基于机器视觉的笔芯球珠表面缺陷检测系统研究

Research on Detection System for Surface Defects of Refill Ball Based on Machine Vision

针对笔芯球珠表面缺陷检测识别问题,设计并实现了基于机器视觉的笔芯球珠表面缺陷检测系统。笔芯球珠在球面展开机构作用下,通过图像采集模块获取5张可以完全覆盖整个球面的图像。通过对每幅图像进行缺陷图像提取后,采用基于轮廓角点匹配的方法实现对每幅图像中缺陷图像的拼接;基于提取的有效特征组合通过KNN分类算法对完整的缺陷图像进行缺陷识别。试验结果表明,该方法能够对笔芯球珠表面缺陷进行精确有效的检测与识别。

To detect the surface defects of refill ball, this paper designs a detecting system for its surface defects based on machine vision. Under the effect of the spherical unfloding module, five images which can entirely cover the whole ball surface are obtained by image acquisition module. The defect image extracted from the 5 images is mosaiced based on contour. Then, based on the combination of the extracted effective features, the whole defect image is identified by KNN classification algorithm. The experimental result shows that the method can be used to detect and recognize the ball surface defects accurately and effectively.