Elliptical Wafer Shape Recognition Using Coordinate Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional shape recognition methods for semiconductor wafers struggle with accurately determining the shape due to noise interference at the edge, making it difficult to achieve precise recognition.
Innovation Solution
A shape recognition method and device that utilizes coordinate distributions and statistical calculations, such as average values and standard deviations, to determine center coordinates and axial lengths of elliptical shapes, thereby reducing the impact of edge noise and enabling quick and robust shape recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If edge coordinate extraction method is used, then shape recognition can be performed, but noise at the edge significantly affects recognition accuracy
Solution Approach 1:
The patent extracts only the necessary coordinate information (coordinates of pixels belonging to the elliptical shape) from the image, separating the useful shape data from the noisy edge information. By working with extracted coordinate points rather than full edge images, the method eliminates the harmful effect of edge noise while preserving the essential shape characteristics needed for accurate recognition.
Solution Approach 2:
The patent introduces coordinate distribution analysis as an intermediary step between image capture and shape parameter calculation. Instead of directly measuring from noisy edge images, the method uses coordinate distributions along with statistical parameters (average values and standard deviations) as intermediaries to derive shape parameters, thereby filtering out noise and achieving accurate shape recognition.
2Manufacturing precision
If conventional edge-based methods are used, then shape can be determined, but processing time and computational resources increase
Solution Approach 1:
The patent replaces complex mechanical/image processing operations (edge detection, noise filtering, contour tracing) with simpler mathematical operations on coordinate data. By substituting the mechanical image processing system with a coordinate-based mathematical system using statistical calculations, the method achieves the same shape determination goal with significantly reduced processing time and computational resources.
Solution Approach 2:
The patent changes the fundamental parameters used for shape determination from pixel intensity values and edge coordinates to statistical parameters (average values and standard deviations of coordinate distributions). This parameter transformation simplifies the calculation process while maintaining shape recognition accuracy, thereby reducing processing time without sacrificing precision.
Data Source
AI summary
According to one embodiment, a shape recognition method executed by a shape recognition device includes acquiring coordinate values in a first direction and a second direction in a captured image of an upper surface of a shape recognition target having an elliptical shape, calculating center coordinates of the elliptical shape based on distributions of the coordinate values, and calculating axial lengths of the elliptical shape based on the distributions of the coordinate values.