3D Data Refinement Using Occupancy Geometry Attribute Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D data decoding methods cannot perform refinement processing that utilizes the relationships between occupancies, geometries, and attributes, limiting the ability to improve image quality through filter processing.
Innovation Solution
A 3D data decoding apparatus that includes a refinement information decoder, geometry decoder, attribute decoder, and a refiner to decode and refine 3D data, allowing for the selection of occupancy, geometry, or attribute refinement targets and performing refinement processing based on specified information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If filter processing is performed only on attributes using existing SEI methods, then attribute refinement is achieved, but refinement using information between images (occupancy and geometry) is not possible
Solution Approach 1:
The patent segments the refinement processing into three distinct target types: occupancy refinement, geometry refinement, and attribute refinement. Each target can be processed independently by the refiner based on the decoded refinement target information, allowing selective application of refinement to specific data components rather than treating all data uniformly.
Solution Approach 2:
The refiner is designed as a multi-functional component that can perform refinement processing on multiple types of data (occupancy, geometry, and attribute frames). The single refiner unit adapts its operation based on the decoded refinement target information, making it versatile enough to handle different refinement scenarios without requiring separate processing units for each data type.
2Device complexity
If refinement processing is limited to attribute-only filtering, then processing complexity is reduced, but coding distortion cannot be further reduced using auxiliary information
Solution Approach 1:
The patent performs preliminary decoding of refinement target information from the coded data before executing the refinement processing. This preliminary action allows the system to determine in advance which type of refinement (occupancy, geometry, or attribute) should be applied, enabling the refiner to prepare the appropriate processing path and minimize coding distortion through targeted refinement.
3Manufacturing precision
If multiple refinement targets (occupancy, geometry, attribute) are supported, then refinement quality is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal refiner that can handle multiple refinement targets (occupancy, geometry, and attribute frames) through a single processing unit. The refiner's operation is controlled by the decoded refinement target information, which directs the refiner to apply appropriate refinement processing to the selected target type, thereby supporting multiple refinement functions without requiring separate dedicated processors for each target.
Data Source
AI summary
A 3D data decoding apparatus for decoding coded data and decoding 3D data including attribute information includes a header decoder that decodes characteristics information of refinement and activation information of the refinement, and a refiner that performs refinement processing of an attribute frame or an geometry frame. Refinement target information indicating which of an occupancy, a geometry, or an attribute is to be used is decoded from coded data of the characteristics information and the refinement is performed using an image specified according to the refinement target information.