Image Processor Common Curve Generation for Dark Image Brightening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies fail to effectively improve image quality, particularly in brightening dark images while minimizing color changes, especially for memory colors like skin tones, which are difficult to view and prone to unnatural lightening.
Innovation Solution
An image processor that generates a common gradation correction curve based on the average luminance of an entire image and a predetermined-colored region, using a weighting factor to balance the luminance correction, ensuring natural brightening of dark areas without significantly altering bright areas, thereby improving image quality and maintaining color accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a γ correction is performed based on display characteristic alone, then the display compatibility is improved, but the image quality for dark images deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the correction approach based on image content. Dark images are identified and processed with a different γ curve (first γ curve) compared to normal images (second γ curve). This allows the system to adapt the correction characteristics to the specific luminance conditions of the input image, improving image quality for dark content while maintaining compatibility for normal content.
Solution Approach 2:
The patent changes the γ correction parameter dynamically based on the luminance characteristics of the input image. By calculating the average luminance and comparing it against a threshold, the system selects between different γ curves, effectively changing the correction parameters to suit the image content. This resolves the contradiction by allowing parameter adaptation rather than using a fixed correction curve.
2Illumination intensity
If the entire image is brightened to improve viewability, then the visibility of dark images is improved, but the memory colors become unnaturally lightened
Solution Approach 1:
The patent applies local quality by using different γ curves for different luminance ranges. The first γ curve is specifically designed for dark images to provide appropriate brightening, while the second γ curve is designed for normal images to preserve color accuracy. This localized approach ensures that dark images are brightened without causing unnatural lightening of memory colors in normal images.
Solution Approach 2:
The patent applies partial action by selectively applying strong correction only where needed. For dark images, a stronger correction (first γ curve) is applied to improve visibility, while for normal images, a milder correction (second γ curve) is applied to maintain color fidelity. This selective application prevents excessive correction that would cause unnatural color lightening.
Data Source
AI summary
An image processor includes a curve generator and a gradation correcting unit. Based on a predetermined correspondence relationship between average luminance of an image and a gradation correction curve, the curve generator generates, in accordance with average luminance of an input image, a common curve, i.e. a gradation correction curve used to correct a gradation of the input image. The gradation correcting unit uses in common the common curve generated by the curve generator for red, green, and blue color signals of the input image to correct gradations of the red, green, and blue color signals. Based on first average luminance, i.e. an average value of luminance of a first region in the input image, second average luminance, i.e. an average value of luminance of a predetermined-colored second region included in the first region, and an area of the second region, the curve generator generates a common curve.