It is needed to process the image real-time in intelligent active security system that based on computer vision and model identify. The system based on PC is not real-time, expensive and can't work smoothly in bad environment. Aimed to the default, this thesis uses the TMS320C6201 to process the image, and development the intelligent active security system.It is first thing to select DSP chip while design the hardware system. This thesis studies the DSP history, condition and using. TMS320C6201 is selected. This chip has great performance and agrees with processing image real-time.The hardware picture of whole system and image collect system is designed. The selection of DSP chip and DSP software must be valued in EVM board, so TMS320C6201 EVM board has been introduced, and the hardware system based on EVM is designed.There some function that intelligent active security system will perform: detect road, detect obstruction and detect the distance. The algorithm based on PC and VC is displayed. The CCS and initiation of DSP code and memory allocation has been studied deeply. The question that VC code translated to CCS has been discuss and got the good advice. Compare the result, study the method how to test the algorithm based on PC and EVM. Hardware the excellent algorithm will promise the real-time. We research the method how to improve the algorithm.With the action image process theory and well-rounded DSP technology, with the research and apply, we designed the high quality, low price, high reliability DSP software and hardware system that can collect and process action image real-time.Based on commonly procedure, with intelligent active security system, computer vision and action image process, we summarized a course to develop and test the DSP hardware and software system. |