Research On The Segmentation Of Moving Object Detecting

Intelligent Video Surveillance System (IVSS) is a frontier topic in Computer Vision domain. It spans many subjects including computer science,machine vision,image engineering,pattern analysis,artificial intelligence,etc.Moving object detection is the key to intelligent video surveillance system technology, besides the results of detection have a great effect in the follow-up treatment such as tracing. Due to its characteristic that moving objects can be detected independently in the situation of the existence of the moving targets by Intelligent Video Surveillance System, moving objects detection technology has been widely used in all aspects of social life. In recent yeas, it has attended by scholars from various countries.As the background scene and the state of the environment such as weather, light, shadow and confusion interference often changes, the difficulties of moving object detection are increased. Therefore, the research of the algorithm of the moving objects detection for objects detection system is to enhance the adaptability of the dynamic, to ensure that the moving objections can be detected accurately for following processing treatment in the case of the outside interference.Based on familiar materials, the motion detection technology and current research are introduced at first. Then, a motion detection and extraction process has been designed and implemented completely on the analysis of background modeling. For overcoming the setback of processing on a single pixel, an approach of the image segmentation has been presented, which is divided into two steps: first, the coarse segmentation based on image regions; second, the fine regions based on the pixel. While combining the parameters of the Gaussian with the characteristics of the average energy of image regions, the optimal threshold value of difference image can be obtained dynamically. Meanwhile, the use of the result of whether the image region matches the current Gaussian model updates the background image dynamically. The experiments show that the motion detection algorithm can detect moving objects correctly, and improve the detecting effect of real-time motion detection system.