New Video Target Tracking Algorithm Based on KNN

Journal of Multimedia

关键词: KNN;    Target Tracking;    Video;   

Others  :  1017222
DOI  :  10.4304/jmm.9.5.709-714

【 摘 要 】

Compressed sensing is an advanced technique used in visual tracking problem. However, current research uses too many pre-knowledge, and cannot deal with quick object sensitively. Based on the problems, the paper proposes a new real-time visual tracking algorithm based on compressive sensing. On one hand, the algorithm uses the characteristics of compressive sensing, and make sure of the time constraint of the tracking. On the other hand, KNN classifier is used to differentiate the target and the background. Results show that the algorithm can provide the accuracy of the real-time tracking.

【 授权许可】

   
@ 2006-2014 by ACADEMY PUBLISHER – All rights reserved.