Multi-View Rendering of Non-Diffuse Objects Under Bandwidth Limits

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for compressing and transmitting multiple views in immersive video fail to accurately render non-diffuse pixels due to limited transmission rates and computational constraints, leading to incorrect reflectance in alternate views.

Innovation Solution

A method that adjusts the amount of data transmitted for non-diffuse pixels based on pixel and transmission rates, selecting and encoding these pixels to maintain overall view quality, using techniques such as entropy coding and reflectance models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If 2D images with depth maps are used to represent 3D scenes, then bandwidth consumption is reduced, but rendering accuracy for non-diffuse objects deteriorates

Engineering Contradiction:
Improvebandwidth consumptionVSAvoidrendering accuracy
Core Design Contradiction:
Loss of energyVSManufacturing precision

Solution Approach 1:

The patent segments the 3D scene representation into multiple components: 2D images, depth maps, and virtual light sources. This segmentation allows each component to be optimized independently - 2D images provide base visual information, depth maps add spatial data, and virtual light sources enable accurate lighting calculations for non-diffuse objects, thereby resolving the contradiction between bandwidth efficiency and rendering accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces virtual light sources as intermediary elements that mediate between the compressed 2D image data and the final rendered output. These virtual light sources carry lighting information that enables accurate rendering of non-diffuse objects without requiring transmission of complete 3D geometric and material data, thus achieving both bandwidth reduction and rendering accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Manufacturing precision

If complete 3D scene data is transmitted, then rendering accuracy is improved, but bandwidth consumption increases

Engineering Contradiction:
Improverendering accuracyVSAvoidbandwidth consumption
Core Design Contradiction:
Manufacturing precisionVSLoss of energy

Solution Approach 1:

The patent creates virtual copies of light sources in the virtual environment that replicate the lighting behavior of real-world sources. These virtual light source copies enable accurate lighting rendering without transmitting actual 3D scene geometry and material properties, achieving rendering accuracy while minimizing bandwidth consumption through selective data transmission.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent changes the representation parameters from complete 3D geometric and material data to a simplified set comprising 2D images, depth maps, and virtual light source parameters. This parameter transformation maintains essential rendering information while significantly reducing data transmission requirements, resolving the bandwidth-accuracy contradiction.

Inventive Principle:
Principle #35Parameter changes

3Device complexity

If 2D images are used for rendering, then device complexity is reduced, but the ability to render non-diffuse objects deteriorates

Engineering Contradiction:
Improverendering system complexityVSAvoidrendering capability for non-diffuse objects
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent performs preliminary extraction and transmission of lighting-relevant information (depth maps and virtual light source data) before rendering. This preliminary action enables the rendering system to accurately handle non-diffuse objects by having the necessary lighting and spatial information pre-prepared, without requiring complex real-time calculations or complete 3D scene data during rendering.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4523417B1Methods, servers and devices for transmitting and rendering multiple views comprising non-diffuse objects
Publication Date: 2026.05.06 TENCENT EUROPE (FRANCE) SAS
  • EP4523417B1 patent drawingFigure 1A
  • EP4523417B1 patent drawingFigure 1B
  • EP4523417B1 patent drawingFigure 2A

AI summary

A method implemented by a server (1SVR) for transmitting data used by a device (1CTL) for rendering multiple views (Vref, Vi) of a single scene, said multiple views comprising at least one basic view (Vref) and at least one additional view (Vi), said method comprising steps of: - generating (SS10) a stream (STR) by encoding said views (Vref, Vi); - selecting (1SS30) non-diffuse pixels (SNDP) among detected non-diffuse pixels (NDP); - generating (1SS40) data (1DND) representative of the selected non-diffuse pixels (SNDP; and - transmitting (SS50) said stream (STR) and said data (1DND) to said device, wherein a size of said data is a function of a pixel rate (PR) of said device and/or of a transmission rate (TR) between said server and said device.