Immersive Video Atlas Encoding for Accurate 6DoF Depth Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current omnidirectional video technologies support only rotational motion, lacking the necessary 6 Degrees of Freedom (DoF) for immersive experiences, and depth map estimation processes are inefficient, leading to inaccuracies and increased complexity.
Innovation Solution
An immersive video encoding method that generates depth atlases through recursive block splits and metadata, enhancing depth map estimation accuracy by transmitting high-quality depth information efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth map estimation is performed using conventional methods, then the process can be completed, but the speed is slow and accuracy is insufficient
Solution Approach 1:
The patent performs depth map estimation during the image preprocessing step before encoding, rather than during rendering. High-quality depth maps are generated in advance and stored, eliminating the need for time-consuming estimation during playback. This preliminary action resolves the contradiction by achieving both high accuracy (through dedicated preprocessing) and high speed (by avoiding real-time computation).
Solution Approach 2:
The patent creates depth atlases that are copies of depth information extracted from multiple view images during preprocessing. These depth atlases are then used directly during rendering without requiring re-estimation. The copying approach allows accurate depth information to be reused efficiently, simultaneously improving accuracy and speed.
2Measurement precision
If high-quality depth map information is transmitted using conventional methods, then depth information can be provided, but the data amount is large and terminal complexity increases
Solution Approach 1:
The patent divides the depth atlas into multiple main blocks and further segments each block into sub-blocks. This segmentation allows selective transmission of only necessary depth information at appropriate resolutions, reducing overall data amount while maintaining accuracy where needed. The segmented structure also simplifies terminal processing by enabling block-based operations.
Solution Approach 2:
The patent applies different processing and transmission qualities to different regions of the depth atlas based on their importance. High-priority regions receive higher quality depth information, while less critical regions use lower quality representations. This local quality approach maintains overall accuracy while significantly reducing total data transmission requirements and terminal complexity.
3Adaptability or versatility
If 6DoF omnidirectional video is implemented, then immersive experience is improved, but the technology complexity and processing requirements increase
Solution Approach 1:
The patent performs extensive preprocessing to generate texture atlases and depth atlases from multiple view images before encoding. This preliminary action transforms complex 6DoF data into organized atlas structures that can be efficiently decoded and rendered. By doing the complex work in advance, the system achieves 6DoF capability without overwhelming terminal devices during playback.
Solution Approach 2:
The patent creates atlas copies (texture atlases and depth atlases) that consolidate information from multiple views into unified structures. These atlases serve as simplified representations that enable 6DoF rendering without requiring the terminal to process the original complex multi-view data, thus reducing system complexity while maintaining 6DoF functionality.
Data Source
AI summary
Disclosed herein are an immersive video encoding method and an immersive video decoding method. The immersive video encoding method includes setting basic view images among input images corresponding to multiple views, generating an atlas image using the basic view images, performing encoding on the atlas image, and generating metadata about an encoded atlas image, wherein the atlas image includes a texture atlas or a depth atlas.