Point Cloud Geometry Encoding Using Sensing Paths and Coarse Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud codecs fail to combine simplicity, low latency, and high compression performance, particularly when encoding and decoding sparse geometry data sensed by sensors with flexible sensing paths, such as those used in autonomous vehicles.
Innovation Solution
A method and apparatus that encode and decode point clouds using a sensing path defined in a two-dimensional angular coordinate space, incorporating azimuthal and elevation angles, to represent potential point locations, and utilize coarse points for efficient encoding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If existing point cloud codecs are used to compress point cloud data, then compression is achieved, but encoding complexity and latency increase, making real-time processing difficult
Solution Approach 1:
The patent segments the point cloud data processing into distinct stages: generating candidate points from sensed data, selecting optimal candidate points, and encoding relative to selected points. This segmentation allows each stage to be optimized independently, reducing overall encoding complexity while maintaining compression efficiency.
Solution Approach 2:
The patent performs preliminary actions by generating multiple candidate points before the actual encoding process. These candidate points are pre-computed based on sensed point cloud data, allowing the encoder to select the most suitable candidates in advance. This preliminary generation reduces the computational burden during real-time encoding, lowering latency while preserving compression performance.
2Loss of energy
If existing point cloud codecs are used to compress point cloud data, then compression is achieved, but latency increases, preventing real-time application
Solution Approach 1:
The patent performs preliminary actions by generating multiple candidate points before the actual encoding process. These candidate points are pre-computed based on sensed point cloud data, allowing the encoder to select the most suitable candidates in advance. This preliminary generation reduces the computational burden during real-time encoding, lowering latency while preserving compression performance.
Solution Approach 2:
The encoding method uses the sensed point cloud data itself to generate candidate points and determine encoding references, rather than relying on external pre-processing or complex external algorithms. This self-service approach reduces dependency on external processing stages, thereby reducing overall latency while maintaining compression efficiency.
3Ease of manufacture
If sparse geometry data from spinning Lidar is encoded using conventional methods, then encoding is performed, but compression performance deteriorates due to data sparsity
Solution Approach 1:
The patent applies local quality by generating candidate points and selecting references based on local spatial relationships in the point cloud data. Rather than applying uniform encoding across all points, the method adapts to local density and distribution characteristics, improving compression performance for sparse regions while maintaining simplicity in dense regions.
Solution Approach 2:
The patent changes parameters by representing points relative to selected candidate points rather than using absolute coordinates. This parameter transformation adapts the encoding to the local structure of sparse data, improving compression efficiency without increasing encoding complexity. The relative representation captures spatial relationships more efficiently for sparse Lidar data.
Data Source
AI summary
A method and apparatus for encoding/decoding a point cloud may use any type of sensor following a sensing path. The method obtains coarse representations of sensed points and encodes the sensing path and the coarse representations. The sensing path and coarse representations of points are decoded, and points of the point cloud are reconstructed from the decoded sensing path and the decoded coarse representations.