Study On Binocular Vision System In Robotic Chess Player

Binocular three-dimensional vision is an important branch of machine vision technology.It can solve the problem like as robot navigation,vision detecting and industry robot arm snatching. Binocular three-dimensional vision take a same scene photos with moving or circumrotating in one or two cameras in different position,and it get a point 3D information by computing parallax theory in the two photos.The research of the binocular vision include camera calibration,ima-ge processing,stereo matching,object recognition,etc.The aim of this article is designing a set of binocular system in chess robot.The function of this machine vision system offer chess 3D parameter for robot by binocular vision system.The research of this article closely enclose binocular vision system in Chinese chess robotThe emphases of research in chess robot are vision information processing and chess player recognition.The pivotal technologies are gain chess space parameters and recognizing chessman. Gaining space parameter is by the operation of point to point As long as the dots in left and right vidicon image plane can match each other,we can gain their corresponding three-dimensional coordinates.The means of circle statistic is used in chessman recognition.We constructed a binocular vision hardware system with one image collection card controlling two cameras based on personal computer to capture the object's images from two different space points.Calibrated cameras by Tsai two steps method and got the inner and external parameters.We transform the color image by image collection card to gray image and processed the images.After processing two images,we can match the character points.The left and right image character points had matched by their gray relativity. At last,on the base of getting the inner and external parameter and finishing character points matching,we could get the central coordinates of chessman and recognize it.