3D Image Depth Adjustment for VR Sweet Spot Degradation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current virtual reality applications face challenges in maintaining high-quality 3D experiences when users move outside the limited 'sweet spot' due to significant degradation in image quality and introduction of artefacts, particularly due to lack of de-occlusion data, which existing solutions fail to optimize in terms of user experience, data rate, complexity, and resource usage.

Innovation Solution

An image generating apparatus that determines view poses for each eye and adjusts depth values based on the difference between the current and reference poses, synthesizing output images by view shifting the reference images, allowing for improved user experience and reduced degradation by modifying depth values to maintain a consistent 3D perception even when the user moves outside the sweet spot.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If view shifting is performed based on original depth values when user moves outside the sweet spot, then image quality degrades significantly and artefacts are introduced, but if depth values are modified to reduce the difference from reference depth, then image quality degradation is reduced and user experience is improved

Engineering Contradiction:
Improveimage qualityVSAvoiduser movement freedom
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent modifies depth values dynamically based on the difference between current and reference view poses. When the user moves outside the sweet spot, the system adjusts the depth parameters to reduce the difference from reference depth values, thereby maintaining image quality while allowing greater user movement freedom. This parameter adjustment prevents the severe quality degradation that would otherwise occur with large pose differences.

Inventive Principle:
Principle #35Parameter changes

2Loss of energy

If a limited sweet spot is enforced to maintain high image quality, then data rate is reduced, but user movement freedom is restricted

Engineering Contradiction:
Improvedata rateVSAvoiduser movement freedom
Core Design Contradiction:
Loss of energyVSEase of operation

Solution Approach 1:

The system dynamically adjusts depth values based on the user's current view pose relative to the reference pose. Instead of enforcing a fixed sweet spot boundary, the patent continuously modifies depth parameters to maintain image quality across a wider range of movements. This dynamic adaptation allows users to move more freely while keeping data transmission efficient, as the system only transmits necessary depth adjustments rather than complete scene data.

Inventive Principle:
Principle #15Dynamics

3Reliability

If depth values are heavily modified to maintain image quality outside the sweet spot, then user experience is improved, but computational complexity increases

Engineering Contradiction:
Improveuser experience consistencyVSAvoidprocessing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies depth value modification selectively based on the specific view pose difference. Rather than uniformly processing all pixels or applying complex algorithms throughout the scene, the system focuses computational effort on adjusting depth values in regions where the pose difference impacts image quality. This localized approach maintains consistent user experience while reducing overall processing complexity compared to global scene re-rendering.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3871408B1Image generating apparatus and method therefor
Publication Date: 2024.05.22 KONINKLIJKE PHILIPS NV
  • EP3871408B1 patent drawingFigure 1
  • EP3871408B1 patent drawingFigure 2
  • EP3871408B1 patent drawingFigure 3

AI summary

An apparatus comprises a determiner (305) which determines a first-eye and a second eye view pose. A receiver (301) receives a reference first-eye image with associated depth values and a reference second-eye image with associated depth values, the reference first-eye image being for a first-eye reference pose and the reference second-eye image being for a second-eye reference pose. A depth processor (311) determines a reference depth value, and modifiers (307) generate modified depth values by reducing a difference between the received depth values and the reference depth value by an amount that depends on a difference between the second or first-eye view pose and the second or first-eye reference pose. A synthesizer (303) synthesizes an output first-eye image for the first-eye view pose by view shifting the reference first-eye image and an output second-eye image for the second-eye view pose by view shifting the reference second-eye image based on the modified depth values. The terms first and second may be replaced by left and right, respectively or vice verse. E.g. the terms first-eye view pose, second-eye view pose, reference first-eye image, and reference second-eye image may be replaced by left-eye view pose, right-eye view pose, reference left-eye image, and reference right-eye image, respectively.