| The speed of image processing frequently touches the tracking performance in imaging tracking system. In this paper, one method is proposed to improve the speed of image processing, i.e. image parallel processing system based on DSP and FPGA. This image processing system has faster real time image processing capacity and is able to provide the target's position steadily for tracking system.DSP is choosed as central processing unit and FPGA is choosed as coprocessor in the image processing system. So it takes advantage of the two components' superiorities. Image processing algorithms used by the system are checked up and compared, then appropriate algorithms are adopted by the system. Median filter algorithm is used in image pretreatment; threshold segmentation algorithm is used in image segmentation; gate tracking algorithm is used in calculating target's position.As the speed of image processing is improved, tracking performance of imaging tracking system is also improved. and also some fine algorithms may be used in image processing system to enhance recognition capacity . |
