Intelligent Video Surveillance System Based On DSP

In recent years, with the continuous improvement of the people’s living standards as well as the continuous progress of science and technology, security issues have already become one of the hottest issues. The video surveillance system appears as an important security measures in our lives, video surveillance system tend to be digital and intelligent which has been promoted by electronic technology, computer technology as well as video processing technology. Intelligent video surveillance system is not only used in the field of security, but also used in transportation, military, finance, industry and other fields, which has a good application and great economic values.This thesis has designed a intelligent video surveillance system which based on a multimedia video processing chip TMS320DM642. The system mainly combines image processing technology, computer vision technology and network communication technology. Firstly described the background of the research, and state the research status of video surveillance systems all over the world. Then we will build a hardware platform structures which is based on DM642as well as the software program design. On the other hand, the thesis do a lot of research about moving target detection algorithm, and describe the traditional three moving target detection algorithms, and give a new method which is based on these three algorithm, Gaussian mixture model background method.It designs the overall framework of the system, each module of the hardware and the software of DSP, and an efficient moving target detection algorithm is applied to the system in this thesis. The program design of DSP is based on TI’s integrated development environment CCStudio. The system uses an embedded operating system DSP/BIOS, controlling each of thread schedules, using TI frame of reference RF5and network development kit NDK to program. It communicates between client and server using network transport protocol TCP/IP, the monitoring screen of the real-time system could work stably and smoothly. It achieved the desired goal. In addition, it describes the analysis of the deficiencies of the system, and points out that the subsequent optimization.