Immersive Video Atlas Alignment for 6 DoF Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual reality services provide limited 6 degrees of freedom (DoF) by failing to include left-and-right translation and up-and-down translation in immersive images, necessitating improved methods for dividing, packing, and merging atlases to enhance stereoscopic image presentation.
Innovation Solution
An immersive image encoding method that generates geometry atlases aligned with texture atlases, allowing efficient packing and merging, and a decoding method that obtains and signals packing information to reconstruct immersive images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If geometry atlases are divided and packed to provide full 6 DoF immersive images, then the completeness of stereoscopic image presentation is improved, but the complexity of atlas processing and merging increases
Solution Approach 1:
The geometry atlas is divided into multiple packed geometry atlases that can be separately processed and then merged with texture atlases. This segmentation allows the system to handle large-scale 6 DoF immersive images by breaking them into manageable chunks while maintaining the ability to reconstruct the complete stereoscopic view
Solution Approach 2:
The patent merges packed geometry atlases with texture atlases in efficient positions to create a unified immersive image representation. This merging process combines the spatial structure information from geometry atlases with the visual information from texture atlases, enabling complete 6 DoF rendering without requiring separate processing pipelines
2Productivity
If geometry atlases are aligned with texture atlases to improve merging efficiency, then the merging speed is improved, but the flexibility in handling different atlas sizes is reduced
Solution Approach 1:
The alignment process is made dynamic by allowing the geometry atlas height to be adjusted to match the texture atlas height. This dynamic adaptation enables efficient merging while maintaining flexibility, as the system can resize geometry atlases based on the specific dimensions of accompanying texture atlases rather than requiring fixed-size inputs
Solution Approach 2:
The patent changes the height parameter of geometry atlases to align with texture atlas dimensions. By modifying this specific parameter, the system achieves compatible dimensions for efficient merging while maintaining the ability to handle various atlas size configurations through parameter adjustment rather than structural redesign
3Manufacturing precision
If atlas packing information is signaled and obtained to enable proper reconstruction, then the accuracy of immersive image reconstruction is improved, but the bitstream overhead increases
Solution Approach 1:
Packing information is extracted and signaled separately in the bitstream to enable accurate reconstruction without embedding it throughout the entire data structure. This extraction approach allows the decoder to obtain necessary packing details (such as atlas dimensions and arrangement) as metadata, improving reconstruction accuracy while minimizing overhead compared to redundant encoding throughout the bitstream
Data Source
AI summary
Disclosed herein are an immersive image encoding/decoding method and apparatus, and a method for transmitting a bitstream generated by the immersive image encoding method. An immersive image encoding method according to the present disclosure may include: generating geometry atlases for an immersive image; aligning the geometry atlases to make an alignment height of the geometry atlases correspond to a height of a texture atlas; and generating a geometry bitstream by encoding the aligned geometry atlases.