Feature Points Based Video Object Tracking for Dynamic Scenes and Its FPGA System Prototyping
In Video based Surveillance Systems, moving object tracking is the key step. In this paper a novel approach for tracking moving object from a video sequence based on edge detection and Feature matching is presented. In the canny edge detector filter the video sequence is sub-divided into its constituent frames. This method proceeds by considering two consecutive frames. Canny’s operator is used to detect the edges present in respective frames, which helps to estimate the object boundaries. Frame difference of object boundaries is performed, to eliminate most of the edges belonging to background. To emphasize the regions containing the moving object, Feature extraction technique is performed on the edge difference image. At last using Feature matching method track the exact moving object.