Correlation Filter Tracking Algorithm Based on Spatio-Temporal Context

Abstract

Object tracking has been widely used in artificial intelligence, military reconnaissance, security monitoring and other fields. It has become a research hotspot of computer vision. To handle the drift problem in the presence of occlusions, a tracker combined with spatio-temporal context information and correlation filter is proposed in this paper. HOG (Histogram of Oriented Gradient), CN (Color Name) and gray features are extracted to learn the correlation filter. Meanwhile, the spatio-temporal context model is trained. The response map of correlation filter and spatio-temporal context model are normalized and fused. Experimental results show that the proposed algorithm can accurately track the object, and has better performance in terms of successful rate, center position error and distance precision.