Stereoscopic Image Matching Brightness Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for stereoscopic image matching, such as those using brightness values, face challenges in achieving sufficient matching accuracy due to differences in brightness caused by lens conditions and imager errors.
Innovation Solution
An image processing apparatus that extracts sub-images from stereoscopic images, compares their brightness, and excludes pairs outside a predetermined brightness range to enhance matching accuracy, while also correcting brightness variations and outputting information on abnormal brightness ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If brightness values are used for matching stereoscopic images, then matching can be performed, but matching accuracy is insufficient due to brightness differences caused by lens conditions and imager errors
Solution Approach 1:
The patent applies preliminary action by performing brightness comparison and exclusion of abnormal sub-image pairs before the actual matching process. The brightness comparing unit compares brightness of sub-image pairs in advance, and the exclusion unit removes pairs with abnormal brightness ratios before matching is performed, preventing brightness differences from degrading matching accuracy
Solution Approach 2:
The patent applies parameter changes by using brightness ratio as a filtering parameter to identify and exclude abnormal sub-image pairs. By calculating the brightness ratio between corresponding sub-images and comparing it against a predetermined threshold, the system dynamically filters out pairs with abnormal brightness characteristics, thereby improving the reliability of subsequent matching operations
2Measurement precision
If all sub-image pairs are matched regardless of brightness differences, then matching coverage is complete, but matching accuracy deteriorates due to inclusion of abnormal brightness pairs
Solution Approach 1:
The patent applies the taking out principle by extracting and removing abnormal sub-image pairs from the matching process. The exclusion unit identifies pairs with abnormal brightness ratios and extracts them from the set of pairs to be matched, preventing them from degrading overall matching accuracy while allowing normal pairs to be processed efficiently
Data Source
AI summary
In an image processing apparatus, an image acquiring unit acquires a first image and a second image that form stereoscopic images. A first sub-image extracting unit extracts first sub-images from the first image. A second sub-image extracting unit extracts second sub-images from the second image. A matching unit matches each pair of the first and second sub-images to determine a degree of similarity therebetween. A similar sub-image setting unit sets the second sub-image having a highest degree of similarity to the first sub-image to be a similar sub-image to the first sub-image. A brightness comparing unit compares in brightness each pair of the first and second sub-images. The matching unit is configured to, if a result of comparison made by the brightness comparing unit between a pair of the first and second sub-images is out of a predetermined brightness range, exclude such a pair of the first and second sub-images.