A fast approach for person detection and tracking. (13th September 2012)

• Record Type: Journal Article
• Authors: Ali, Ahmed; Terada, Kenji
• Is Part Of: International journal of computer applications technology. Volume 44:Number 3(2012)
• Page Start: 210
• Page End: 216
• Publication Date: 2012-09-13
• Subjects: Kalman filter -- mean shift algorithm -- object tracking -- object detection -- video processing -- surveillance systems -- target tracking -- detecting people -- tracking people; Technology -- Data processing -- Periodicals; 620.00285
• Languages: English
• ISSNs: 0952-8091

Abstract

Object detection and tracking is one of the most popular areas of video processing and the essential requirement of any surveillance system. A fast approach for person detection and tracking is presented. This work proposes to do target tracking with Kalman filter. In case of a misdetection, which would lead a wrong update of the filter, a fast mean shift iteration based on integral computation is performed to propose a more accurate detection. Tracking results are demonstrated for complex scenes and evaluation of the proposed tracking approach is presented.