Multifunction Digital Video Surveillance System Based On Heterogeneous Dual-Core Embedded Platforms

With the rapid development of the security industry and development trend of monitoring equipment’s miniaturization and portability, the using of digital video surveillance products is more and more widely. Digital monitoring devices with independent processing capability based on the embedded platform occupy a major share of the market due to its advantages on reliability, size and cost. But embedded digital video surveillance products are also faced with many problems as demands of video surveillance growing continuously. On the one hand, embedded product has weaker management function and poorer compatibility than PC-based video surveillance systems. It’s diffcult to upgrade software and new requirements cannot be achieved on the old products. On the other hand, intelligent monitoring systems require more computing resources. But the single-core embedded platform with limited resources has been unable to afford more algorithms. Moreover, the difficulty of adding new algorithms in embedded platforms also restricts the development of embedded intelligent monitoring devices. In such a case, hardware manufacturers have introduced heterogeneous dual-core hardware architecture and supporting software development kit, trying to solve these problems. Therefore, intelligent monitoring platform based on heterogeneous dual-core hardware architecture has broad market prospects.In this paper, an intelligent monitoring platform based on TMS320DM8168Davinci digital media processor has been studied to finely provide a management system of embedded platform. The main results which this paper has achieved include:1. This paper designs and implements a developing and debugging system based on TMS320DM8168Davinci processor. Features implemented include reading and decoding the stream file locally, calling algorithms to process, displaying or storing the results generated by algorithms, which enable developers to debug algorithms easily on this platform.2. This paper designs an algorithm interface based on McFW architecture which enable algorithms to be switched dynamicly and added easily in the ALG LINK.3. This paper designs and implements an embedded digital video surveillance management system based on heterogeneous dual-core hardware. The system can be administrated effectively via Web and load modules intelligently.