Design And Implementation Of Video Tracking And Encoding System Based On FPGA

The technology of moving target detection and tracking has been widely used in the field of visual surveillance, industrial detection and so on. The processing speed of moving target detection and tracking has been a huge difficulty to system implementation, and the compression and storage of processed video is also very important. This system adopts FPGA as the platform, using its parallel processing mechanism and powerful computing ability to improve system processing speed and performance.Through consulting a great number of relevant references at home and abroad, it is summarized that, in embedded systems, core algorithms of detection and tracking are mostly implemented by software. Such systems cannot meet real-time demands because of huge data volume and computation, and the video after processing cannot be stored in real time and replayed. Therefore, this paper puts forward a method to design and implement system which is based on FPGA. The system is implemented in parallel by hardware to make the processing speed improved greatly and satisfy system’s demand better.This paper designs video tracking and encoding system through constructing hardware platform and software system, and divides it into modules to realize. The video collection module is realized through using D5M camera to collect video images which are converted into RGB format to be cached in SDRAM. Using image preprocessing, morphology operation, and frame difference to get moving target region, and centroid tracking algorithm to track target. The video encoding module is realized through encoding the RGB image data after tracking by MJPEG, generating AVI video flow. Calling SD IP core and using the4wires mode of SD bus mode to store video. Then the real-time detection and tracking result is displayed through VGA. At the same time, NIOS Ⅱ processor is adopted as the core of controlling the modules. At last, the system is verified and implemented on Terasic DE2-115development platform.Throuh testing, it is proved that, this system has the speed of25fps for processing RGB image which is320×240. It has realized the real-time processing, and the tracking accuracy can reach more than90%. Working in20M frequency, the video encoding IP core can use MJPEG to encode image data into AVI video flow which can be played coherently and stored in SD card. The results show that the system achieves the desired function, has better real time and reliability, and portability with a wide range of applications.