A Wafer Surface Defect Detection Method Based on Gabor Feature and Random Dimensionality Reduction

Abstract

The invention discloses a wafer surface defect detection method based on Gabor features and random dimensionality reduction. Use a CCD camera to collect wafer surface images, and then preprocess the images; design 40 Gabor filters for obtaining wafer surface texture features, and then perform convolution operations on 40 Gabor filters and images to obtain 40 feature images ; Use random dimensionality reduction for 40 feature images: perform threshold segmentation on the image after dimensionality reduction, construct an objective function for the segmentation threshold, solve the objective function to obtain the final segmentation threshold, use the final segmentation threshold to segment the image into foreground and background, and determine the segmentation Threshold, and ultimately accurately detect wafer surface defects. The method of the invention can well identify and locate the wafer surface defect, and the identification efficiency is also greatly improved.

Description

technical field [0001] The invention relates to a wafer surface defect detection method based on Gabor features and random dimensionality reduction, which belongs to the field of image recognition. Background technique [0002] With the rapid development of integrated circuit manufacturing technology, the feature size of wafers is continuously reduced, resulting in more tiny defects. Defects on the wafer surface have become a major obstacle affecting yield. How to accurately and automatically detect wafer defects is a complex and challenging task. Defect detection technology has become a key technology in the integrated circuit industry. [0003] The traditional method is generally to manually detect wafers with relatively large defects, but the detection effect is often not good and the efficiency is low. Machine detection mostly relies on image detection methods, among which template matching is the most commonly used method, but the matching time is too long and it is ...

AI Technical Summary

Problems solved by technology

Wafer surface defects have become a major obstacle to yield

How to accurately and automatically detect wafer defects is a complex and challenging task

[0003] The traditional method is generally to manually detect wafers with large defects, but the detection effect is often not good and the efficiency is low

Machine detection mostly relies on image detection methods, among which the method based on template matching is the most commonly used, but the matching time is too long and it is difficult to achieve real-time performance, so it is difficult to meet industrial needs

Images