Adaptive Keypoints Grouping for Robust Visual Tracking

Lao Mingjie, Yazhe Tang, AngZong Yao Kevin, Lin Feng

Peer Reviewed

Abstract: This paper introduces a keypoints based adaptive grouping algorithm for long term visual tracking. The algorithm employs a collaborative scheme with keypoints matching and estimation for the purpose of global scope tracking. The motion flow of keypoint is measured in optical flow and points with different motion are marked as the outliers. The keypoints group can adaptively change based on motion consensual clustering. The class changing may improve the flexibility of system because the keypoints may change it group attribute adaptively based on its motion similarity to the other group. Finally, a series of experiments is presented to verify the performance of proposed algorithm on the public benchmark.
Published in: Proceedings of the ION 2017 Pacific PNT Meeting
May 1 - 4, 2017
Marriott Waikiki Beach Resort & Spa
Honolulu, Hawaii
Pages: 155 - 161
Cite this article: Mingjie, Lao, Tang, Yazhe, Kevin, AngZong Yao, Feng, Lin, "Adaptive Keypoints Grouping for Robust Visual Tracking," Proceedings of the ION 2017 Pacific PNT Meeting, Honolulu, Hawaii, May 2017, pp. 155-161.
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In