Color Invariant Feature Matching for Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition algorithms, such as SIFT, SURF, MSER, BRISK, and FREAK, fail to accurately recognize objects when color is involved, leading to false matches due to conversion to grayscale, which cannot distinguish objects of different colors with similar grayscale values.
Innovation Solution
A processor-implemented method using color invariant feature matching, which identifies interest points, determines color values for each pixel in primary color channels, calculates a color difference factor, and determines a false point value to accurately match objects based on color differences, with threshold values to prevent false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If object recognition algorithms convert color images to grayscale for processing, then the algorithms can process images using existing grayscale-based methods, but objects of different colors with similar grayscale values cannot be distinguished, leading to false matches
Solution Approach 1:
The patent changes the parameter space from grayscale to color space by introducing color difference factors (CDF) that operate on color channels. Instead of converting color to grayscale, the system maintains color information and uses CDF to measure color differences between corresponding pixels in template and actual images, enabling accurate discrimination of objects with different colors.
2Device complexity
If algorithms use grayscale conversion for object recognition, then computational complexity is reduced, but color information is lost causing inability to discriminate graphical objects with same shape but different colors
Solution Approach 1:
The patent introduces color difference factor (CDF) as an intermediary metric that bridges template matching and color discrimination. CDF quantifies color differences between corresponding pixels while maintaining the overall structure of template matching, allowing the system to incorporate color information without completely redesigning the recognition algorithm.
3Productivity
If existing algorithms compare only grayscale values, then processing is simpler and faster, but false positives occur when different colored objects have similar grayscale equivalents
Solution Approach 1:
The patent applies partial color analysis by computing color difference factors only for inlier pixels (those that match the template within a certain threshold) rather than all pixels. This selective approach maintains processing efficiency while improving reliability by focusing color comparison on relevant regions where actual color differences matter for discrimination.
Data Source
AI summary
A method of verifying graphical objects using color invariant feature matching is provided. The method comprises determining, for each of a plurality of primary color channels, a color value for each inlier pixel in an actual object for that primary color channel; determining, for each color channel, a color difference factor (CDF) for an expected object by comparing, for each of the plurality of primary color channels, the color value for each inlier pixel in the expected object to the color value for each corresponding inlier pixel in the actual object; determining a false point value for the expected object from the CDF; and determining that the expected object matches the actual object when the false point value for the expected object is lower than the false point value for other expected objects.