Tracking And Alignment System Based On Machine Vision

When the motion target is scanned and tracked,a single fixed camera is used to collect the panorama of the target motion.The camera's field of view is too large to cause the low resolution of the camera so that the target cannot be located high precision,but when the camera is used to track the target to obtain the target,the field of view of the camera can obviously be reduced very much.Many,so as to solve the problem of its low resolution.This requires the system to ensure real-time and accuracy in design,so as to achieve the function of tracking and alignment.This paper mainly designs a tracking and alignment system based on DSP control,CMOS industrial camera image feedback,which mainly solves the problem of the practical application of the system.This system combines the image processing ability of PC and the advantage of motor control of DSP,and realizes the function of tracking alignment through RS232 communication,and the tracking error is within 5 pixels.In view of the control system,starting with the traditional PID control theory,the motor mathematical model of this system is created,and the simulation is carried out by Simulink to adjust the PID parameters to the satisfactory effect.Then,the PID based on sliding mode speed controller and PID control based on single neuron are discussed respectively.It can be obtained from the simulation results that the PID control precision based on sliding mode speed controller can be higher.The PID control based on single neuron requires more time to learn.On the basis of the oretical analysis,TMS320F28335 is used as the control core of hardware implementation,and the feedback information of image denoising and centroid extraction from PC is used to calibrate the camera parameters by the calibration method of Zhang Zheng friends camera,thus the functions of positioning,tracking and alignment are realized.In this paper,the function of tracking alignment is designed and implemented,and the experimental data and system error analysis are given.