Spinning Sensor Head Point Cloud Encoding for Low-Latency Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud codecs fail to combine simplicity, low latency, and high compression performance, particularly in encoding and decoding sparse geometry data captured by spinning sensor heads for applications like autonomous driving, due to non-optimal coding structures and lack of interaction between sensors.
Innovation Solution
A method and apparatus for encoding and decoding point clouds captured by a spinning sensors head, involving ordering points based on lexicographic angles and indices, encoding the number of duplicated points, and using entropy coding to optimize the bitstream, allowing for efficient compression and low latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If existing point cloud codecs are used, then compression performance can be achieved, but encoding complexity and latency increase
Solution Approach 1:
The patent segments the point cloud encoding process into distinct modules: ordering points by lexicographic order of (azimuthal angle, sensor index), counting duplicated points, and entropy coding. This segmentation allows each module to be optimized independently, reducing overall complexity while maintaining compression efficiency.
Solution Approach 2:
The patent performs preliminary ordering of points by lexicographic order before encoding. This preliminary action organizes the data in advance, enabling simpler and faster encoding operations. The ordering is based on azimuthal angles and sensor indices, which are pre-computed from the captured point cloud data.
2Loss of energy
If existing point cloud codecs are used, then compression can be achieved, but decoding latency increases
Solution Approach 1:
The patent performs preliminary ordering of points by lexicographic order before encoding. This preliminary action organizes the data in advance, enabling simpler and faster encoding operations. The ordering is based on azimuthal angles and sensor indices, which are pre-computed from the captured point cloud data.
3Loss of time
If simple encoding methods are used, then latency is reduced, but compression performance deteriorates
Solution Approach 1:
The patent uses the inherent structure of spinning sensor data - specifically the regular distribution of points captured by multiple sensors at different azimuthal angles - to enable self-compression. The lexicographic ordering and duplication counting exploit the natural organization of the data, eliminating the need for complex external compression algorithms while achieving efficient compression.
4Device complexity
If duplicated points are not handled, then encoding is simpler, but data redundancy increases
Solution Approach 1:
The patent performs preliminary ordering of points by lexicographic order before encoding. This preliminary action organizes the data in advance, enabling simpler and faster encoding operations. The ordering is based on azimuthal angles and sensor indices, which are pre-computed from the captured point cloud data.
Solution Approach 2:
The patent uses the inherent structure of spinning sensor data - specifically the regular distribution of points captured by multiple sensors at different azimuthal angles - to enable self-compression. The lexicographic ordering and duplication counting exploit the natural organization of the data, eliminating the need for complex external compression algorithms while achieving efficient compression.
Data Source
AI summary
There is provided methods and apparatus of encoding/decoding a point cloud captured by a set of sensors. The encoding method obtains a list of ordered points of the point cloud by ordering the points according to a lexicographic order based on azimuthal angles and sensor indices associated with points of the point cloud; obtains a next point following a current point in the list of ordered points and having an order index different of an order index of the current point, and obtains a number of duplicated points equals to a number of points following the current point that have an order index equals to the order index of the current point; encodes the number of duplicated points and encodes an order index difference representing a difference between the order index of the next point and the order index of the current point.