Research On Moving Object Detection Algorithm Based On Computer Vision

Machine vision is a new discipline, since its initial research, has experienced more than half a century of development. Its mature develops together with the computer technology. Especially, the ability of computer data processing is improved greatly, which has made great progress for the machine vision.Machine vision is closely linked with cognitive science, statistics and applied mathematics, computer engineering, physics, signal processing and neurophysiology and many other key disciplines, so that the practical application of machine vision technology from theoretical research has went further, and now the machine vision technology plays a very important role in image compression and processing, visual reconstruction, intelligent video surveillance, automatic driving and unmanned driving vehicles, safety monitoring, industrial automation, medical operation, satellite measurement and control fields, and it has a very broad application prospects.The main research contents in the machine vision include the calibration of the motion target, the detecting and tracking of the motion target, etc., the work starts with the 2D image information based on visual sensor, and obtains the required information finally. This paper mainly has researched and implemented the detecting and tracking algorithms of the motion target based on vision. In contrast with many kinds of commonly motion target detection algorithms, this paper proposed a kind of spatial edge orientation histograms(SEOH) algorithm for motion target detection. Aimed at the light sensitive, changing with the natural environment, detecting difficultly with the shadow effect, the algorithm protrude the spatial edge orientation information, making up the defects of the conventional algorithm. At the same time, in order to improve the speed and the robustness of the algorithm, this paper combined SEOH and Kalman filter, and further to modify the predicting motion parameters. Experiments showed that, this algorithm improves the anti interference in the light and environmental changes.