空洞缺陷是BGA焊接缺陷中比较常见的一种,主要由回流焊时产生的气体没有及时排出而导致。X射线无损检测技术可以使空洞缺陷在焊球图像上显示为白色区域供技术人员查看,但在噪声、不均匀照射、存在与缺陷类似目标的干扰下如何准确地自动提取缺陷一直是个难题。提出使用Otsu算法分割焊球,使用数学形态学中的开闭运算、顶帽变换提取空洞缺陷的技术方案,实现每个焊点空洞缺陷的准确提取和自动分析。实验结果表明,提出的技术方案鲁棒性强,可应用于BGA焊点的空洞缺陷检测。 Hole defects are the more common type of BGA soldering defects and are mainly caused by the gas evolved during reflow and are not discharged in time. X-ray non-destructive testing technology can make the hollow defect appear as a white area on the solder ball image for the technician to view. However, it is always a difficult problem to accurately and automatically extract defects under the interference of noise, inhomogeneous irradiation and existence of similar defects. The Otsu algorithm is proposed to segment the solder ball, the technical scheme of opening and closing operation in mathematical morphology and top hat transform to extract the cavity defects, and the accurate extraction and automatic analysis of the cavity defects in each spot are achieved. The experimental results show that the proposed technical scheme is robust and can be applied to the detection of hollow defects in BGA solder joints.