Point Cloud Cuboid Encoding With Interleaved TriSoup Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud compression methods, particularly TriSoup coding schemes, suffer from high transmission latency when occupancy and Trisoup information are transmitted over a single channel, which is problematic for real-time applications like self-driving cars.
Innovation Solution
The method encodes and decodes occupancy and Trisoup information in an interleaved manner within a bitstream, allowing simultaneous transmission over a single channel, reducing latency by interleaving occupancy and Trisoup information for each cuboid volume in the octree structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If occupancy and Trisoup information are transmitted sequentially over a single channel, then channel simplicity is maintained, but transmission latency increases
Solution Approach 1:
The patent segments the transmission of occupancy and Trisoup information into interleaved data units, where each data unit contains both occupancy information for a specific cuboid volume and Trisoup information for another cuboid volume. This segmentation allows both types of information to be transmitted simultaneously over a single channel without increasing latency, while maintaining channel simplicity.
Solution Approach 2:
The patent introduces a temporal dimension to the transmission process by interleaving occupancy and Trisoup information across multiple data units. Instead of transmitting all occupancy information before Trisoup information (sequential transmission), the system alternates between the two types of information across different time slots, effectively adding a temporal dimension to reduce latency.
2Loss of time
If separate channels are used for occupancy and Trisoup information, then transmission latency is reduced, but channel complexity increases
Solution Approach 1:
The patent merges the transmission of occupancy and Trisoup information into a single channel by interleaving them in alternating data units. Each data unit contains both occupancy information for one cuboid volume and Trisoup information for another cuboid volume, allowing both information types to share the same channel resources without increasing latency, thus avoiding the need for separate channels.
Solution Approach 2:
The patent implements a dynamic transmission structure where the composition of data units varies. Instead of fixed sequential transmission, the system dynamically alternates between occupancy and Trisoup information across different data units, allowing flexible and efficient use of a single channel while maintaining low latency performance.
3Productivity
If all occupancy information is transmitted before Trisoup information, then encoding simplicity is maintained, but transmission time increases
Solution Approach 1:
The patent segments the transmission process into multiple data units, each containing a mix of occupancy and Trisoup information. This segmentation allows the system to transmit both types of information concurrently rather than sequentially, doubling the effective transmission speed while maintaining encoding simplicity through the structured interleaving pattern.
Solution Approach 2:
The patent ensures continuous useful action by interleaving occupancy and Trisoup information transmission throughout the data units. Instead of completing all occupancy transmission before starting Trisoup transmission (which would create idle time), the system continuously alternates between the two types of information, maximizing the utilization of transmission time and improving overall productivity.
Data Source
AI summary
A method of encoding, into a bitstream comprising at least one data unit, position information of points of a point cloud, includes: encoding occupancy information of nodes from depth 0 to depth d−1 of an octree structure, the occupancy information being representative of presence of points of the point cloud included in cuboid volumes; and for each current cuboid volume including at least one point of the point cloud, encoding occupancy information of a leaf node and Trisoup information of another leaf node in the data unit in an interleaved way, wherein the Trisoup information is representative of presence of vertices on edges of the current cuboid volume, and vertices positions along the edges.