Research And Implementation Of Motion Target Detecting Algorithm Based On Optical Flow Algorithm

The technology of motion target detection and tracking has been widely used in consumer electronics, industrial inspection, security surveillance and intelligent transportation. The processing speed of target detection and tracking has been a huge gap for system implementation. Nowadays, the research for moving target detection and tracking algorithm is so mature that the improving space is not large. The development method that using C language in PC can’t be satisfied the demands of the performance requirements of processing speed.This paper design a motion target tracking system based on optical flow. Combined with simple and effective detection and tracking algorithm, the processing speed improved greatly.This paper designs moving target detecting and tracking system through constructing hardware platform system, and divides it into module to design and realize. The system obtains real-time image data by digital image acquisition kit. The system uses the SDRAM to realize the image data access. The motion target detection of image series is using optical flow algorithm. With the detection data, the system realizes motion target tracking based on characteristic point. The LEDs display the velocity of motion target and the frame that the system has processed. At last, the outcome of image series is showed on the screen by VGA.This system has the speed of16.9fps for processing true color images which is640×480. It realized the real-time process, and tracking accuracy can reach more than93%. The experiment results show that the system achieves the desired function and has better real-time and reliability.