Welcome to Acta Armamentarii ! Today is Share:

Acta Armamentarii ›› 2017, Vol. 38 ›› Issue (6): 1140-1146.doi: 10.3969/j.issn.1000-1093.2017.06.013

• Paper • Previous Articles     Next Articles

Multi-feature Compressive Sensing Target Tracking Algorithm Based on Redundant Dictionary

ZHU Su1, 2, BO Yu-ming1, HE Liang1   

  1. (1.School of Automation, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China;2.School of Electronic Information and Optoelectronic Technology, Zijin College, Nanjing University of Science and Technology, Nanjing 210046, Jiangsu, China)
  • Received:2016-10-24 Revised:2016-10-24 Online:2017-12-15

Abstract: In consideration that the basis of signal sparse representation is an orthogonal matrix in the multi-feature compressed sensing algorithm, the multi-features of infrared and visible images are extracted to construct a sparse representation in a subspace of redundant dictionary, and the selection of sensing matrix and the reconstruction of sparse signal in the algorithm are analyzed. A redundant dictionary-based target tracking algorithm of multi-feature compressed sensing in the framework of particle filter is proposed by reconstructing the signal sparse representation, which can automatically detect dynamic targets in complex environment. Experimental results show that, compared with other classical algorithms, the proposed algorithm has better robustness and real-time in complex environment like illumination change and interference object occlusion. Key

Key words: informationprocessingtechnology, redundantdictionary, compressivesensing, particlefilter, targettracking

CLC Number: