Rotation Angle Detection for Semiconductor Circuit Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for detecting rotation angles from images of semiconductor circuit patterns captured by SEM are inadequate, as they are designed for document images with regular character arrangements, failing to accurately handle images with vertical and horizontal wiring lines and no regular small object arrangements.
Innovation Solution
The method involves Fourier transform-based edge extraction, evaluation value calculation for each angle using a concentric circle area in the Fourier-transformed image, and pseudo-Radon transform to determine the rotation angle, with optional smoothing filters to stabilize angle detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If existing document image processing techniques are applied to SEM images of circuit patterns, then the processing method is simple, but the rotation angle detection accuracy deteriorates because circuit pattern images lack regular small object arrangements like characters
Solution Approach 1:
The invention changes the detection parameters by using Fourier transform to convert spatial domain image data into frequency domain data. This transformation allows the detection method to adapt from document image characteristics to circuit pattern characteristics, maintaining simplicity while improving accuracy for the target application.
Solution Approach 2:
The invention replaces traditional mechanical projection-based angle detection with a mathematical Fourier transform approach. This substitution enables accurate detection of rotation angles in circuit patterns by analyzing frequency components rather than relying on projection methods designed for document images.
2Productivity
If the image is compressed into 8*8 size and approximate angle is determined, then the calculation speed is improved, but the angle detection precision deteriorates
Solution Approach 1:
The invention transitions from spatial domain processing to frequency domain processing using Fourier transform. This dimensional change allows the method to achieve high precision angle detection without requiring large image sizes or excessive calculations, as the frequency domain representation efficiently captures rotational information.
3Measurement precision
If the image is rotated repeatedly by micro-angles to determine normal position, then the angle detection accuracy is improved, but the calculation time increases significantly
Solution Approach 1:
The invention replaces the mechanical approach of repeatedly rotating images by micro-angles with a mathematical Fourier transform method. This substitution achieves high precision angle detection in a single transformation operation, eliminating the need for multiple rotations and significantly reducing calculation time while maintaining or improving accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for accurate and relatively fast detection of rotation angles from SEM images of semiconductor circuit patterns, reducing angle detection errors even for images with widely broken-apart lines.
Implementation Method 1
a first image obtained by subjecting the photographed image to preprocessing is Fourier-transformed
Data Source
AI summary
An evaluation value indicative of the extent of lines in each direction is calculated for a pre-processed image in which 0s are filled in and extended in the lateral direction of the inputted image and which has been reduced ⅛th in the longitudinal direction. To obtain the angle of rotation of an image from the change in the evaluation value obtained while the angle relative to the lateral direction of the pre-processed image is modified in small steps, a parallel line is drawn for each direction, a projection is taken, and the sum of squares serves as the evaluation value of the direction. The direction having the highest evaluation value serves as the obtained direction of rotation from the normal position. The projection of each direction references the point of intersection between the parallel line drawn for each direction and the coordinate line of the horizontal axis.