Intelligent Monitoring System Based On Davinci

As the fast development of information technology and the rapid spread of the internet, the video surveillance system is striding forward to networking and smartness. Traditional analogous video surveillance systems and digital video surveillance system is gradually replaced by smart monitoring system. Based on the dual-core architecture of Davinci DM6446 processor, a smart networking video surveillance system framework, including hardware and software, is proposed in this dissertation.To exploit the resources of the dual-core structure of DM6446, multiple applications, including video recording and playback, user interface control, video signals transmitted over the network, the external device control and so on, are running in the ARM-side, and those high computational complexity routines, such as AdaBoost-based face detection, Mean-Shift based face tracking, and H.264 video codec, are executed in DSP-side. A DirectShow technology based client program is also developed to enable the remote surveillance system control.In the DSP subsystem software design, the algorithm generation, xDM standardization, algorithm optimization, algorithm packaging, executable code generation for Codec Server, etc, are discussed in detail. To make good use of limited resources, a new resources allocation strategy is proposed, and the resource utilization rate is significantly increased. Multi-threaded programming is used in ARM subsystem designed to maximally exploit its system control ability. Based on DirectShow technology network player, the client side can not only transmit monitoring video via internet, but also control camera PAN/TILT and LANS, effictively expanding the scope of surveillance. Consequently, the networking and smartness of the proposed video survillance system is achieved.