. Meanwhile, CF-based methods usually update filters at every frame even when occlusion occurs, which degrades the capability of discriminating the target from background. A novel scale-adaptive object-tracking method is proposed in this paper. Firstly, the features are extracted from different layers of ResNet to produce response maps, and then, in order to locate the target more accurately, these response maps are fused based on AdaBoost algorithm. Secondly, to prevent the filters from updating when occlusion occurs, an update strategy with occlusion detection is proposed. Finally, a scale filter is used to estimate the target scale. The experimental results demonstrate that the proposed method performs favorably compared with several mainstream methods especially in the case of occlusion and scale change.
Funder
National Natural Science Foundation of China
Publisher
Springer Science and Business Media LLC
Subject
Electrical and Electronic Engineering,Information Systems,Signal Processing
Reference53 articles.
1. G. Zhang, J. Yang, W. Wang, Y. H. Hu, J. Liu, Adaptive visual target tracking algorithm based on classified-patch kernel particle filter. EURASIP J. Image Video Process.2019(1), 20 (2019).
2. B. -G. Kim, G. -S. Hong, J. -H. Kim, Y. -J. Choi, An efficient vision-based object detection and tracking using online learning. J. Multimed. Inf. Syst. (KMMS). 4:, 285–288 (2017).
3. B. -G. Kim, D. -J. Park, Novel target segmentation and tracking based on fuzzy membership distribution for vision-based target tracking system. Image Vis. Comput.24:, 1319–1331 (2006).
4. G. -S. Hong, S. -H. Yang, B. -G. KIm, Y. -S. Hwang, K. -K. Kwoni, Fast multi-feature pedestrian detection algorithm based on discrete wavelet transform for interactive driver assistance system. Multimed. Tools Appl.75:, 15229–15245 (2016).
5. S. Jung, Y. Kim, E. Hwang, Real-time car tracking system based on surveillance videos. EURASIP J. Image Video Process.2018(1), 133 (2018).
Cited by
68 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献