The Implementation Of Image Detection And Target Tracking System Based On DSP

Safety of our life and asset have been paid more and more attention as the times go by, and human resource is on the trend of high cost, thus study of real-time surveillance system becomes one of the important issues. But, modern surveillance system can only provide limited function. Except simple surveillance system, we still hope that we can detect object under complicated background, thus we proposed detection of moving object on indoor plaza and advanced automatic tracking system.It combines the TI TMS320DM642 EVM with CCS software system to be the research developing platform. At first it use EDMA controller to create DMA gateway, and directly store image materials in capture buffer to wait for process. First, it builds the background model by observing the scene without people in it. When the detecting process starts, an adaptive background subtraction method is utilized to segment the moving regions and locate the positions of human bodies. Thus, multiple human bodies can be tracked. In real-time motion tracking, it get moving objects block by Background Image Subtraction Algorithm, and it build every moving object Color Histogram Model. When moving objects don't cross, it use Correlation Based Block Matching Method; on the other hand, when they cross, it use Color Histogram Model Matching to complete our tracking system. Finally it exports processed images on the screen by EDMA controller. Experimental results show that the proposed system performs well in indoor, complex environments.