Octree Predictive Tree Point Cloud Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The storage and transmission of large volumetric data sets, such as point clouds, are costly and time-consuming due to their size, limiting their use in real-time applications and requiring significant network resources.
Innovation Solution
A system that compresses spatial and attribute information using a combination of octrees and predictive trees, allowing for efficient encoding and decoding of point clouds, which reduces storage space and transmission time by partitioning points into cubes and sub-cubes and using predictive trees to encode duplicate points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If volumetric data sets are stored and transmitted in their original format, then complete spatial and attribute information is preserved, but storage space and transmission time increase significantly
Solution Approach 1:
The point cloud data is segmented into multiple octants using an octree structure, where the volumetric space is recursively divided into eight sub-volumes. This segmentation allows efficient encoding by representing only occupied regions rather than the entire volume, reducing storage requirements while preserving complete spatial information where needed.
Solution Approach 2:
The patent employs nested octree structures where octants are further subdivided into sub-octants, creating a hierarchical nesting pattern. This nested organization enables multi-level compression where common patterns at parent levels can predict details at child levels, reducing the total data quantity needed to represent the complete volumetric information.
2Loss of information
If volumetric data sets are stored and transmitted in their original format, then complete spatial and attribute information is preserved, but transmission time increases significantly
Solution Approach 1:
The data is pre-organized into an octree structure with predictive relationships established before transmission. By pre-computing the hierarchical organization and predictive models, the encoding efficiency is improved during transmission, allowing complete information to be conveyed in less time without sacrificing data fidelity.
3Quantity of substance
If octree structures are used to partition points into cubes and sub-cubes, then storage efficiency is improved, but encoding complexity increases
Solution Approach 1:
The octree structure is designed to be self-describing, where the hierarchical partitioning into cubes and sub-cubes automatically encodes spatial relationships without requiring external indexing structures. Each node in the octree contains information about its children, making the structure self-contained and reducing the complexity of encoding and decoding operations.
4Quantity of substance
If predictive trees are used to encode duplicate points, then storage efficiency is improved, but the complexity of the encoding structure increases
Solution Approach 1:
Duplicate points are merged into a single representative node in the predictive tree structure, with a count indicator specifying the number of duplicate occurrences. This merging approach consolidates redundant information into one location, reducing storage requirements while maintaining the ability to reconstruct all original points, thereby improving storage efficiency without proportionally increasing structural complexity.
Data Source
AI summary
An encoder is configured to compress point cloud geometry information using an octree/predictive tree combination geometric compression technique that embeds predictive trees in leaf nodes of an octree instead of encoding additional octree occupancy symbols for the leaf nodes. Alternatively an encoder may be configured to embed octrees in leaf nodes of a predictive tree structure. Similarly a decoder is configured to generate a reconstructed three-dimensional geometry from a bit stream including combined octree and predictive tree encoding information.