Hybrid Semiconductor Wafer Inspection Framework via Autonomous Data Annotation
Changheon Han,Heebum Chun,Jiho Lee,Fengfeng Zhou,Huitaek Yun,ChaBum Lee,Martin Jun
DOI: https://doi.org/10.1115/1.4065276
2024-04-05
Journal of Manufacturing Science and Engineering
Abstract:Abstract Semiconductors play an indispensable role in data collection, processing, and analysis, ultimately enabling more agile and productive operations. Given the importance of wafers in semiconductor fabrication, the purity of a wafer is essential to maintain the integrity of the overall manufacturing process. To tackle this issue, this study proposes a novel Automated Visual Inspection (AVI) framework for scrutinizing semiconductor wafers from scratch, capable of both identifying defective wafers and pinpointing the location of defects through autonomous data annotation. Initially, this proposed methodology leveraged a texture analysis method known as Gray Level Co-occurrence Matrix (GLCM) that categorized wafer images—captured via a stroboscopic imaging system—into distinct scenarios for clear and noisy wafer inspection. GLCM approaches further allowed for a complete separation of noisy wafers into defective and normal wafers as well as the extraction of defect images from noisy defective wafers, which were then used for training a Convolutional Neural Network (CNN) model. Consequently, the CNN model excelled in localizing defects on noisy defective wafers, achieving an F1 score exceeding 0.901. In clear wafers, a background subtraction technique represented defects as clusters of white points. The quantity of these white points not only determined the defectiveness of clear wafers but also pinpointed locations of defects on clear wafers. Lastly, the application of a CNN further enhanced performance, robustness, and consistency irrespective of variations in the ratio of white point clusters. This technique demonstrated accuracy in localizing defects on clear wafers, yielding an F1 score greater than 0.993.
engineering, mechanical, manufacturing