Application of exception recognition and fuzzy decision in surveillance

The method of fuzzy decision is proposed according to specific application environment after extracting characteristic parameters effectively. The area and moving distance of the moving object are used to judge the recognition and the alarm of the objective. Abnormal phenomena are caught and stored. At last, alarm signal will be sent. The research shows that our novel method is very effective and efficient in making the surveillance system eliminating the region noise caused by fluorescent lamp, local moving shade and dithering in dynamic image sequences.