Sparkle Processing for Natural Highlight Representation

Solution Overview

Problem

Existing image processing technologies often result in unnatural, artificially compressed bright areas due to dynamic range limitations, leading to a loss of natural representation in images, especially when displaying high-luminance regions like car headlights and reflections.

Innovation Solution

An image conversion unit that selectively increases the luminance values of small, high-luminance areas using a high-pass filter and gain control, ensuring that only bright, localized regions like highlights are enhanced, while maintaining original luminance values in larger and darker areas, thereby recreating a more natural 'sparkle' effect.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If dynamic range compression is applied to cameras to handle high luminance range, then the camera can capture images with high dynamic range, but the bright areas are strongly compressed resulting in artificial representations

Engineering Contradiction:

Improveluminance rangeVSAvoidnatural representation

Core Design Contradiction:

Illumination intensityVSManufacturing precision

Solution Approach 1:

The patent applies local quality by selectively enhancing only small regions with high luminance values (highlights) while leaving the rest of the image unchanged. The filter identifies localized bright areas and applies gain control only to those regions, preserving natural representation in those specific areas without over-brightening the entire image.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent segments the image into different luminance regions by using a high-pass filter to identify and separate small high-luminance areas from the rest of the image. This segmentation allows independent processing of highlights versus midtones and shadows, enabling natural representation in the highlighted regions while maintaining overall image quality.

Inventive Principle:

Principle #1Segmentation

2Adaptability or versatility

If contrast reduction is applied to images, then the dynamic range is limited for display, but the bright areas are strongly compressed

Engineering Contradiction:

Improvedisplay compatibilityVSAvoidbright area luminance

Core Design Contradiction:

Adaptability or versatilityVSIllumination intensity

Solution Approach 1:

The patent applies local quality by selectively enhancing only small regions with high luminance values (highlights) while leaving the rest of the image unchanged. The filter identifies localized bright areas and applies gain control only to those regions, preserving natural representation in those specific areas without over-brightening the entire image.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent applies partial action by enhancing only a portion of the image (the highlights) rather than the entire image. The high-pass filter identifies and processes only the small high-luminance regions, applying gain control selectively to those areas while leaving midtones and shadows unaffected.

Inventive Principle:

Principle #16Partial or excessive action

3Illumination intensity

If the entire image luminance is increased to enhance highlights, then the bright areas become more visible, but the average luminance increases excessively

Engineering Contradiction:

Improvehighlight brightnessVSAvoidaverage luminance

Core Design Contradiction:

Illumination intensityVSQuantity of substance

Solution Approach 1:

The patent applies local quality by selectively enhancing only small regions with high luminance values (highlights) while leaving the rest of the image unchanged. The filter identifies localized bright areas and applies gain control only to those regions, preserving natural representation in those specific areas without over-brightening the entire image.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent segments the image into different luminance regions by using a high-pass filter to identify and separate small high-luminance areas from the rest of the image. This segmentation allows independent processing of highlights versus midtones and shadows, enabling natural representation in the highlighted regions while maintaining overall image quality.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS8068691B2Sparkle processing

Publication Date: 2011.11.29 TOP VICTORY INVESTMENTS LTD

AI summary

An image conversion unit (100) for converting an input image (602) into an output image (606) is disclosed. The image conversion unit (100) comprises: a filter (104) for selecting a group of connected input pixels from the input image (602), the group of connected input pixels having a size which is relatively small relative to the size of the input image (602), the input pixels of the group of connected input pixels having high luminance values relative to the range of input luminance values; and a pixel processor (104) for assigning luminance values to output pixels of the output image (606) on basis of the respective luminance values of the input pixels of the input image (602), whereby the luminance values of the output pixels corresponding to the group of connected input pixels are higher than the respective luminance values of the connected input pixels.