Researeh On The Key Technologyo Of Video Surveillance System Based On Embedded Linux

With the advancement of science and technology, people’s quality of life alsomade increasing demands. The comfortable, affordable, convenient and safe homeenvironment is gradually being favored by the public. As an important part of thehome security system, video surveillance is gradually getting people’s attention. Theembedded video monitoring system due to small size, powerful, stable performanceand low cost advantages make home security and other applications in place. Thingsintelligent home security concepts proposed will drive the use of embedded videosurveillance in the home security system. Because of the open-source Linux systemwith a streamlined core and excellent performance, therefore, in the embeddedoperating system, Linux operating system is gaining more and more attention.According to the performance requirements of enhance home security system，todesign a video surveillance system. By analyzing the architecture of ARMmicroprocessors and the Linux operating system works, given Linux2.6operatingsystem and application programs on the ARM9microprocessor transplant after cutting,to achieve intelligent home network video surveillance system purposes. According tothe various needs of home security applications, the paper added video surveillanceprocess moving target face detection section to detect abnormal. By building a Webserver in the embedded Linux systems can use your computer browser way to accessreal-time video image data. In this paper, how to improve the efficiency of imageacquisition and household scenes of moving target detection are two basic problemspeople face are discussed and studied.Firstly, the current video surveillance technology is moving in the intelligentdirection, mainly refers to the increased recognition or face detection, imageprocessing part. It is because of real-time video surveillance system put forwardhigher requirements, so, how to improve the collection efficiency of the image is thefirst research question of this article. This paper turns double URB way to improvecommunication efficiency image capture, in turn achieve URB communicationprocess needs to be studied and improved V4L agreement.Secondly, there are a lot of image processing technologies, but no single methodcan be applied to all scenarios, and to find a suitable way to not be able to fully apply.This need for a particular scene, select the appropriate method and further improvements. This study is to detect the scene in a private home moving targetdetection and face detection technology to achieve the image moving human face.Which the moving target detection using background subtraction and select improvedGaussian mixture background modeling to meet the testing needs of home backgroundin complex environments face a moving target; while face detection method is the useof color matching.