基于Hough变换的汽车零件中圆形检测的改进算法

摘要

针对汽车零件中的圆检测实际需求，在分析了基本Hough变换和随机Hough变换进行圆检测的技术特征基础上，对基本Hough变换进行算法改进。采用Sobel算子提取图像边缘，利用圆参数范围已知的先验知识确定感兴趣区域和圆半径的检索范围，使得计算量大大减少，从而实现圆心坐标和半径的快速检测，满足工业生产实时性的要求。%In the automobile components detection for the actual needs, the analysis of the basic Hough transform and ran-dom Hough transform for circle detection technical characteristics, it improves the basic Hough transform algorithm. Using Sobel operator to extract image edge, and circular range of parameters prior knowledge is used to determine the region of interest and the range of radius, so the calculation is greatly reduced, thus realizing the rapid detection of the center coordi-nates and radius, and it meets the real-time requirements of industrial production.