基于稀疏成像的半导体薄膜材料界面缺陷检测  

Interface Defect Detection of Semiconductor Thin Film Materials Based on Sparse Imaging

作　　者：李聪[1] 谭明 刘小标 李辉[1] LI Cong;TAN Ming;LIU Xiao-biao;LI Hui(College of Science,Henan Agricultural University,Zhengzhou Henan 450000,China)

机构地区：[1]河南农业大学理学院,河南郑州450000

出　　处：《计算机仿真》2024年第1期197-200,226,共5页Computer Simulation

基　　金：国家自然科学基金(22005087)。

摘　　要：复杂背景下检测半导体薄膜材料界面微小缺陷具有一定的难度,为了精准检测半导体薄膜材料界面缺陷,提出一种稀疏成像下半导体薄膜材料界面缺陷检测方法。扫描采集半导体薄膜材料界面二维图像,对含有噪声的半导体薄膜材料界面实施小波分解,获取不同频带的子图像。低频图像保持不变,选择对应的模板对高频图像滤波处理,将滤波处理后的高频图像和低频图像两者合成,获取去噪后的图像。通过机器视觉定位薄膜材料界面的缺陷位置,提取缺陷区域特征,采用稀疏成像对特征参数修正,完成半导体薄膜材料界面缺陷检测。仿真结果表明,采用所提方法可以获取更加精准的检测结果,用时比较短,满足高效与高精度检测需求。Detecting interface defects of semiconductor thin film materials under complex backgrounds is challenging.In order to accurately detect interface defects,a method for detecting defects in the interface of semiconductor thin film materials was proposed based on sparse imaging.Firstly,the two-dimensional images of the interface of semiconductor thin film materials were scanned and collected.Then,wavelet decomposition was applied to the interface with noise thus obtaining sub-images with different frequency bands.Under the condition that the low-frequency image remained unchanged,corresponding templates were selected to filter high-frequency images.After filtering,the high-frequency and low-frequency images were combined to obtain a denoised image.Meanwhile,the defect of the film material interface was located by machine vision.Meanwhile,the feature of the defect region was extracted from the position Moreover,feature parameters were corrected by sparse imaging,and ultimately,defect detection in the interface of semiconductor thin film materials was completed.Simulation results show that the proposed method can obtain more accurate test results with shorter processing time,which meets the requirements for efficient and high-precisiontests.

关 键 词：稀疏成像 半导体 薄膜材料 界面缺陷检测 小波分解 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置]