Chroma Subsampling Optimization for Image Artifact Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Chroma subsampling in image compression often results in visually disturbing artifacts, particularly in images with text, striped patterns, and computer-generated graphics, due to the reduction in color information while retaining brightness resolution.

Innovation Solution

A method for optimizing chroma subsampling by measuring quality metrics for input and reconstructed images, calculating error values, and iteratively modifying the subsampled representation to reduce visual artifacts, ensuring the final image meets specific quality thresholds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If chroma subsampling is applied to reduce data amount, then data compression ratio is improved, but visual quality deteriorates due to artifacts in text and graphic images

Engineering Contradiction:
Improvedata amountVSAvoidvisual quality
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies different chroma subsampling strategies to different regions of the image based on content analysis. Text regions, graphic regions, and natural image regions are identified and processed differently, with more aggressive subsampling applied to natural regions where artifacts are less noticeable, and reduced or no subsampling applied to text and graphic regions where quality must be preserved.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts chroma subsampling parameters based on image content characteristics. By analyzing local image features such as edge density, color variation, and pattern complexity, the system modifies subsampling ratios and reconstruction filter parameters to optimize the balance between compression and visual quality for different content types.

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If chroma subsampling reduces spatial resolution of color information, then data compression is improved, but color accuracy deteriorates

Engineering Contradiction:
Improvedata compressionVSAvoidcolor accuracy
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent performs preliminary analysis of the image content before applying chroma subsampling to determine appropriate processing strategies. By pre-identifying regions with text, graphics, or smooth gradients, the system can prepare region-specific subsampling parameters and reconstruction approaches that preserve color accuracy where needed while achieving compression where permissible.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent employs feedback mechanisms where the reconstructed image quality is evaluated and used to adjust subsampling parameters. Quality metrics are computed based on the reconstructed output, and this feedback is used to iteratively refine the subsampling strategy, ensuring color accuracy is maintained within acceptable thresholds while maximizing compression.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3180910B1Method for optimized chroma subsampling, apparatus for optimized chroma subsampling and storage device
Publication Date: 2021.09.15 GOOGLE LLC
  • EP3180910B1 patent drawingFigure 1~2
  • EP3180910B1 patent drawingFigure 3
  • EP3180910B1 patent drawingFigure 4

AI summary

A method (400) includes obtaining (410), at one or more computing devices, an input image; determining (420) a first value of a quality metric for the input image; generating (430) a first chroma subsampled representation of the input image; and generating (440) a reconstructed image based on the chroma subsampled representation. The method also includes determining a second value of the quality metric for the reconstructed image; determining an error value based on the first value of the quality metric and the second value of the quality metric; and generating a second chroma subsampled representation of the input image based in part on the error value.