Research On Calibration And Reconstruction Of Binocular Stereo-vision System

Binocular stereo vision as an important branch of computer vision hasbeen the focus and hot spots of research. The basic principle of binocularstereo vision is simulating the process of handing scene of human vision, wecan get different images with two different viewpoints and then calculate theposition deviation of images pixels based on triangular measure. Finally wecan reconstruct the3-D coordinates of the object.According to the principle of binocular stereo vision, the authorconstructs the3-D reconstruction system of binocular stereo vision anddiscusses about the relevant key technologies of the system, such as cameracalibration, stereo matching,3-D reconstruction and so on. By using VC++and Halcon software development platform and relevant hardware equipment,a3-D reconstruction system of binocular stereo vision is built successfully.Analyzing the principle of camera calibration, Zhang-Zhengyou’smethod is adopted to get the exterior and inner parameters of two cameras,and the binocular camera calibration result is got too. Stereo matchingalgorithms based on feature and region are studied, considering experimentenvironment, epipolar constraint technique is adopted in the processing ofstereo matching. Discussing about the method of restoring the object’s3-Dcoordinates in binocular stereo vision system. In this paper, using theparameters of the camera calibration, we calculate3-D coordinates of thematching points and then get the depth information of the object in the scene.Finally, to reconstruct and display3-D scene in different viewpoints by usingVisual Studio2008.