Dynamic Measurement Technology Research Body The Appearance Of Geometric Feature Size

With the development of computer technology, stereoscopic vision in theoretical research and practical applications have achieved remarkable results, the use of stereo vision method to establish the human body model is also increasingly sophisticated technology. It is used in aerospace, medical rehabilitation, traffic regulation, sports analysis, virtual reality and other aspects of the three-dimensional motion detection and more widely. In this paper, based on the key technology to achieve a three-dimensional reconstruction of binocular stereo vision includes camera calibration, feature extraction and matching, and the depth of information to determine the three-dimensional coordinates of the calculation.Calibration for binocular stereo vision camera, this paper OpenCV camera calibration based on the analysis of existing models and binocular stereo vision system camera calibration methods, gives consideration camera lens distortion, calibration based on OpenCV binocular stereo camera methods. Experimental results show that the method calibration parameters stable and precision to meet the requirements of three-dimensional reconstruction of binocular stereo vision and other applications.For three-dimensional reconstruction of the feature extraction, matching, this paper describes the matching criteria epipolar constraint, the constraint will match point on a pole line, can effectively eliminate false matches and greatly improve the matching efficiency. This paper uses the SIFT feature extraction matching algorithm. and designed two experiments to test the performance of SIFT algorithm, experiments show that the algorithm can solve the image rotation, zooming, panning and other stereo matching case, matching accuracy than the high, is a very good algorithm.Depth information to determine and calculate the three-dimensional coordinates of the three-dimensional reconstruction of the core steps, this paper analyzes three-dimensional camera imaging model, according to the basic principles of the proposed stereoscopic three-dimensional coordinates calculation. And then design experiments to generate a disparity map, and calculate the spatial coordinates of the point D, the data on the subject body to be measured to verify the results of the calculation and the error cause and the measured results are analyzed, this method proved to be effective.System is measured by body height, arm length, shoulder and other data for comparison with the actual measured data, there are about2%of the number of errors, the results agree well, therefore convenient method with the traditional manual measurement, fast and the dimensions of the various actions can be collected.