一种基于智能感知的晶圆微结构缺陷全域判读方法
By applying a gridding process and an adaptive merging strategy to the wafer image, the problem of significant differences in features between different regions in the wafer image was solved, achieving a defect detection effect with high detection rate and low false alarm rate.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHAANXI SUN MOON CORE SEMICON CO LTD
- Filing Date
- 2026-05-07
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot effectively address the significant differences in features between different regions in highly complex wafer images, resulting in over-enhancing in dense areas or insufficient enhancement in sparse areas by the global uniform contrast enhancement strategy, making it difficult to achieve high detection rates and low false alarm rates.
By segmenting the wafer grayscale image into multiple minimum grids, calculating the merging similarity of adjacent grids, adaptively setting the merging threshold and merging region blocks layer by layer, setting a limiting value based on the number of merging levels, performing differentiated contrast enhancement, and eliminating block effects through boundary distance weighting smoothing.
It achieves precise physical segmentation of functional areas on the wafer surface, improves the detection accuracy of defects and reduces the false alarm rate, and ensures global optimization of image quality and robust interpretation.
Smart Images
Figure CN122175965B_ABST