Research On Bridge Collision Avoidance Technology Based On Video Monitoring

The increasing number of cross-sea and cross-river bridges and the increasing trend of ships to large-scale development put forward higher requirements for bridge anti-collision technology.As a branch of artificial intelligence,the rapid development of machine vision has greatly improved the degree of automation in all stages of manufacturing.In this paper,the bridge collision avoidance technology is researched based on video monitoring combined with the urgent needs of bridge collision prevention and related technologies of machine vision.The main tasks are as follows:The principle of camera calibration was analyzed in detail.Based on the Hikvision camera,the camera calibration was performed using the Zhang Zhengyou chessboard calibration method to obtain the camera internal and external parameters.Combining the advantages and disadvantages of the inter-frame difference method and the background subtraction method,the three-frame difference method and the mixed Gaussian background difference method comprehensive detection algorithm are used to detect the foreground object.According to the geometric information such as the area and aspect ratio of the circumscribed rectangle of the foreground object,it can detect the sports ship in the bridge area accurately,in real time and completely.A ship yaw detection method based on virtual channel is proposed.According to the triangle similarity principle and the internal and external parameter matrix obtained by camera calibration,the image pixel coordinates of the virtual channel end point are calculated.Using the OpenCV linear rendering function and the Bezier curve,a virtual and curved channel are constructed between the endpoints to detect whether the ship is yawing.The experimental results show that the bridge anti-collision technology based on video monitoring adopted in this paper is feasible and practical,and can accurately detect the ship and whether it is yawing.