Method Of The Robot Stereo Vision System Design Based On A New Calibration

The subject of this article from the new robot binocular stereo vision system of academic research projects, thesis to robot for the target recognition and location as the research object, this paper proposes a between the traditional calibration and the method of separation of internal and external parameters calibration, in order to improve the precision of robot localization method.Stereo vision is an important branch of computer vision, has always been the hot spot of the computer vision research, gradually formed its own method and theory.Access to information space is a three-dimensional scene from the most basic of computer vision research content.Stereoscopic vision is the most important in sensing technology computer passive measurement method.This article first introduced the topic background, research significance, the construction of a stereo vision system, this paper proposes a between the traditional calibration and the method of separation of internal and external parameters calibration, use physical method to extract;When external parameters calibration properly selecting the reference coordinate system, through the experimental analysis of the two cameras outside parameter model, and through experiment illustrates the effectiveness of the method. Secondly,, stereo matching method based on features are studied. Finally, combining with the camera calibration results, and the extracted feature points, the use of camera imaging depth of target feature points to restore inverse transformation method, thereby fixing the goal.The experimental results show that this topic using the binocular stereo vision positioning method has higher precision.