Research & Implementation Of Vehicle Backover Obstacle Detection Algorithm Based On Vision

With the growth of world economy and the decrease of automobiles'price, the quantity of private vehicles increases quickly in recent years. While enjoying the convenience, the traffic accident problems become more prominent day by day. The accidents caused by driver at the parking time have reached approximately 30%, because of the blind spot behind the vehicle. Therefore, the parking assistant technology which has backover obstacle detection function receives much attention. It has become the development trend of intelligent parking technology at present.After analyzing and comparing lots of obstacle detection algorithms, the thesis has proposed a robust obstacle detection algorithm based on estimation of ego-motion. Firstly, the algorithm based on ego-motion model, through traversal parameter of movement search space, searching the maximum motion parameter denoted ego-motion probability between the corresponding road surface regions which in continual two images, thus estimates the ego-motion of the vehicle related to the road accurately. Secondly, the thesis get the differential image from sequence images based on the motion compensation using the movement difference between obstacle and road surface regions, and implements the obstacle detection through removing the noise spot and withdrawing the obstacle region in differential images. Finally, the thesis has carried out the design and high-performance implementation in the view of the special embedded platform form overall performance angle.The proposed algorithm in the thesis has been implemented in Recognition Engine which belongs to Neusoft Advanced Automotive Electronic Research Center. The method has been tested in real world environments and has achieved expected results in recognition rate and real-time fields.