Image Haze Correction with Evaluation Maps for Brightness and Color
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing haze correction techniques can result in images becoming too dark and unnatural in color saturation after haze removal, especially in scenes shot under appropriate brightness, and do not adequately consider the relationship among color components for saturation correction.
Innovation Solution
An image processing apparatus that generates an evaluation image and transmission map for haze correction, applies haze correction based on the map, and adjusts brightness or saturation based on the change in the evaluation image before and after haze correction, without altering a predetermined color component.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If haze correction is carried out using a known technique, then haze removal effect is achieved, but the image becomes dark and underexposed
Solution Approach 1:
The patent applies parameter changes by adjusting the strength or intensity of the haze correction processing. By controlling the degree of correction applied, the system achieves effective haze removal while preventing excessive darkening of the image, thus resolving the contradiction between haze removal effectiveness and image brightness preservation.
2Reliability
If haze correction is carried out using a known technique, then haze removal effect is achieved, but the colors become unnatural and saturation is excessive
Solution Approach 1:
The patent uses parameter changes to control the saturation level during haze correction. By adjusting the correction strength parameter, the system achieves effective haze removal while maintaining natural color saturation levels, avoiding the unnatural appearance that occurs with standard haze correction techniques.
3Ease of manufacture
If enhancement suppression region is defined to suppress haze correction intensity, then color naturalness is improved, but the haze removal effect decreases
Solution Approach 1:
The patent applies dynamics by making the haze correction intensity adaptive rather than static. The correction strength is dynamically adjusted based on image characteristics and regions, allowing the system to maintain natural colors in certain areas while preserving strong haze removal effects in others, thus resolving the contradiction between color naturalness and haze removal effectiveness.
Data Source
AI summary
An evaluation image generation unit generates, based on an image, an evaluation image for haze correction. A transmission map generation unit generates, based on the evaluation image, a transmission map for the haze correction. A first correction unit applies the haze correction that is based on the transmission map to the evaluation image. A second correction unit applies the haze correction that is based on the transmission map to the image. A third correction unit corrects, based on an amount of change in the evaluation image before and after the application of the haze correction, brightness of the image to which the haze correction has already been applied.