Title :
Research and implementation of distributed intelligent video analysis system based on DSP
Author :
Liang Chengwu ; Wang Hui ; Guan Qing ; Xu Sheng
Author_Institution :
Henan Univ. of Urban Constr., Pingdingshan, China
Abstract :
This paper aims to realize an distributed intelligent video analysis system based on TI multimedia Digital Signal Processor (DSP) TMS320DM642. The system we proposed provides users with fast, precise and smart video analysis services. This system collects real time field data via analog cameras and IP cameras based on Network or Wireless System. The image data then is transmitted to the embedded high speed DSP real-time image processing board, where target, including person, object and vehicle, is extracted from background and compared with predefined rules. Users can freely set virtual alarm line, alarm region and violation behaviors through rule setting interface. DSP analyze video images timely, and detect violations. Target segmentation, feature extraction, alarm and automatic tracking of multi-channel and multi-camera are implemented at the same time in this system. This system can be widely used in many fields, such as security monitoring, and has great potential in real applications with very high commercial values.
Keywords :
digital signal processing chips; multimedia systems; video signal processing; DSP real-time image processing board; IP cameras; analog camera; automatic tracking; distributed intelligent video analysis system; feature extraction; image data; multi-camera; multi-channel; multimedia digital signal processor TMS320DM642; precise video analysis service; rule setting interface; security monitoring; smart video analysis service; target segmentation; violation behavior; virtual alarm line; wireless system; Cameras; Data mining; Digital signal processing; Digital signal processors; Image processing; Intelligent systems; Multimedia systems; Real time systems; Signal analysis; Vehicles; DSP Implementation; Feature Extraction; Intelligent Video Analysis; Target Detection; Target Segmentation; Target Tracking;
Conference_Titel :
Mechanic Automation and Control Engineering (MACE), 2010 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-7737-1
DOI :
10.1109/MACE.2010.5535620