基于双目立体视觉的目标测距系统

Distance measurement system based on binocular stereo vision

双目立体视觉能够在多种条件下感知三维场景的立体信息,基于双目立体视觉的测距以其特有的优势在测距中扮演着越来越重要的角色。研究了双目测距的原理、摄像机的标定、立体校正、立体匹配等关键技术,在此基础上提出了用MATLAB和OpenCV相结合的方法实现双目立体视觉的测距。使用MATLAB进行相机的标定,然后借助OpenCV实现立体校正和立体匹配。通过立体匹配得出的视差值,可以求解出目标和相机之间的距离实现测距。通过实验的结果表明,提出的双目测距方法在一定范围内达到了精度要求。

Binocular stereo vision can understand three-dimensional information of obiects in a variety o（ conditions. Distance measurement based on binocular stereo vision plays a more and more important role in the distance measurement system. This paper studies the principle of distance measurement, camera calibration, stereo rectification, stereo matching and other key technologies, based on which this paper propose a method to calculate the distance by MATLAB and OpenCV. Using MATLAB to calibrate camera, using OpenCV to rectify and match images, distance between camera and object can be get via the disparity value. The experiments show that this distance measurement method can be effective within a certain range.