• Infrared and Laser Engineering
  • Vol. 50, Issue 1, 20200105 (2021)
Faling Chen1、2、3、4、5, Qinghai Ding1、6, Haibo Luo1、2、4、5, Bin Hui1、2、4、5, Zheng Chang1、2、4、5, and Yunpeng Liu1、2、4、5
Author Affiliations
  • 1Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China
  • 2Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang 110169, China
  • 3University of Chinese Academy of Sciences, Beijing 100049, China
  • 4Key Laboratory of Opto-Electronic Information Processing, Chinese Academy of Sciences, Shenyang 110016, China
  • 5The Key Lab of Image Understanding and Computer Vision, Shenyang 110016, China
  • 6Space Star Technology Co., LTD, Beijing 100086, China
  • show less
    DOI: 10.3788/IRLA20200105 Cite this Article
    Faling Chen, Qinghai Ding, Haibo Luo, Bin Hui, Zheng Chang, Yunpeng Liu. Anti-occlusion real time target tracking algorithm employing spatio-temporal context[J]. Infrared and Laser Engineering, 2021, 50(1): 20200105 Copy Citation Text show less

    Abstract

    An anti-occlusion real time target tracking algorithm employing spatio-temporal context was proposed to solve the problems of tracking instability or even failure, which were caused by illumination variation, background clutters, target deformation or occlusion. Firstly, in the framework of spatio-temporal context model, the adaptive dimensionality reduced color features were adopted to describe the target to promote the distinguish ability in complex scene. Secondly, the peak and the peak-to-sidelobe ratio of confidence map response were combined to evaluate the target tracking status. Then, occlusion was discriminated by the correlation coefficient between target templates. Finally, when the target tracking status fluctuated, the update speed of target model was reduced, and the target coordinates were corrected by the Kalman filter. When the target was occluded seriously, the target coordinates was predicted according to the Kalman filter, and the target model was stopped to update for recapturing and tracking the target again after occlusion released. 36 color sequences with multiple challenging attributes were selected to evaluate the performance of the proposed algorithm, and it was compared with other excellent target tracking algorithms. The experimental results demonstrated that this algorithm has strong anti-occlusion ability, and improved the robustness of target tracking effectively under the influence of disturbance factors such as illumination variation, background clutters and target deformation. Meanwhile, it met the real time requirement of target tracking.
    Faling Chen, Qinghai Ding, Haibo Luo, Bin Hui, Zheng Chang, Yunpeng Liu. Anti-occlusion real time target tracking algorithm employing spatio-temporal context[J]. Infrared and Laser Engineering, 2021, 50(1): 20200105
    Download Citation