Depth Map Coding Using Region-Growing Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for encoding depth information in 3D video applications are inefficient, leading to high bit rates and artifacts around sharp edges, which complicates the rendering of virtual views in 3D television and free viewpoint video applications.

Innovation Solution

A region-growing mode for depth map coding is introduced, which uses neighboring depth values and structural similarity between depth maps and video images to segment and encode depth information, reducing bit rates while maintaining quality by estimating depth values accurately and preserving depth edges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If conventional depth encoding methods are used, then encoding simplicity is maintained, but bit rate increases and artifacts appear around sharp edges

Engineering Contradiction:
Improvebit rateVSAvoidencoding complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the depth map into multiple regions based on depth values and corresponding video image features. Different encoding strategies are applied to different segments: regions with sharp edges use one encoding approach while smooth regions use another, thereby reducing overall bit rate while maintaining quality and avoiding artifacts around edges.

Inventive Principle:
Principle #1Segmentation

2Reliability

If conventional encoding is used, then encoding process is simple, but rendering quality of virtual views deteriorates

Engineering Contradiction:
Improverendering qualityVSAvoidencoding process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements local quality by adapting the encoding precision and method to local characteristics of the depth map. Regions containing sharp edges or important features are encoded with higher quality and more detail, while smooth regions use coarser encoding. This localized approach improves virtual view rendering quality where it matters most without unnecessarily increasing overall encoding complexity.

Inventive Principle:
Principle #3Local quality

3Quantity of substance

If depth information is compressed aggressively, then storage efficiency improves, but artifacts around sharp edges increase

Engineering Contradiction:
Improvedata sizeVSAvoidartifacts
Core Design Contradiction:
Quantity of substanceVSObject-affected harmful factors

Solution Approach 1:

The patent employs dynamic encoding where the compression level and method are adaptively adjusted based on local depth variations and corresponding video content. In regions with sharp edges or high frequency content, the encoder dynamically reduces compression to preserve edge integrity and avoid artifacts. In smooth regions, aggressive compression is applied to reduce data size. This dynamic adaptation resolves the contradiction between compression efficiency and artifact prevention.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP2748794B1Depth coding
Publication Date: 2017.07.12 THOMSON LICENSING SA
  • EP2748794B1 patent drawingFigure 1~2
  • EP2748794B1 patent drawingFigure 3~5
  • EP2748794B1 patent drawingFigure 6~7

AI summary

Various implementations address depth coding and related disciplines. In one particular implementation, a segmentation is determined for a particular portion of a video image in a sequence of video images. The segmentation is determined based on reference depth indicators that are associated with at least a portion of one video image in the sequence of video images. Target depth indicators associated with the particular portion of the video image are processed. The processing is based on the determined segmentation in the particular portion of the video image. In another particular implementation, a segmentation is determined for at least a given portion of a video image based on depth indicators associated with the given portion. The segmentation is extended from the given portion into a target portion of the video image based on pixel values in the given portion and on pixel values in the target portion.