Lidar Pose Matrix Compression for Faster Autonomous Driving Data Transfer

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The large volume of data generated by lidar scanners in autonomous driving systems leads to low transmission efficiency, making it difficult for the system to adjust vehicle direction and speed in a timely manner.

Innovation Solution

A data compression method that involves obtaining a pose matrix and distance matrix at the current moment, retrieving a previous pose matrix to calculate a residual matrix, and encoding and transmitting this data to the decoding end, allowing the decoding end to reconstruct the current pose matrix and store it along with the distance matrix.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If lidar scanner data is transmitted in full detail, then data completeness is maintained, but transmission efficiency deteriorates due to large data volume

Engineering Contradiction:
Improvedata completenessVSAvoidtransmission efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent extracts only the essential components needed for autonomous driving control from the complete lidar data. Specifically, it extracts pose matrix information (position and orientation) and distance matrix information (range to objects), discarding redundant detailed scan data. This extraction principle resolves the contradiction by maintaining the reliability needed for control decisions while dramatically improving transmission efficiency through selective data transmission.

Inventive Principle:
Principle #2Taking out (Extraction)

2Productivity

If data transmission volume is reduced through compression, then transmission efficiency is improved, but data processing complexity increases at the decoding end

Engineering Contradiction:
Improvetransmission efficiencyVSAvoiddata processing complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments lidar data into distinct functional components: pose matrix (containing position and orientation information) and distance matrix (containing range information). Each segment is processed and transmitted independently with appropriate compression. This segmentation reduces overall processing complexity compared to compressing the entire dataset, as each segment can be handled with specialized, optimized algorithms tailored to its specific characteristics.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms raw lidar scan data into parameter-based representations (pose matrices and distance matrices). Instead of transmitting raw point cloud data with millions of coordinates, the system converts this into compact matrix forms with key parameters such as position coordinates, rotation angles, and distance values. This parameter transformation significantly reduces data volume while maintaining the essential information needed for autonomous driving control, resolving the contradiction between compression efficiency and processing complexity.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4718120A1Data compression method and apparatus, and device and medium
Publication Date: 2026.04.01 SHANGHAI CAMBRICON INFORMATION TECH CO LTD
  • EP4718120A1 patent drawingFigure 1~2
  • EP4718120A1 patent drawingFigure 3~4
  • EP4718120A1 patent drawingFigure 5~6

AI summary

The present disclosure provides a data compression device including an obtaining unit, a processing unit and a transmitting unit. Based on the technical solution provided by this disclosure, the amount of data transmitted is greatly reduced, thereby improving transmission efficiency.