3D Video Comfort Noise Processing for Artifact Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current 3D television displays experience visual discomfort due to keystone distortion, depth plane curvature, magnification, miniaturization effects, shear distortion, cross-talk, and lack of depth cues, as well as discomfort from differences in film grain noise between left-eye and right-eye views, with no existing techniques for adding 'comfort noise' to mitigate these issues.

Innovation Solution

A method for adding noise to 3D images by generating comfort noise and film grain based on image depth, using a comfort noise module that separates and processes video and depth information to synchronize noise between left and right views, employing additive noise models and warping techniques to ensure coherent noise across views.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If comfort noise is added to reduce visual discomfort from coding artifacts, then subjective quality improves, but no existing technique exists for 3D images

Engineering Contradiction:
Improvesubjective qualityVSAvoidapplicability to 3D images
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent segments the 3D image processing into separate left-eye and right-eye view processing channels, applying comfort noise independently to each view. This segmentation allows the adaptation of 2D comfort noise techniques to 3D images by treating each eye's view as a separate 2D image stream that can be processed individually while maintaining depth information integrity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extends 2D comfort noise addition techniques into the third dimension by incorporating depth information. It uses depth maps and disparity data to guide the noise addition process, effectively transitioning from 2D image processing to 3D-aware processing while maintaining the benefits of comfort noise for reducing coding artifacts.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Manufacturing precision

If different noise is added to left-eye and right-eye views, then coding artifacts are reduced, but visual discomfort increases due to film grain differences

Engineering Contradiction:
Improvecoding artifact reductionVSAvoidvisual discomfort from film grain differences
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The patent applies local quality by using depth information to control the characteristics of comfort noise in different regions of the image. Areas with different depth values receive differently tailored noise patterns, allowing the system to reduce coding artifacts while maintaining consistent film grain appearance across stereoscopic views by adapting noise properties to local depth characteristics.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent employs feedback mechanisms by using depth maps and disparity information to guide the comfort noise generation process. The system continuously references depth data to ensure that noise added to left and right views maintains appropriate relationships, preventing the visual discomfort caused by inconsistent film grain while preserving artifact reduction benefits.

Inventive Principle:
Principle #23Feedback

3Manufacturing precision

If comfort noise is added to improve image quality, then coding artifacts are reduced, but device complexity increases due to additional processing requirements

Engineering Contradiction:
Improveimage qualityVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent achieves universality by designing a comfort noise addition system that handles multiple functions: it processes both left and right eye views, integrates depth map processing, manages disparity compensation, and maintains film grain consistency across views. This multi-functional approach consolidates what would otherwise require separate processing chains into a unified system, reducing overall complexity while improving image quality.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent applies preliminary action by pre-processing depth information and disparity data before the comfort noise addition stage. By preparing depth maps and disparity compensation data in advance, the system reduces the computational burden during the actual noise addition process, thereby improving image quality while managing processing complexity through staged preparation of necessary data.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2569950B1Comfort noise and film grain processing for 3 dimensional video
Publication Date: 2015.07.08 THOMSON LICENSING SA
  • EP2569950B1 patent drawingFigure 1
  • EP2569950B1 patent drawingFigure 2
  • EP2569950B1 patent drawingFigure 3

AI summary

Noise, either in the form of comfort noise or film grain, is added to a three dimensional image in accordance with image depth information to reduce human sensitivity to coding artifacts, thereby improving subjective image quality.