基于机器视觉的套筒件尺寸测量研究  

Research on dimensional measurement of sleeve pieces based on machine vision

作　　者：蔡舒 吴超华[1] 罗威 史晓亮[1] CAI Shu;WU Chaohua;LUO Wei;SHI Xiaoliang(Wuhan University of Technology,Wuhan 430070,China)

机构地区：[1]武汉理工大学,武汉430070

出　　处：《现代制造工程》2025年第4期123-131,共9页Modern Manufacturing Engineering

基　　金：国家自然科学基金项目(52375201)。

摘　　要：针对传统工业检测中测量套筒尺寸效率低、精度低等问题,提出一种基于机器视觉的套筒尺寸测量方法。首先,通过Halcon软件校准后的CMOS相机采集套筒图像,再经过双边滤波和阈值分割预处理;然后,采用改进的Canny算法结合Otsu算法获取的最佳分割阈值,获得边缘像素级坐标;再然后使用改进的Zernike矩算法对边缘进行精确定位;最后,采用最小二乘法拟合内、外圆求解出套筒内、外径。实验结果表明,该方法平均测量精度小于0.02 mm,相对未改进算法在时间上减少了28.3%,测量精度和效率满足企业快速、精确的检测要求。Aiming at the problems of low efficiency and low precision of measuring sleeve size in traditional industrial inspection, a sleeve size measurement method based on machine vision was proposed. Firstly, the sleeve image was captured by the CMOS camera calibrated by Halcon software, and pre-processed by bilateral filtering and threshold segmentation. The improved Canny algorithm combined with the optimal segmentation threshold obtained by the Otsu method was used to obtain the pixel-level coordinates of the edges. The improved Zernike moments algorithm was used to locate the edges accurately. Finally, the inner and outer diameters of the sleeve were solved by fitting the inner and outer circles using the least squares method.The experimental results show that the average measurement accuracy of the method is less than 0.02 mm, and the measurement accuracy and efficiency meet the requirements of enterprises for rapid and accurate inspection.

关 键 词：机器视觉 套筒 亚像素 尺寸测量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]