OpenCV Implementation Of Moving Object Detection And Tracking In Video

Development of computer vision technique involves many disciplines, is quite extensive and optics, signal processing, artificial intelligence, pattern recognition, image processing, and computer technology. As signal processing theory, calculation theory, Visual theory, and human visual mechanism, related technology of development, computer vision also has development fast, and its non-contact, automation, visual, real-time and intelligence, features widely are used for industrial production and medical health, aviation aerospace, scientific research and the defense construction, many national and industry pay attention to it.Based on video analysis and tracking of moving objects in the image is one of the most complex computer vision research, Because of the plenty of video, slow processing speed, at the same time meet the requirements of time, have to make efficient image processing algorithms. Study on algorithm for researchers, if every image processing function to write from scratch yourself, it is difficult to ensure commonality of function and cause a waste of time. Select Intel's OpenCV image processing library for secondary development platform, OpenCV is open source projects and is a computer vision and pattern recognition, which can be used in many fields. It has the following features:open source; Based on Intel processor instruction set optimization code; Defines the structure and function of the uniform; Appropriated images and matrix calculations; Flexible and convenient user interface; Support for multiple platforms.The main research contents of this article as follows:1. For image noise removal method for comparison, summarizes the different filter image processing algorithm is particularly useful in that regard.2, the image of leveling and the realization of binary code, proposing a binarization algorithm for gray-scale image, realized on the image characteristic analysis and identification.3.This paper has analysis research on the movement target detection and track algorithm on video, for detection process in the video monitoring background very easy was interference of situation, creation of the Gaussian model is used to detect the foreground image. 4. OpenCV vision library and VS2005write applications, which has a secondary development with OpenCV. the realization of the detection and tracking of moving objects.5. System design of a framework for target detection and tracking, and on target detection and tracking using OpenCV implementation.Figure28reference60...