Image Matching via Color Histogram Bin Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image comparison techniques are prone to errors when dealing with image distortion, such as noise, cropping, or combining images, leading to false matches or missed detections.
Innovation Solution
The method uses color histograms to compare images by generating and adjusting bin values, determining a comparison metric based on differences between the histograms, and identifying matches based on a threshold, allowing for reliable matching even in the presence of distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional image comparison techniques are used, then the comparison process is simple, but the accuracy deteriorates in the presence of image distortion
Solution Approach 1:
The patent transforms images into color histogram representations, changing the parameter space from pixel values to color distribution frequencies. This transformation makes the comparison robust to distortion because color histograms capture the essential color content while being invariant to geometric transformations like cropping, rotation, and noise addition.
Solution Approach 2:
The patent introduces color histograms as an intermediary representation between the original images and the comparison process. Instead of directly comparing distorted images, the method compares their color histogram signatures, which serve as a mediator that preserves matching information while being resilient to distortion.
2Measurement precision
If color histograms are used for image comparison, then the accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent segments the color space into discrete bins, transforming continuous color values into discrete histogram bins. This segmentation simplifies the computation by reducing the infinite continuous color space to a finite number of bins, making histogram generation and comparison computationally efficient while maintaining precision.
Solution Approach 2:
The patent replaces complex pixel-level image processing with histogram-based statistical analysis. Instead of performing computationally intensive operations on individual pixels, the method uses histogram computations that aggregate color information, significantly reducing computational complexity while preserving matching accuracy.
3Reliability
If bin values are adjusted using scale factor and offset, then the robustness to distortion improves, but the processing time increases
Solution Approach 1:
The patent performs preliminary normalization of color histograms by adjusting bin values using scale factors and offsets before comparison. This preliminary action prepares the histograms to accommodate variations in lighting, exposure, and color balance, making the subsequent comparison more robust to distortion without requiring complex real-time adjustments during the matching process.
Data Source
AI summary
Methods, apparatus and articles of manufacture for image comparison using color histograms are disclosed. An example method disclosed herein to compare a first image and a second image comprises obtaining a first color histogram for a first set of pixels sampled from the first image, obtaining a second color histogram for a second set of pixels sampled from the second image, determining a comparison metric based on differences between bin values of the first color histogram and adjusted bin values of the second color histogram, and determining whether the first image and the second image match based on the comparison metric.