针对拉深缺陷人工检测的不足问题,提出了一种生产线上表面缺陷检测系统的设计方案。通过Halcon和VC++的联合开发,实现图像处理流程以及检测结果输出,包括摄像机标定、对图像进行平滑处理、利用动态阀值分割表面缺陷特征,通过形态学处理识别破裂缺陷,并对破裂缺陷分类和测量以及输出破裂缺陷外形和尺寸等,实现冲压生产线拉深件表面缺陷的自动检测功能。其中在相机标定过程中,采用Halcon的相机标定方法,快速获取图像坐标与世界三维坐标的对应关系,并实现缺陷特征的准确测量。此方案具有一定的稳定性,可提高检测效率。 Aimed at the deficiency of the manual inspection of the deep drawing defect, a design scheme of the on-line surface defect inspection system was proposed. Through the joint development of Halcon and VC ++, the image processing flow and the output of the test results are realized, including camera calibration, image smoothing, dynamic threshold segmentation of surface defects, morphological processing to identify rupture defects, classification of rupture defects and Measurement and output rupture defect shape and size, etc., to achieve the stamping production line drawing parts surface defects automatic detection. In the process of camera calibration, Halcon’s camera calibration method is used to quickly obtain the correspondence between image coordinates and the world’s three-dimensional coordinates, and to achieve accurate measurement of defect features. This program has a certain stability, can improve the detection efficiency.