双目立体视觉测量方法研究

提出了一种简易快速的特征点提取算子，并与经典算法作了实验对比，证实了所提算法简单、有效、可靠，对外界光照条件具有较强的适应性，且抗干扰能力强。在进行特征点匹配时，在现有约束条件的基础上，提出了基于被测物体形态与摄像机相对位置的视差范围约束，根据视差方向性并采用动态搜索方法来缩小匹配点的搜索范围，从而提高了搜索速度，降低了错误匹配概率。通过对具有已知三维坐标标准件的计算机仿真实验，证实了所提出的特征点识别与提取及匹配方法是切实有效的。

Study on Measurement Method Based on Binocular Stereo Vision

A kind of simple and fast arithmetic operators to extract feature points is presented. It is proved that the algorithm is simple, effective and reliable; moreover, the algorithm has strong adaptability to outside illumination condition and stronger anti jamming capability after comparison with classic algorithms by experiment. During matching, disparity scope restriction based on the shape of measured object and relative position between the camera and the measured object is mentioned also. Based on the directionality of disparity, the searching scope of matching feature points is reduced so that the searching speed is improved and the error probability on matching is dropped down after the dynamic searching method being used. It is proved that the algorithm to extract feature points and the matching method is feasible by computer imitated experiment on standard piece with given 3D coordinates.