BGA焊点气泡缺陷X射线图像的动态阈值分割方法

Dynamic Threshold Segmentation of Bubble Defects in BGA Solder Balls

BGA焊点使用X射线检测时,气泡缺陷的灰度对比度较低,如何准确分割一直是研究热点。提出一种动态阈值分割算法,首先对原始图像进行平滑处理,然后用原始图像减去平滑图像,对得到的差图像设定阈值进行分割,实现目标区域的提取。实验中使用典型BGA焊球的X射线图像,通过控制图像的平滑程度和差图像阈值的大小,分别提取焊球区域和气泡区域,二者相与后实现气泡缺陷的最终分割。实验过程中与OTSU分割算法做对比,表明本算法具有更好的适应性和分割结果。

Accurate segmentation of bubble defects is a research hot spot because of low gray contrast when the BGA solder joints are detected by x-ray.A dynamic threshold segmentation is proposed which the original image is first smoothed,and then subtract the smoothed image to obtain a new image which is segmented to get target area by using a specific defined threshold.Typical BGA xray images were used to test the segmentation by controlling the degree of image smoothing and threshold of difference image to extract the solder ball area and bubble areas respectively which were performed logic and to achieve the final segmentation.The experimental results show that the proposed algorithm has better adaptability and segmentation results compared with the OTSU segmentation algorithm.