Point Cloud Transmission With Hierarchical Attribute Encoding

The patent segments point cloud data into multiple octrees based on spatial division, and further segments attribute data into subgroups with tree structure. This segmentation allows parallel processing of different octrees and subgroups, reducing computational complexity while maintaining high throughput by processing data in manageable chunks rather than handling the entire point cloud at once.

The patent introduces a hierarchical structure with multiple levels of detail (LoDs) that adds an organizational dimension to the data. By structuring attribute data into subgroups with parent-child relationships and creating multiple LoDs, the system transforms the flat data structure into a hierarchical one that enables efficient parallel processing and reduces computational complexity.

By segmenting point cloud data into multiple octrees and attribute data into subgroups, the patent enables parallel encoding and decoding operations across different segments. This segmentation reduces the time required to process the entire data set by performing operations simultaneously on multiple chunks, thereby reducing latency while maintaining high throughput.

The patent performs preliminary organization of attribute data into subgroups with tree structure and generation of multiple LoDs before the actual encoding process. This preliminary structuring prepares the data in advance, making the subsequent encoding and decoding operations more efficient and reducing processing time.

The patent segments attribute data into subgroups with tree structure, where each subgroup is associated with specific octrees. This segmentation allows the encoding process to handle smaller, more manageable portions of data while maintaining a systematic approach. The hierarchical structure provides clear organization that simplifies the encoding process despite the large volume of data.

The patent introduces a hierarchical dimension to attribute data organization through subgroups and multiple LoDs. This additional organizational layer transforms the encoding process from handling a flat large dataset to processing structured hierarchical data, which simplifies the encoding logic while managing the complexity introduced by data volume.

The patent segments point cloud data into multiple octrees that can be independently processed and encoded. This segmentation enables scalable services by allowing different portions of the point cloud to be processed at different resolutions and qualities. The hierarchical structure of subgroups and LoDs provides flexibility for adaptive coding, enabling the system to scale services according to computational resources and requirements.

The patent creates a dynamic hierarchical structure with multiple LoDs that can be adjusted based on processing requirements. This dynamic structure allows the encoding and decoding processes to adapt to different service levels, providing scalability. The system can dynamically select which LoDs to process based on available resources and service requirements, managing complexity while maintaining service flexibility.