Multi-Layer RGBD Video Encoding With Depth-Color Layer Constraints

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video encoding and decoding technologies struggle to efficiently handle multi-layer RGBD video profiles, particularly in aligning and encoding depth maps with color videos, due to differences in sensor capture and resolution, leading to challenges in real-time volumetric rendering and device compatibility.

Innovation Solution

A method for encoding and decoding multi-layer RGBD video profiles by constraining the bitstream with specific layer limits, using confidence maps, and adhering to defined constraints such as maximum layers and chroma sampling, allowing for efficient decoding and rendering of volumetric video.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If depth maps and color videos are encoded separately without alignment constraints, then encoding flexibility is improved, but decoding complexity and rendering accuracy deteriorate

Engineering Contradiction:
Improveencoding flexibilityVSAvoiddecoding complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces specific parameter constraints in the bitstream (max_layers_minus1, chroma_sampling constraints, layer association flags) that define the relationship between depth and color layers. These parameter changes enable the encoder to maintain flexibility in processing while providing the decoder with predetermined structural rules, thereby reducing decoding complexity without sacrificing encoding adaptability.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent establishes layer association relationships and constraints during the encoding phase (preliminary action), where the encoder pre-defines which depth layers correspond to which color layers and sets maximum layer limits. This preliminary structuring allows the decoder to proceed with predictable operations, reducing its complexity while preserving encoding flexibility.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If multiple depth layers are encoded with high resolution, then depth map quality is improved, but bitstream size and processing load increase

Engineering Contradiction:
Improvedepth map qualityVSAvoidbitstream size
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies different quality levels to different layers locally. Instead of encoding all depth layers at maximum resolution, the encoder can assign full resolution to critical depth layers while using lower resolution for less important layers. The bitstream constraints allow selective association of high-quality depth maps with specific color layers, maintaining overall depth map quality while reducing total bitstream size.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements partial action by encoding only the necessary number of depth layers at high resolution based on the max_layers_minus1 constraint. Rather than encoding all possible depth layers at full quality, the system encodes a controlled subset that satisfies the application requirements, thereby reducing bitstream size while maintaining adequate depth map quality for the intended use case.

Inventive Principle:
Principle #16Partial or excessive action

3Productivity

If layer association constraints are enforced in the bitstream, then decoding efficiency is improved, but encoding complexity increases

Engineering Contradiction:
Improvedecoding efficiencyVSAvoidencoding complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent implements self-service through automated layer association mechanisms. The encoder automatically manages the association between depth and color layers according to the predefined max_layers_minus1 constraint and chroma_sampling rules, without requiring complex manual configuration. The bitstream structure itself carries the association information (sdi_associated_layer_idx, sdi_aux_id), allowing the system to self-organize the multi-layer data efficiently, improving decoding efficiency while keeping encoding complexity manageable through standardized procedures.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4657857A1Methods for video encoding and video decoding, video encoder and video decoder for supporting multi-layer RGBD video profiles
Publication Date: 2025.12.03 BEIJING XIAOMI MOBILE SOFTWARE CO LTD
  • EP4657857A1 patent drawingFigure 1
  • EP4657857A1 patent drawingFigure 2
  • EP4657857A1 patent drawingFigure 3

AI summary

A video encoder (100) for encoding according to an embodiment is provided. The video encoder (100) is configured for encoding a first video sequence into a first layer of a group of layers of a bitstream. Moreover, the video encoder (100) is configured for encoding a second video sequence into a second layer of the group of layers of the bitstream. Furthermore, the video encoder (100) is configured for writing or encoding profile and level information into the bitstream, wherein the profile and level information comprises a first constraint for the group of layers.