Stereo Matching Accuracy via Adaptive Algorithm Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing stereo matching technologies face challenges in accurately obtaining depth information from two-dimensional images displayed on three-dimensional display devices, as they struggle to accurately search matching objects between images with different views, leading to potential errors in disparity information and incorrect depth perception.
Innovation Solution
The method involves performing stereo matching computations using two different algorithms (dynamic programming and scan-line optimization) on image blocks to generate multiple matching results, calculating matching errors and similarities, and determining the most accurate result based on error and similarity thresholds, ensuring the selection of an accurate stereo matching result adapted to object characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single matching algorithm is used for stereo matching, then the device complexity is low, but the matching accuracy is insufficient due to inability to adapt to different object characteristics
Solution Approach 1:
The system dynamically selects between different matching algorithms (dynamic programming and scan-line optimization) based on the characteristics of the image block being processed. The determination module chooses the appropriate algorithm adaptively, making the system flexible and responsive to different object types rather than using a fixed single algorithm approach.
Solution Approach 2:
The system changes the algorithm parameter (which matching algorithm to use) based on the input image characteristics. By adjusting the algorithm selection parameter according to object features, the system achieves high accuracy across different scenarios without requiring a complex unified algorithm.
2Measurement precision
If multiple matching algorithms are used to improve accuracy, then the matching precision is improved, but the computation time increases
Solution Approach 1:
The image is divided into multiple image blocks, and different matching algorithms are applied to different blocks based on their characteristics. This segmentation allows the system to use computationally intensive algorithms only where necessary while using faster algorithms for other regions, reducing overall computation time while maintaining accuracy where needed.
Solution Approach 2:
The system applies multiple algorithms only partially - specifically, it uses two different algorithms (dynamic programming and scan-line optimization) only for determining the final stereo matching result, rather than applying multiple algorithms to every step of the processing pipeline. This partial application reduces unnecessary computational overhead.
3Loss of information
If matching objects are searched between images with different views, then depth information can be obtained, but matching errors occur when landscapes are not entirely the same
Solution Approach 1:
The system uses a determination module that evaluates the matching results from multiple algorithms and selects the most reliable one. This feedback mechanism compares the outputs of different algorithms and uses error analysis to identify and correct matching errors, improving the reliability of depth information extraction from stereo images with different landscapes.
Solution Approach 2:
The system combines the results from multiple matching algorithms (dynamic programming and scan-line optimization) to produce a final stereo matching result. By compositeing the strengths of different algorithms, the system achieves more reliable matching accuracy even when images have different landscapes or object characteristics.
Data Source
AI summary
An image matching method is utilized for performing a stereo matching from a first image block to a second image block in a stereo matching system. The image matching method includes performing a matching computation from the first image block to the second image block according to a first matching algorithm to generate a first matching result; performing the matching computation between the first image block and the second image block according to a second matching algorithm to generate a second matching result and a third matching result; obtaining a matching error and a matching similarity of the first image block according to the second matching result and the third matching result; and determining a stereo matching result of the first image block according to the matching error and the matching similarity.