Skip to main content

Robust Tracking For Motion Blur Via Context Enhancement

Zhongjie Mao, Xi Chen, Yucheng Wang, Jia Yan

  • SPS
    Members: Free
    IEEE Members: $11.00
    Non-members: $15.00
    Length: 00:05:00
20 Sep 2021

Motion blur is pervasive in object tracking, especially in applications such as unmanned aerial vehicles or pods. However, the focus of tracking research has been on generic visual tracking rather than specific scenarios, such as motion blur, which degrades the performance in these scenarios. In this work, we propose an effective method for tracking in motion blur by employing the framework of D3S (a discriminative single shot segmentation tracker). IQA (image quality assessment) and deblurring components are both introduced into the basic D3S framework to enhance context patch, which improves the tracking accuracy in blurred target tracking. Extensive experiments demonstrate that our tracker can robustly track objects, not only in blurred videos but also in other challenging scenes.

Value-Added Bundle(s) Including this Product

More Like This

  • SPS
    Members: Free
    IEEE Members: $11.00
    Non-members: $15.00
  • SPS
    Members: Free
    IEEE Members: $11.00
    Non-members: $15.00
  • SPS
    Members: Free
    IEEE Members: $11.00
    Non-members: $15.00