Research On The Aerial Moving Target Tracking Technology Of Multi Eye Vision On Ship Carrier

With the development of machine vision, machine vision has been more and more widely used in civil and military fields. In banks, shopping malls, hospitals,schools, traffic junctions and other places, video surveillance equipment that can be seen everywhere guarantee the normal operation of people’s lives; in the military,machine vision can be used to monitor and track the situation of the enemy. As an important application of machine vision, moving object tracking technology has been developed rapidly in recent years, and it has a higher requirement for the accuracy and real-time of tracking.In order to ensure the safety of the ship, there is a need for a full range of monitoring around the ship. Based on this background, this paper focuses on the background of the moving target tracking technology, and consider the impact of the actual environmental impact on the tracking system. A solution is proposed for the influence of ocean waves and the influence of light.A lot of moving target detection and tracking method, for the specific target tracking can usually be with the target texture, color, shape, and other information as the basis for tracking. In the video frame image and template matching, to track the moving target. If the unknown target information is unknown, according to the characteristics of the motion to track the target, this paper mainly studies the moving object tracking technology in this case. The research content of the subject is divided into four parts:(1) According to the background of the subject, this paper analyzes the actual problems in the process of detection and tracking, and puts forward the corresponding solutions. A multi camera vision measurement system is designed to meet the requirements of all azimuth tracking. In view of the influence of ocean wave on the image, this paper proposes a method to eliminate the blur.(2) This paper compared the Gauss background model, the code of this model,LBP model and other different background modeling methods of defects. According to the possible changes of light, this paper improve the background modeling algorithm.(3) Through binocular vision can reconstruct the Three-dimensional,programming to achieve the measurement of the target distance and azimuth angle.(4)Based on the error analysis of the measured data, analyze the data redundancy of the multiple sets of binocular vision measurement system, and the method of neural network is proposed to reduce the error of the system.