Image Feature Point Association Using Positional Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in accurately associating feature points across multiple images, particularly when repeating patterns are present, leading to erroneous correlations due to similar points having high similarity scores at fixed intervals.
Innovation Solution
An image processing device and method that extracts feature points from time-series images, sets a search range for corresponding points, and associates them based on positional relationships, with a three-dimensional point group estimation unit and obstacle estimation unit to prevent erroneous associations by identifying and stopping the association process when a multi-peak property in the score distribution is detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If feature points are associated based on similarity scores alone, then matching speed is improved, but erroneous associations occur in images with repeating patterns
Solution Approach 1:
The patent changes the parameter basis for association from solely similarity scores to a composite of similarity scores and positional relationships. Specifically, it introduces distance constraints and relative position calculations to filter out erroneous matches in repeating patterns while maintaining efficient processing
Solution Approach 2:
The patent introduces positional relationship analysis as an intermediary mechanism between feature extraction and final association. This intermediary layer checks whether candidate points satisfy distance and position constraints before confirming associations, thereby preventing erroneous matches without significantly increasing computational complexity
2Measurement precision
If a comprehensive search range is used to find corresponding points, then matching completeness is improved, but computational complexity increases
Solution Approach 1:
The patent segments the search process by first identifying candidate points within a comprehensive search range, then applying positional relationship constraints to filter and validate these candidates. This two-stage approach ensures complete coverage while managing computational load through structured processing
Solution Approach 2:
The patent performs preliminary filtering of candidate points based on positional relationships before final association confirmation. By pre-checking distance constraints and relative positions, it reduces the number of points requiring full similarity analysis, thereby balancing completeness with computational efficiency
Data Source
AI summary
According to one or more embodiments, an image processing device includes a storage device and a processor. The storage device stores a plurality of time-series images. The processor extracts one or more feature points of a first image of the time-series images, sets a search range for one or more corresponding points of the feature points in a second image of the time-series images, searches for the one or more corresponding points in the search range of the second image, if there are a plurality of corresponding points searched, associates the searched corresponding points with the feature points based on a positional relationship between the corresponding points.