Target Tracking Control System Based On The Ccd Image Sensor

With the introduction of images, audio, video and other media information,the wireless multimedia sensor becomes a new research area of computerscience and technology, the achievements of which has been widely applied inmilitary, environmental, space exploration, health, transportation and otherkinds of field. Tracking objects is a main area of wireless multimedia sensor. Byright of the application of CCD video sensor, the tracked target would becomemore lively and intuitive, and the trace is more accurate.Intelligent video tracking system through the camera of image toautomatically identify and locate the moving targets without human intervention,and control the camera move in accordance with the target acts to real-timetarget tracking. The moving target detection, trajectory prediction of movingtarget detection, and PTZ(Pan Tilt Zoom) control algorithm of video trackingtechnology were mainly studied.In view of existing shortcoming of the camera and the image detection, therequirements of accuracy and rapidity, found in the several commonly useddetection methods, in the detection of the target image, this paper put forwardthree frame differential methods based on motion compensation. In this way, themoving targets with dynamic background can be detected quickly andaccurately.The improved particle filtering algorithm has been used in this paper forthe target trajectory prediction. To improve the robustness, timeliness andaccuracy of tracking, this paper use the particle swarm optimization filteralgorithm to predict the target trajectory, proposed reference model of targetwhich combine the color histogram with the contour histogram, to solve particledegradation of the particle filter. The simulation results show that the improvedparticle filter algorithm is much better than the traditional particle filteralgorithm for the effect of predict.In the control system, the DC servo motor control the PTZ, and a mathematical model of the servo system is established. The nonlinearity,interference and uncertainty of motion objects of video tracking system, thispaper analysis the limitation of the conventional PID controller,and thenestablish fuzzy PID controller and the IMC-PID controller for video trackingsystem. The simulation results show that both of the fuzzy PID controller andthe IMC-PID controller can satisfy the requirements of dynamic characteristicsand robustness, of which the IMC-PID controller with only one adjustableparameter is more available.