Font Size: a A A

The Research Of Intelligent Technologies In Video Monitoring System Based On ARM And DSP

Posted on:2012-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:P P ZhangFull Text:PDF
GTID:2178330335491171Subject:Physical Electronics
Abstract/Summary:PDF Full Text Request
This paper studied some practical technologies of intelligent video surveillance system based on embedded core-ARM and DSP. Designed a real-time video monitoring system using ARM and DSP,with simple equipment,low power consumption,stable performance and low cost.This embedded platform could achieve some operations of intelligent video surveillance system.In the hardware design,S3C2410,ARM microprocessor was used as the main controler, and running Linux operating system to control multiple real-time task such as video signal collection,data transmission and image processing; Information of the video image sequence was processed by the DM642 microprocessor,then,after calculation, DM642 output the location and size of the interest or target regions in the current processing image, and the results are fed back to the ARM,to adjust the controls of video signal collection and data transmission operations and so on.The moving target detection and tracking algorithm is chiefly researched in this article. Firstly,it introduced the image preprocessing techniques, including image filtering, image enhancement processing; Analysis the frequently-used methods in moving target detection and tracking,and their detection and tracking elements.This paper detailedly studied the theory of CAMshift algorithms,discusses the strong points and weak points of this algorithm used in moving target tracking application. The classical CAMshift algorithm is low intelligentize and when the moving target was largely sheltered, CAMshift can not track the target effectively.So this paper researched one approach to improve classical CAMshift algorithm whitch combin with KIM difference algorithm and Kalman filter. That new CAMshift algorithm's tracking ability has been enhanced and it's using scope has been enlarged too.Finally, this paper gave some simulation experiments and showed the experimental result.The result demonstrate that new CAMshift algorithm is high real-time,more robust and better performance,and will be batter used in embedded video surveillance system,effectively achieve the detection and tracking for moving object in a simple scene.
Keywords/Search Tags:Intelligent video surveillance, ARM, DSP, CAMshift, KIM difference algorithm, Kalman filter
PDF Full Text Request
Related items