Wafer Inspection Color Mapping for Defect Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing visual inspection methods for semiconductor wafers, particularly for silicon carbide wafers, lack consistency due to subjective human evaluation, leading to erroneous defect determination.
Innovation Solution
A wafer inspection method and apparatus that converts scanning information into a color image by calculating feature values such as mean, standard deviation, and coefficient of variation, and sets haze thresholds to generate a color image illustrating the wafer texture, thereby reducing subjective interpretation and improving defect determination accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If visual inspection is performed by human eyes, then inspection can be conducted, but determination accuracy deteriorates due to subjective evaluation
Solution Approach 1:
The patent replaces human visual inspection with an automated optical inspection system that uses cameras and image processing algorithms to objectively evaluate wafer flatness and detect defects, eliminating subjective human judgment while maintaining operational capability
Solution Approach 2:
The patent transforms subjective visual assessment into objective quantitative parameters by measuring haze values, calculating feature values (mean, standard deviation, coefficient of variation), and comparing them against predefined thresholds to automatically determine defect status
2Measurement precision
If automated optical inspection is used, then inspection objectivity is improved, but inspection complexity increases
Solution Approach 1:
The patent divides the wafer inspection into discrete unit blocks, calculates feature values for each block independently, and processes them through standardized algorithms, which simplifies the overall inspection process while maintaining high accuracy through systematic breakdown of the inspection task
Solution Approach 2:
The patent converts complex visual assessment into standardized quantitative parameters (haze values, feature values, color values) that can be automatically measured and compared against thresholds, reducing the complexity of decision-making while improving objectivity
3Ease of operation
If manual visual determination is used, then inspection simplicity is maintained, but inspection consistency deteriorates
Solution Approach 1:
The patent replaces manual visual determination with automated image processing that applies consistent algorithms and thresholds to all inspections, ensuring uniform evaluation criteria are applied every time without variation in human judgment
Solution Approach 2:
The patent transforms subjective visual consistency into objective numerical consistency by measuring and comparing quantitative parameters (haze values, feature values) against standardized thresholds, ensuring identical inspection results for identical wafer conditions regardless of inspector
Data Source
AI summary
The disclosure provides a wafer inspection method and wafer inspection apparatus. The method includes: receive scanning information of at least one wafer, wherein the scanning information includes a plurality of haze values; the scanning information is divided into a plurality of information blocks according to the unit block, and the feature value of each of the plurality of information blocks is calculated according to the plurality of haze values included in each of the plurality of information blocks; and converting the feature value into a color value according to the haze upper threshold and the haze lower threshold, generating the color value corresponding to the at least one wafer according to the converted color value according to the feature value, whereby the color graph displays the texture content of the at least one wafer.