Image Processing Apparatus Detail Clarity Restoration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques using self-similarity struggle to effectively restore detail clarity, especially when enlarging images, as they can render thin lines as thick, delete fine patterns, and cause discontinuity at block borders, and adjusting bandwidth can lead to degradation and loss of detail clarity.
Innovation Solution
An image processing apparatus that acquires matching taps and prediction taps around a pixel of interest, identifies similarity maximizing pixels, and computes pixel values by mixing computed values using a pixel value mixing control unit, which adjusts blend ratios based on similarity and stationarity to restore image detail clarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If existing self-similarity processing is used to improve image quality, then image continuity in boundary portions is improved, but thin lines are rendered as thick lines and fine patterns are deleted or degraded
Solution Approach 1:
The patent applies local quality by differentiating processing based on local image characteristics. The stationarity determination unit identifies whether each region contains fine patterns (non-stationary) or smooth areas (stationary), and the pixel value mixing control unit adjusts blend ratios accordingly - using lower blend ratios for non-stationary regions to preserve fine patterns while using higher blend ratios for stationary regions to improve continuity.
2Illumination intensity
If bandwidth adjustment is performed to increase sharpness, then image sharpness is improved, but detail clarity is lost and degradation becomes noticeable
Solution Approach 1:
The patent implements dynamics by making the processing parameters adaptive rather than fixed. The blend ratio is dynamically adjusted based on local stationarity characteristics detected in each region, allowing the system to optimize between sharpness and detail preservation for each local area rather than applying uniform processing across the entire image.
3Productivity
If pixels are replaced on a block basis to improve processing efficiency, then processing productivity is improved, but discontinuity appears in border portions of blocks
Solution Approach 1:
The patent merges the benefits of block-based processing with continuous boundary handling. While processing is organized in blocks for efficiency, the pixel value mixing control unit applies blending at block boundaries using similarity maximizing pixels from adjacent blocks, seamlessly merging the processed regions to eliminate visible discontinuities while maintaining processing efficiency.
Data Source
AI summary
An image processor includes a first acquirer that acquires a matching tap having a target pixel at the center, a second acquirer that acquires plural matching taps each having, at the center, one of pixels in a search area including the pixels surrounding the target pixel, a similarity identifier that identifies, among the matching taps acquired by the second acquirer, a similarity maximizing pixel representing a central pixel of the matching tap having the highest similarity to the matching tap acquired by the first acquirer, and a mixing controller that computes a pixel value of the target pixel by mixing a pixel value obtained by performing a predetermined arithmetic process on pixel values of a prediction tap having the target pixel at the center with a pixel value obtained by performing a predetermined arithmetic process on pixel values of a prediction tap having the similarity maximizing pixel at the center.