An improved fisher discriminant dictionary learning for video object tracking

  • Authors:
  • Ji Zhang;Hong-yuan Wang;Fu-hua Chen

  • Affiliations:
  • School of Information Science and Engineering, ChangZhou University, ChangZhou, China,ChangZhou Key Laboratory for Process Perception and Interconnected Technology, ChangZhou, China;School of Information Science and Engineering, ChangZhou University, ChangZhou, China,ChangZhou Key Laboratory for Process Perception and Interconnected Technology, ChangZhou, China;Department of Natural Science and Mathematics, West Liberty University, West Virginia, United States

  • Venue:
  • IScIDE'12 Proceedings of the third Sino-foreign-interchange conference on Intelligent Science and Intelligent Data Engineering
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Video object tracking plays an important role in modern computer vision, and many algorithms have been proposed in recent years. ℓ1-tracker, a novel generative tracking method based on sparse coding, has demonstrated very promising performance in numerous challenging sequences. But the high computational cost, which is caused by the large size of dictionary, influences its application in tracking severely. In this paper, based on original Fisher discriminant dictionary learning(FDDL) and our improved version, we present a novel tracking algorithm, called FD2LT. In our framework, tracking is considered as a problem consisting of three components, including object location, training samples selection, and dictionary updating. Experimental results demonstrate the effectiveness of the proposed tracking algorithm.