Study of Moving Object Detecting and Tracking Algorithm for Video Surveillance System
Tao Wang,Rongfu Zhang
DOI: https://doi.org/10.1117/12.865457
2010-01-01
Abstract:This paper describes a specific process of moving target detecting and tracking in the video surveillance. Obtain high-quality background is the key to achieving differential target detecting in the video surveillance. The paper is based on a block segmentation method to build clear background, and using the method of background difference to detecing moving target, after a series of treatment we can be extracted the more comprehensive object from original image, then using the smallest bounding rectangle to locate the object. In the video surveillance system, the delay of camera and other reasons lead to tracking lag, the model of Kalman filter based on template matching was proposed, using deduced and estimated capacity of Kalman, the center of smallest bounding rectangle for predictive value, predicted the position in the next moment may appare, followed by template matching in the region as the center of this position, by calculate the cross-correlation similarity of current image and reference image, can determine the best matching center. As narrowed the scope of searching, thereby reduced the searching time, so there be achieve fast-tracking