Point Cloud Stream Grouping for View-Dependent G-PCC Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional G-PCC stream distribution technologies result in unnecessary high bit rates for high-definition Point Clouds, leading to potential reproduction interruptions, and reducing bit rates for lower definition, causing sparse point representation, thus not efficiently utilizing network bandwidth for high-quality viewing experiences.
Innovation Solution
An information processing apparatus and method that generate files with spatial positional information and grouping information for segmented Point Cloud data, along with directional and definition degree information, to optimize the distribution of encoded Point Cloud data streams, allowing for high-definition viewing while reducing bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the three-dimensional structure information of a Point Cloud object is uniformly compressed by octree encoding to achieve high definition, then the Point Cloud data quality is improved, but the bit rate increases unnecessarily leading to reproduction interruptions
Solution Approach 1:
The patent applies local quality by differentiating compression levels across different spatial regions of the Point Cloud. The viewing direction information is used to identify a front half region (visible area) and a rear half region (invisible area), applying different compression ratios to each. The visible region maintains high definition quality while the invisible region uses lower compression, optimizing bandwidth utilization without sacrificing user-perceived quality.
2Reliability
If the bit rate is reduced to restrain reproduction interruptions, then network transmission stability is improved, but the compression ratio is uniformly increased resulting in sparse point representation and low definition
Solution Approach 1:
The patent implements local quality by applying differential compression strategies to different spatial regions. The front half region (visible to user) maintains high point density and definition quality, while the rear half region (not visible) uses higher compression ratios. This ensures reproduction stability through optimized bit rate while preserving definition quality in the user's field of view.
Solution Approach 2:
The patent applies preliminary action by pre-calculating and embedding viewing direction information and region division information into the Point Cloud data before transmission. This allows the receiving device to automatically select appropriate compression regions based on the predetermined viewing direction, ensuring both reproduction stability and definition quality without real-time processing delays.
3Ease of manufacture
If uniform compression is applied to the entire Point Cloud object, then encoding simplicity is maintained, but network bandwidth is not utilized efficiently for high-quality viewing experiences
Solution Approach 1:
The patent applies local quality by dividing the Point Cloud into distinct compression regions (front half and rear half) based on viewing direction. Each region is encoded with appropriate compression parameters, allowing efficient bandwidth utilization while maintaining high quality in visible areas. The region division information and viewing direction information guide this differentiated encoding process.
Solution Approach 2:
The patent introduces dynamics by making the compression strategy adaptive to viewing direction. The front half and rear half regions are dynamically identified based on viewing direction information, allowing the encoding process to adapt to different viewing scenarios. This dynamic region-based approach optimizes bandwidth utilization while maintaining encoding feasibility through automated region selection.
Data Source
AI summary
The present disclosure relates to an information processing apparatus and an information processing method that are capable of distributing higher-quality G-PCC streams. When G-PCC streams obtained by encoding Point Cloud data according to G-PCC are generated, there are generated spatial positional information indicating spatial positions of respective pieces of partial Point Cloud data that represent individual parts into which the Point Cloud data is segmented, and grouping information that groups the partial Point Cloud data. The present technology is applicable to a generating apparatus that generates G-PCC streams, for example.