Color Calibration Outlier Detection Histogram Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing color calibration methods, such as histogram matching, suffer from artifacts and noise when source and reference images are not perfectly aligned, leading to decreased image quality and increased noise, especially with large exposure differences.
Innovation Solution
An apparatus and method that includes a histogram matching unit, probability distribution computation unit, outlier detection unit, and outlier correction unit to improve color calibration by generating a conditional probability distribution of the reference image, detecting outliers based on this distribution, and correcting pixel intensities to reduce artifacts and noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If histogram matching is performed to improve color calibration between images, then color similarity is improved, but artifacts and noise increase due to mismatches from camera or object motion
Solution Approach 1:
The patent applies preliminary action by performing outlier detection and correction before final histogram matching. The method identifies and corrects mismatched pixels in advance by comparing histograms of overlapping regions and correcting outliers based on statistical analysis, thereby preventing artifacts and noise from being introduced during the color calibration process
Solution Approach 2:
The patent implements feedback by using the detected outliers to refine the histogram matching process. The method calculates the difference between histograms of overlapping regions, identifies outliers based on statistical thresholds, and uses this feedback information to correct pixel values before completing the color calibration, thus improving overall accuracy while reducing artifacts
2Manufacturing precision
If histogram matching is performed to match color histograms between images, then color distribution similarity is improved, but image noise increases due to brightening dark regions
Solution Approach 1:
The patent applies local quality by treating different regions of the image differently during histogram matching. The method identifies dark regions and applies selective outlier correction only where needed, rather than uniformly processing all pixels. This localized approach maintains histogram matching accuracy while minimizing noise amplification in specific regions
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting the outlier detection thresholds and correction parameters based on local image characteristics. The method modifies statistical parameters such as mean and standard deviation of histogram values in different regions, allowing adaptive noise control while maintaining overall color calibration accuracy
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The present invention relates to an apparatus and a method for mapping the colors of at least one source image to the colors of a reference image. The apparatus comprises a histogram matching unit (52) adapted to match the histogram of the source image to the histogram of the reference image so as to generate a histogram matched image of the source image, a probability distribution computation unit (54) adapted to generate a conditional probability distribution of the reference image, and an outlier detection unit (55) adapted to detect outliers in the histogram matched image on the basis of the conditional probability distribution.