Image Texture Enhancement via Color Variation Modulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques that aim to emphasize texture in images often introduce noise, such as coarseness or color speckles, due to the direct addition of minute amplitude components, which can damage the fine image structure and result in poor texture representation.
Innovation Solution
An image processing method that extracts local variation components from color information, produces a pseudo-texture component from these variations, and adds it to brightness information, while modulating the pseudo-texture component based on brightness or color information to minimize noise and enhance texture representation, using techniques like smoothing, difference calculation, and random number modulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If minute amplitude components are directly added to image data to emphasize texture, then texture representation is improved, but noise level increases causing coarseness and color speckles
Solution Approach 1:
The patent introduces an intermediary processing stage between extracting minute amplitude components and adding them to image data. Specifically, it uses a conversion part that transforms the variation component into a pseudo-texture component through modulation with brightness information or random numbers, and an addition part that selectively adds this processed component. This intermediary processing filters out harmful noise while preserving useful texture information, resolving the contradiction between texture enhancement and noise reduction.
Solution Approach 2:
The patent changes the parameters of the variation component through modulation operations. The conversion part modulates the variation component by multiplying it with brightness information or random number values, transforming its amplitude and frequency characteristics. This parameter transformation allows the system to emphasize texture in specific frequency bands while suppressing noise, thereby improving texture representation without increasing visible noise levels.
2Object-affected harmful factors
If smoothing is applied to reduce noise, then noise level is reduced, but fine image structure and texture are damaged
Solution Approach 1:
The patent segments the image processing into distinct functional parts: a variation extraction part that isolates texture information, a conversion part that processes this information, and an addition part that reintroduces it. By segmenting the processing pipeline, the system can apply smoothing selectively to certain components while preserving fine structures in others, resolving the contradiction between noise reduction and structure preservation.
Solution Approach 2:
The patent extracts the variation component from the original image data using a variation extraction part that calculates differences between the image and a smoothed version. This extraction separates texture information from noise, allowing subsequent processing to enhance texture without affecting the fine image structure that would be damaged by aggressive smoothing. The extracted variation component can then be processed and added back in a controlled manner.
Data Source
AI summary
An image processing device extracts a local variation component from color information in image data. Using the variation component of the color information, a pseudo-texture component of brightness information is produced. The thus produced pseudo-texture component is added to the brightness information. By this image processing, an image having improved texture is produced.