Image Matching With Chroma Features and Adaptive Patch Sizes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image matching techniques face challenges in low-light conditions due to low resolution and noise, leading to ambiguity and errors, and inefficiencies arise from using fixed feature sizes that are not adaptable to varying image resolutions.
Innovation Solution
The method involves converting images to a CIELAB color space for chroma-enhanced features and adjusting feature sizes based on image resolution, using adaptable feature sizes for improved accuracy and reduced resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed feature sizes are used for image matching, then the processing pipeline is simple, but the accuracy deteriorates when image resolution varies
Solution Approach 1:
The patent implements dynamic feature size adjustment where the feature size is adaptively determined based on the input image resolution. The system automatically selects appropriate feature sizes (e.g., 7x7, 15x15, or 31x31 patches) according to the image dimensions, transforming the static fixed-size approach into a dynamic adaptive one that maintains accuracy across varying resolutions without significantly increasing pipeline complexity
2Measurement precision
If chroma-enhanced features are used, then feature detection accuracy is improved, but computational resources are increased
Solution Approach 1:
The patent applies local quality enhancement by converting only the relevant color channels (chroma channels a* and b* in CIELAB color space) rather than processing the entire image uniformly. This selective chroma enhancement focuses computational resources on the specific channels that provide the most benefit for feature detection, improving accuracy while controlling overall computational cost
Solution Approach 2:
The system changes the color space parameter from standard RGB to CIELAB color space, which provides perceptually uniform color representation. This parameter transformation enhances the chromatic information in a way that is more aligned with human vision, improving feature detection accuracy without requiring excessive computational resources
3Measurement precision
If feature size is increased for higher resolution images, then feature matching accuracy is improved, but processing time increases
Solution Approach 1:
The patent implements dynamic feature size selection where smaller feature sizes (e.g., 7x7) are used for low-resolution images and larger feature sizes (e.g., 31x31) are used for high-resolution images. This adaptive approach ensures that processing time is proportional to the image resolution requirements, maintaining accuracy without unnecessary computational overhead
Data Source
AI summary
Certain aspects of the present disclosure provide techniques for feature-based image matching. A method generally includes obtaining a first image frame and a second image frame; generating a modified first image frame and a modified second image frame, wherein to generate the modified first image frame and the modified second image frame comprises to convert a first color space of the first image frame and a second color space of the second image frame to a third color space; extracting a first plurality of features from the modified first image frame; extracting a second plurality of features from the modified second image frame; and determining at least one first matching cost based on the first plurality of features and the second plurality of features.