Multi-way Video Moving Target Surveillance System Design And Implementation

Intelligent Video Surveillance is employed to obtain the moving target's orientation, recognition and track according to analyse the real-time image automatically by the methods of digital image processing. The core technologies of video surveillance are to detect and perform image information real-time and effectively. This topic presents the design and the realization of multi-channel video surveillance system for moving targets based on Blackfin DSP.According to the practical requirements of engineering, the video detection and tracking algorithm based on static background is firstly discussed in this thesis, the main content includes: denoising the image by mean filter method, achieving image detection and segmentation of the moving targets by using frame-difference and threshold selection methods, analysing the moving target eigenvalue and achieving the goal of tracking movement according to search algorithm order. The multi-channel video moving objects detection system is then analysed and designed, which can be devided into four sub-systems: the control system, responsible for the controlling of user interface, buttons and system communications; video moving target detection and tracking systems, responsible for capturing video image, data processing, the video character superposion and so on; board communication system, responsible for providing power signal and connecting of the internal communications lines; alarm system, responsible for the alarm output. A test for each module of the electric circuits is carried on, some of the problems encountered during the process of commission are analysed and resolved. Finally, the feasibility of the design is verified in the qualitative way, it is proved that the Video Surveillance system is able to meet real-time detection and moving target tracking requirements.