Template Matching Apparatus for Periodic Pattern Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing template matching techniques face challenges in accurately detecting the correct position of a pattern in images with periodic structures, leading to erroneous recognition due to morphology changes and noise, especially when similar patterns are shifted by one or two periods, and struggle with poor signal-to-noise ratios.
Innovation Solution
A template matching apparatus and method that selects a template area, detects similar images, forms difference images, modifies the template to penalize similar images, and evaluates similarity degrees using a predetermined formula, thereby reducing erroneous recognition by using a modified template for evaluation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If template matching is performed on images with periodic structures, then pattern detection can be conducted, but erroneous recognition occurs when similar patterns are shifted by one or two periods
Solution Approach 1:
The patent segments the template into multiple sub-templates and performs matching separately for each sub-template. This segmentation allows the system to distinguish between the correct pattern position and periodic similar patterns by comparing results across multiple segments, thereby reducing erroneous recognition while maintaining detection accuracy.
Solution Approach 2:
The patent applies different evaluation criteria and weighting to different regions of the template based on their local characteristics. By emphasizing unique local features and suppressing regions prone to periodic confusion, the system achieves both accurate pattern detection and reliable discrimination against erroneous matches.
2Adaptability or versatility
If template matching is performed with morphology changes and noise, then generalization is achieved, but signal-to-noise ratio deteriorates
Solution Approach 1:
The patent performs preliminary preprocessing operations including noise filtering, morphology normalization, and feature enhancement before template matching. These preliminary actions prepare the input image and template to be more robust against noise and morphology variations, improving both adaptability and signal-to-noise ratio.
Solution Approach 2:
The patent dynamically adjusts matching parameters such as similarity thresholds, evaluation weights, and tolerance levels based on the observed morphology changes and noise characteristics. This adaptive parameter adjustment maintains high adaptability to different patterns while compensating for noise interference in real-time.
3Measurement precision
If strict template matching is applied, then detection precision is maintained, but false negatives increase due to minor variations
Solution Approach 1:
The patent implements dynamic threshold adjustment and adaptive evaluation criteria that respond to the complexity and characteristics of the input image. This dynamic approach maintains strict precision requirements for clear matches while becoming more tolerant of minor variations in complex or noisy scenarios, thereby reducing false negatives without sacrificing detection precision.
Data Source
AI summary
A similar image having a high correlation is selected through autocorrelation performs a template original image selected from an image photographed for a template, and a difference image between the similar image and template original image is formed. An image extracting a real difference is formed by removing noises and edges in unstable areas from the difference image. This image is added to the template original image to form a modified template. Template matching is performed by using the modified template as a template. The image extracting the real reference and added to the modified template functions to add an evaluation penalty to the similar image during matching evaluation to lower an evaluation value of the similar image so that a probability of erroneously recognizing the similar image as the image to be detected.