一种改进的微波芯片金锡共晶焊接X光图像空洞检测方法  

An Improved Cavity Detection Method for X-ray Image of Gold-Tin Eutectic Soldering of Microwave Chip

作　　者：林俊宇 康陈子 邢增繁 郑博文 邱晓晖[1] LIN Junyu;KANG Chenzi;XING Zengfan;ZHENG Bowen;QIU Xiaohui(Nanjing University of Posts and Telecommunications,Nanjing 210003,China)

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003

出　　处：《传感器世界》2022年第7期17-23,共7页Sensor World

基　　金：江苏省自然科学基金(No.BK2011789)。

摘　　要：空洞检测是保证微波芯片金锡共晶焊接质量的关键之一。随着图像处理技术的长足发展,基于X光图像的空洞检测逐渐成为该领域自动无损检测的有效手段。文章使用数学形态学方法进行X光图像空洞的检测,将不同结构元素的毛刺及干扰导线的去除力作对比,同时,针对中值滤波对图像边缘腐蚀的问题,采用优化的中值滤波进行简单的四角修复,进一步使用Symmetric型中值滤波来解决图像撕裂问题;其次,针对烧结空洞提取面积存在缺失现象,提出基于区域生长的改进OTSU阈值分割方法,该方法以OTSU阈值分割结果为种子点,通过区域生长生成完整的烧结空洞,有效改善了烧结空洞的提取完整性;最后,金锡共晶X射线图像对比处理结果,从主观及品质因数fom客观评价上均验证了文章改进方法的有效性。Morphological image processing is a series of image processing techniques that process image shape characteristics based on mathematical morphology. In the paper, mathematical morphological methods are used to detect the holes in X-ray images, comparing the burrs of different structural elements and the removal forces of interfering wires. At the same time, in view of the problem of median filter corrosion on the edge of the image, the paper uses the optimized median filter to perform a simple four-corner repair, and further uses the Symmetric median filter to solve the image tearing problem. Secondly,in response to the problem that the OTSU threshold segmentation method will reduce the extraction area of the sintered hole, we use the regional growth method to use the residual white point as the seed point to grow an area similar to the area of the sintered cavity that should have been displayed at the location. Finally, through the objective evaluation of the quality factor data fom, it is found that the image detection effect is better after the improvement of the paper.

关 键 词：数学形态学 中值滤波 区域生长 

分 类 号：TN957.52[电子电信—信号与信息处理]