NDT Map Compression Using Mean and Covariance Transform Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The large data size of 3D point cloud maps in autonomous driving systems limits the scalability of driving areas and hinders real-time efficient operation, as the size of the NDT map data increases with the expansion of the driving area.

Innovation Solution

A method is introduced to compress NDT map data by processing mean vector and covariance matrix data using transformed center and offset vectors, and standard deviation and rotational transform angles, within specific reference ranges, allowing for efficient data representation and storage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of stationary object

If 3D point cloud map data is used for autonomous driving, then the driving area coverage is improved, but the data size increases making it difficult to manage and process in real-time

Engineering Contradiction:
Improvedriving area coverageVSAvoiddata size
Core Design Contradiction:
Area of stationary objectVSQuantity of substance

Solution Approach 1:

The patent transforms the 3D point cloud data from raw coordinate format into NDT map data by changing the parameter representation. Each lattice space is modeled as a normal distribution with parameters including mean vector (position), covariance matrix (shape and orientation), and standard deviation. This parameter transformation compresses the data while preserving the essential geometric information needed for autonomous driving navigation and mapping.

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If NDT map data is compressed to reduce data size, then storage efficiency is improved, but registration performance may deteriorate

Engineering Contradiction:
Improvedata sizeVSAvoidregistration performance
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent applies partial compression by selectively processing different components of the NDT map data with different compression strategies. The mean vector data is compressed using one method while the covariance matrix data is compressed using another method, allowing optimization for both data size reduction and registration performance preservation. This selective partial compression maintains critical registration information while reducing overall data volume.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4236323A1Method, apparatus and computer program for compressing NDT map data for autonomous driving system
Publication Date: 2023.08.30 RIDEFLUX INC
  • EP4236323A1 patent drawingFigure 1
  • EP4236323A1 patent drawingFigure 2
  • EP4236323A1 patent drawingFigure 3~4

AI summary

Provided are a method, apparatus, and a computer program for compressing NDT map data for an autonomous driving system. Various embodiments of the present invention provide a method of compressing normal distribution transform (NDT) map data generated by modeling a three-dimensional (3D) point cloud for a predetermined area as a normal distribution set for an autonomous driving system, which is performed by a computing device, the method including: processing mean vector data included in the NDT map data; processing covariance matrix data included in the NDT map data; and generating compressed NDT map data using the processed mean vector data and the processed covariance matrix data.