Design Of Indoor Intelligent Monitoring System Base On Davinci Platform

The video monitoring system has experienced three development stages during the past20years, from the first simulation system to the partial digitization system, and reached the third full digitization system now.With the development of the internet, computer and sensor technology and the emergence of the abundant image processing algorithms, intelligent monitoring attracts more and more attention and turns to be the most advanced application model in the area of monitoring, which is mainly reflected in the thorough application of the intelligent video processing technology in the video monitoring system.Considering the problems exist in the current video monitoring system, such as the poor warning accuracy, the false and failure warning phenomenon and the difficulties in video data analyzing, this thesis designed and realized a set of indoor smart monitoring system with the dual-core processor TMS320DM6446from TI Company. The main work included setting up the efficient software frame and designing and making peripheral circuit board. The system built in the thesis could detect movement objects and its basic external form in high efficiency and accuracy according to the indoor environment. It also makes judgments and emits warnings when the abnormal behaviors such as falling down happen and could resist the interference from the changing illumination to the detect results effectively at the same time.The main contributions and innovations are as below:1. Setting up a dual-core embedded system development platform which is based on the Davinci framework. The ARM operates the MontaVista Linux system, achieving the functions of collecting, processing, revealing and controlling of the video information by using multithreading technology. The DSP operates DSP/BIOS system, taking charge of the realizing of the image processing algorithm. 2. Designing and making the peripheral circuit board according to the core board from Tairui Company and the related demands of the system, which reduces the cost of the production effectively.3. Realizing the detecting of the moving objects:on the basis of background subtraction, improving it by utilizing morphological image processing method and image connectivity analysis, so as to achieve the fast and accurate detecting of the indoor moving objects and resist the interference from the changing illumination effectively at the same time.4. Achieving the judgments of abnormal behaviors:judging if the abnormal behaviors such as falling down exist in the moving objects and start the accordingly warning mechanism afterwards by the analysis of the external form of the moving objects.