一种基于立体视觉的车轮中心测量方法

A wheel center measurement method based on stereo vision

在车辆尺寸参数测量系统中,车轮中心是一个很重要的过程参数,目前该参数都是由人工进行手动测量,为此提出一种基于立体视觉的车轮中心高精度测量方法.该方法以车轮钢圈作为处理对象,首先根据边缘分组算法和椭圆拟合判别算法提取出车轮钢圈的圆形特征;其次利用立体匹配算法获取车轮钢圈圆形特征上点的三维坐标;最后通过三维平面拟合、坐标转换以及平面圆拟合算法获取车轮中心的三维坐标.实验结果表明:该方法能够有效地获得车轮中心的三维坐标,为后续车辆尺寸参数的测量奠定良好基础.

As an important dimensional parameter,wheel center is always measured manually,therefore a high precision measurement method based on stereo vision is proposed in this paper.The edge grouped algorithm and ellipse fitting algorithm are applied to extract the circle shape of the steel ring,then the subsection stereo matching algorithm is used to acquire the 3D coordinate of interest points in the circle shape.Finally through the 3D plane fitting algorithm,coordinate transformation algorithm and circle fitting algorithm,the 3D coordinate of the wheel center is calculated.Measurement experiment results show that the method can acquire the 3D coordinate of the wheel center efficiently,thus lay a foundation for other vehicle parameter measurement.