The Design And Implementation Of The Intelligent Video Surveillance Front-End Based On DM6467

In recent years, by the improvement of computer calculation capacity and the development of the network technology as well as higher real-life demands of the society, video surveillance technology and application are undergoing rapid development. Visual information occupies large proportions in all fields of human social activities information, therefore, video surveillance technology is becoming an important defense and security method.Intelligent video surveillance technology is a new artificial intelligent security technology which involves computer science, machine vision, image processing, pattern recognition, artificial intelligence and various other disciplines. It can automatically analyze image sequences captured by video cameras without human interference, then understand the content and give explanation of the objective scene. Hence, it is becoming a hot topic in the field of video surveillance technology.In this thesis, ARM+DSP dual core processor dm6467 is chosen as the central processor for the test system, embedded Linux is used as the operating system. Based on this system, the intelligent video surveillance front-end is designed and implemented. the intelligent front-end not only is capable of traditional tasks such as video capture, display and storage, but also performs intelligent detection of the video and adjust the coding parameters adaptively. In addition, human-machine interaction is designed on the system.Firstly, according to the "Technical specification of real time video intelligent analyzing device in surveillance system", the overall design scheme and the main function of each module are discussed and given:video capture, analysis, coding, storage, network transmission, display and human-machine interface. The function of the video capture and display module is to capture and display PAL video in the resolution of 720×576. The main task of video analysis and coding module is analyzing the original video data and performing subsequent encoding and compression. Based on real-time motion detection and perimeter intrusion detection, the encoding bitrate and the image resolution are increased automatically when the alarming event happens. The video storage module stores the compressed video on local SATA hard disk. The network transmission module transmits the encoded stream to the remote PC, so that the monitor video can be displayed on the remote PC in real time. The human-machine interface is mainly used as the setting of the structure and function of intelligent video monitoring front-end. The interface has the function of interface switching, board and video display parameters, encoding and decoding parameter, alarm information and video storage etc.Experimental results show that the developed intelligent video surveillance front-end can meet all the design requirements, and the standards of real-time intelligent surveillance.Finally, each module of the intelligent video surveillance front-end is downloaded into the system,and linked to a remote PC via Ethernet to form a complete intelligent video surveillance system.