Template Matching Using Coarse and Fine Search Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional template matching techniques require extensive calculations to handle rotation, scaling, and deformation, making them inefficient.
Innovation Solution
An image processing device and method that transforms a template into multiple extended templates using various values for rotation, scaling, and deformation, allowing for a first coarse search and a second fine search, reducing the number of comparison operations by selecting templates based on specific boundaries and detection results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple extended templates are generated to handle rotation, scaling, and deformation, then pattern detection capability is improved, but calculation amount increases extremely
Solution Approach 1:
The patent segments the template extension process into two distinct phases: a first search using a subset of extended templates to identify candidate regions, and a second search using additional extended templates to refine detection in those specific regions. This segmentation avoids the need to compare all extended templates against the entire image, significantly reducing calculation while maintaining detection capability.
Solution Approach 2:
The patent performs preliminary action by conducting the first search before the second search. The first search uses a smaller set of extended templates to pre-identify candidate regions where the target pattern may exist. This preliminary identification allows the second search to focus computational resources only on relevant regions, rather than performing exhaustive comparisons across the entire image with all extended templates.
2Measurement precision
If all extended templates are compared with the image, then accurate pattern detection is achieved, but processing time increases
Solution Approach 1:
The patent divides the template matching process into two sequential searches: a first search that quickly identifies candidate regions using a subset of extended templates, and a second search that performs detailed matching in those candidate regions. This segmentation maintains detection accuracy by ensuring thorough comparison in relevant areas while avoiding time-consuming comparisons in irrelevant areas.
Solution Approach 2:
The patent applies partial action by performing complete template comparisons only in candidate regions identified by the first search, rather than performing exhaustive comparisons across the entire image. The first search uses a limited set of extended templates to identify where full comparisons are necessary, allowing the second search to focus computational effort partially on specific regions, thereby reducing overall processing time while maintaining accuracy.
Data Source
AI summary
An image processing device includes a template analyzer to transform a pattern included in a template into multiple transformed patterns and generate multiple extended templates consisting of the template and multiple transformed templates including the respective transformed patterns. Further, there is a search processor to perform a first search that searches a to-be-searched image by using multiple first extended templates selected from the multiple extended templates, and perform a second search that searches the to-be-searched image by using multiple second extended templates selected from the multiple extended templates on a basis of a result of the first search. The first search is performed by using the template and two or more transformed templates selected at coarse boundaries from the multiple extended templates, and the second search is performed by using two or more extended templates near one of the extended templates that has been detected in the first search.