Study On Position Control System Based On Computer Vision

Specific to the current status of industrial production,this paper put foward theory using video surveillance system instead of position sensor in the traditional industrial production, in order to attain the information of interesting target quickly and easily and simplify the complexity of the industrial production. The article designed an embedded control system based on computer vision with simple background system to examine the effects of theory in actual industrial production.The main research method is:analysis of the current main target detecting and tracking algorithm, select background subtraction method as the main method of target detection;then choose the kalman filter theory in bias as the target tracking method.On these basis,detecting and tracking a few targets under simple background is achieved.Then information is transferred to the motor control system to control the motor using traditional method PID,and the speed of motor will change the position of the target.Finally,all the subsystem make up a simply closed loop system.By analyzing the results,the algorithm designed in this article can detect and track the moving target under simple background accurately,and instead of position sensor in this case.