A Method And Experiment For Measuring Distance Of Object Parallax From Image

Human using vision to obtain the three-dimensional information from outside,the binocular parallax ranging is a technique of ranging 2D images.Now it has played an increasingly important role in military, aerospace, industry, medicine and other fields.The binocular parallax ranging system includes image acquisition, camera calibration, feature extraction, image matching and 3D reconstruction technology,the key technology is camera calibration and image matching,this paper focuses on these two key technologies.In this paper,based on the domestic and foreign existing binocular stereo vision system,according to the characteristics of experimental object,a method of measuring the distance of the target object and the parallax of the image matching points is designed. After the experimental analysis,the relative error of ranging is less than 3%, which is suitable for near range.In this paper, a large number of matching points are obtained by using automatic matching technology. At present, the automatic matching technology based on the SIFT algorithm has been more mature and precision, and it has been widely used in computer vision field.Therefore, the SIFT algorithm is adopted to obtain the matching points.The main flow of SIFT algorithm is: scale space construction, feature points extraction, feature points matching.