Depth Encoding for 3D Scene Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack a method to encode and decode data representative of a 3D scene that can be rendered as both 3DoF and 6DoF video, limiting the flexibility and immersion in volumetric content viewing experiences.
Innovation Solution
A method and device for encoding and decoding data that quantizes depth values beyond the bit depth limitations, using a second set of quantization parameters common to multiple blocks, allowing for flexible rendering as 3DoF or 6DoF by encoding and decoding quantized depth values according to these parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth data is quantized to exceed encoding bit depth limitations, then the number of quantized depth values increases beyond available encoding values, but this creates a contradiction between measurement precision and loss of information
Solution Approach 1:
The patent divides the picture into multiple blocks and determines a separate set of candidate quantization parameters for each block. This segmentation allows different regions of the image to use different quantization strategies, preserving depth precision in critical areas while managing overall bit rate constraints.
Solution Approach 2:
The patent changes the quantization parameters dynamically based on block characteristics. By selecting from multiple candidate quantization parameters for each block, the system adapts the quantization precision to local depth variations, maintaining measurement precision where needed while reducing information loss in other regions.
2Adaptability or versatility
If a single stream is encoded to support both 3DoF and 6DoF rendering, then versatility is improved, but device complexity increases due to the need for flexible encoding and decoding mechanisms
Solution Approach 1:
The patent creates a universal encoding framework that produces a single bitstream capable of supporting both 3DoF and 6DoF rendering. The depth encoding mechanism is designed to be format-agnostic, allowing the same encoded data to be used for different rendering types without requiring separate encoding streams.
Solution Approach 2:
The patent introduces dynamic selection of quantization parameters based on the desired rendering format. The encoding system can adapt its parameters to optimize for either 3DoF or 6DoF output, and the decoding system can dynamically interpret the same bitstream according to the target rendering requirements.
Data Source
AI summary
Methods and device for encoding/decoding data representative of depth of a 3D scene. The depth data are quantized in a range of quantized depth values larger than a range of encoding values allowed by a determined encoding bit depth. For blocks of pixels comprising the depth data, a first set of candidate quantization parameters is determined. A second set of quantization parameters is determined as a subset of the union of the first sets. The second set comprising candidate quantization parameters common to a plurality of blocks. One or more quantization parameters of the second set being associated with each block of pixels of the picture. The second set of quantization parameters is encoded, and the quantized depth values are encoded according to the quantization parameters.