摘要
针对印刷线路板凹陷和空洞缺陷难检测的问题,提出了一种基于背景连通域的线路板缺陷定位及识别方法。方法首先利用Gerber文件产生线路板标准图像,利用CCD摄像装置获取线路板待测图像;其次,采用定位孔检测法和仿射变换法实现标准图像和待测图像的配准;再次,采用差影算法和形态滤波法对待测线路板缺陷进行定位;最后,采用背景连通域方法和铜料检测法实现缺陷类型识别。通过对带缺陷线路板的检测实验,结果验证所提方法不仅能对常见缺陷准确定位及识别,而且能对凹陷和空洞缺陷准确高效识别。
Owing to the difficulties in the detection of pit and void defects of Printed Circuit Board(PCB),a defect location and recognition method based on background connected domain was proposed.Firstly,the Gerber file was used to get the PCB standard image,and the CCD camera device was used to acquire the PCB determined image.Secondly,the registration of the PCB standard image and the determined image was achieved by detecting the location of holes and affine transform.Thirdly,determined circuit board defects were located by subtraction algorithm and morphological filtering method.Finally,defects were recognized by the background connected domain method and copper material detection method.The PCB detection experiment demonstrates that the method not only locates and recognizes common defects accurately,but also recognizes the pit and void defects accurately and efficiently.
出处
《半导体光电》
CAS
北大核心
2015年第5期820-823,842,共5页
Semiconductor Optoelectronics
基金
国家自然科学基金项目(61462063)
航空科学基金项目(20135756010)
江西省高等学校科技落地计划项目(KJLD13058)
关键词
印刷线路板
背景连通域
差影算法
缺陷定位
缺陷识别
printed circuit board
background connected domain
subtraction algorithm
defect location
defect recognition