Study On Bridge Anticollision System Based On Video Sequence Images Processing

With the national economy development, more and more bridges are built upon the rivers. Now there are more than 20 bridges upon the Yangtze River. Take the Yangtze River as an example, since 1957, it had been collide with the ships 69times, including 68 times into the pier. The bridge is damaged badly, and it is imminent to keep it from being collided with the ships again.With the development of the video surveillance, many researchers turn to work on it achieved some positive results. It is widely used in ITS. But in the area of the pier monitoring, only some manual measures are being used. These measures can mitigate the damage of the bridge pier in a certain extent, but can not resolve the conflicts between the bridge and the ships thoroughly.In the article, an approach is proposed to solve the problem. In this paper, it is focused on two technologies crucial to the system, namely, the ship detection and the ship tracking. The main point of the article is:1. Analyses the two methods for the target detection: The background subtraction and the inter-frame difference. Discusses the difference between the methods, and proposes a method for the target detection;2. Discusses the algorithm of the corner detection-SUSAN corner detection algorithm, and uses this algorithm to locate the target. The experiment shows it is correct and feasible.In the end,this paper put forward a personal view on the further development of this system.