摘要
利用显微镜观察和统计晶体缺陷是一种常规的测试方法,但由于研发初期微管、孔洞等缺陷数通常较多且分布不均匀,统计起来非常费时费力。针对上述问题,介绍了一种利用MATLAB图像处理技术的硫化镉晶片微管检测计数的新方法。借助MATLAB图像处理工具箱,对微分干涉显微镜下的硫化镉晶片照片进行处理,编写晶片微管检测计数程序,统计评价区域内晶片的微管数,并且该程序具有手动调节阈值功能。研究结果证明,利用此方法统计的微管数准确度可达95%,借助图像处理技术检测统计微管数是一种高效、快速、准确的新方法,可实现检测、统计过程的自动化,是一种非常实用的技术。
Using microscope to detect and statistic wafer defection is a general method. In order to solve the problem that defect inspection by manual counting is low accuracy and time consuming,the new method of cadmium sulfide（ Cd S） wafer micropipe detection and statistics was introduced by image processing. By using the MATLAB image processing technology,the differential interference microscope images of Cd S wafer were shown. The micropipe detection counting program of Cd S wafer was wrote,and the number of micropipe were recorded,and the program has the threshold adjustment function. The results show that the average statistical accuracy of the proposed method is 95%,and the MATLAB image processing technology is an effective,rapid and accurate new evaluation method for detection and statistics of wafer micropipe. The technology can be releazied detection and statistics automatically,it is a very usefull technique.
出处
《半导体技术》
CAS
CSCD
北大核心
2015年第2期152-156,共5页
Semiconductor Technology
关键词
晶片
微管
缺陷统计
图像处理
阈值调节
wafer
micropipe
defect statistics
image process
threshold adjustment
