基于机器视觉薄零件高精度测量技术的研究

为快速准确检测薄零件几何量精度,提出了一种基于机器视觉的检测方法。根据具体的视场和精度要求,完成电荷耦合元件(CCD)传感器、镜头、光源等选型与硬件系统设计。开发了软件系统,采用双三次插值法细分图像后提取亚像素边缘,变权重最小二乘法拟合出零件轮廓的圆弧和直线,自动计算零件的圆度误差,中心距尺寸误差,并判断其是否合格。通过齿轮泵中间体零件的检测结果表明,本系统能快速准确检测中高精度(IT7级)薄零件的几何量精度,满足实际应用的需求。

Research on High Precision Measurement Technology of Thin Parts Based on Machine Vision

In order to measure geometrical precision of the thin parts rapidly and accurately, a measurement method based on machine vision was represented. According to the requirement of field of view and precision, hardware and software of the measurement system including the CCD sensor, lens, and the light source were developed. The subpixel edges of the image were extracted after bicubic interpolation subdivision. And the contour of the part is segmented into the arcs and lines by least square fit method with weight function. Then the roundness error and the center distance size error were calculated automatically. The measurement results of gear pump intermediate part show that, this system can used in geometry precision (IT7) measurement of thin parts quickly and accurately.