Mask Defect Detection Using Histogram Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for inspecting masks used in semiconductor manufacturing have low determination accuracy due to unreliable reference images and allowable error ranges, making it difficult to detect defects in electron beam irradiating apparatuses.
Innovation Solution
A method and apparatus that obtain and analyze histograms of mask patterns to detect defects by measuring grey levels and comparing them with pattern information, including noise removal techniques such as smoothing and binarization, to accurately identify bridges and shifts in lattice-like and linear patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image comparison method is used, then inspection process is simple, but determination accuracy is low
Solution Approach 1:
The patent replaces the conventional mechanical image comparison method with a histogram-based quantitative analysis method. Instead of visually or mechanically comparing images, the system converts mask patterns into histograms of grey levels and compares these statistical representations, enabling more precise defect detection through mathematical analysis rather than traditional image superposition.
Solution Approach 2:
The patent transforms the inspection approach by changing from direct image comparison to histogram comparison. By converting spatial image data into statistical parameter distributions (histograms of grey levels), the system achieves higher determination accuracy through quantitative parameter analysis rather than qualitative image matching.
2Reliability
If reference image comparison is used, then inspection is straightforward, but reliability of defect detection is insufficient
Solution Approach 1:
The patent substitutes the conventional reference image comparison approach with a histogram-based statistical analysis system. This replacement enhances reliability by using quantitative histogram data that objectively reflects pattern characteristics, eliminating the subjectivity and error margins inherent in reference image comparison methods.
Solution Approach 2:
The patent introduces histogram comparison as an intermediary between the mask pattern and the defect detection decision. Instead of directly comparing images, the system uses histograms as an intermediate statistical representation that captures essential pattern characteristics, providing a more reliable basis for defect identification.
3Measurement precision
If image comparison with allowable errors is used, then inspection is quick, but measurement precision deteriorates
Solution Approach 1:
The patent replaces the conventional image comparison method with histogram-based quantitative analysis. This substitution enables precise defect detection by analyzing the statistical distribution of grey levels, allowing the system to identify even subtle defects that would be obscured by allowable error margins in traditional image comparison.
Solution Approach 2:
The patent transforms the inspection methodology by changing from direct image comparison to histogram parameter comparison. By analyzing the statistical parameters (grey level distributions) rather than pixel-by-pixel image data, the system achieves high measurement precision while maintaining inspection efficiency through automated histogram generation and comparison.
Data Source
AI summary
In a method of inspecting a mask, an image of a pattern on the mask may be obtained. A histogram of the image by grey levels may be obtained. The histogram may be compared with information of the pattern to detect a defect of the mask. Thus, reliability of defect detection in the mask may be remarkably improved.