LIDAR Extrinsic Matrix Accuracy Detector

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Inaccurate extrinsic matrices used in LIDAR fusion processes in autonomous driving vehicles can lead to errors in object detection and perception, due to faulty calibration or pose movement of LIDAR devices, which are not effectively addressed by existing calibration methods.

Innovation Solution

A computer-implemented method to determine the accuracy of extrinsic matrices by analyzing the distribution of coordinate values from post-LIDAR fusion point clouds, prompting recalibration when multiple peaks are detected, ensuring accurate transformation of point clouds into a uniform frame of reference.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing calibration methods are used, then the calibration process is simple, but the accuracy of extrinsic matrices deteriorates due to faulty calibration or pose movement

Engineering Contradiction:
Improveextrinsic matrix accuracyVSAvoidcalibration method complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements a feedback mechanism by continuously monitoring the distribution of coordinate values from LIDAR fusion point clouds and comparing them against expected patterns. When deviations indicate inaccurate extrinsic matrices, the system automatically triggers recalibration procedures, creating a closed-loop feedback system that maintains calibration accuracy without requiring manual intervention

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent replaces manual calibration procedures with an automated computational approach. Instead of relying on physical calibration objects or manual adjustment mechanisms, the system uses algorithmic analysis of point cloud coordinate distributions to detect and correct extrinsic matrix inaccuracies, substituting mechanical calibration processes with digital signal processing and statistical analysis

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Area of stationary object

If multiple LIDAR devices are used to reduce blind areas, then coverage improves, but the complexity of managing and calibrating multiple extrinsic matrices increases

Engineering Contradiction:
Improvecoverage areaVSAvoidcalibration management complexity
Core Design Contradiction:
Area of stationary objectVSDevice complexity

Solution Approach 1:

The patent merges the calibration management of multiple LIDAR devices into a unified framework. By analyzing the coordinate value distributions from all LIDAR devices simultaneously and using a common reference frame, the system consolidates what would otherwise be separate calibration management tasks into a single coordinated process, reducing overall complexity while maintaining comprehensive coverage

Inventive Principle:
Principle #5Merging (Combining)

3Reliability

If extrinsic matrices are updated frequently to account for pose movement, then accuracy is maintained, but the frequency of recalibration increases system complexity

Engineering Contradiction:
Improvecalibration reliabilityVSAvoidcalibration update efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent performs preliminary analysis of coordinate value distributions continuously in the background, preparing detection signals that indicate when recalibration is needed. This preliminary monitoring action allows the system to maintain reliability by detecting pose movements early, while avoiding unnecessary recalibration operations that would reduce productivity, thereby optimizing the balance between reliability and efficiency

Inventive Principle:
Principle #10Preliminary action

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This method ensures accurate extrinsic matrix calibration, reducing errors in object detection and perception processes by identifying and correcting inaccuracies in LIDAR fusion, thereby enhancing the reliability of autonomous driving systems.

Implementation Method 1

Individual points in the point cloud can be determined by transmitting a laser pulse and detecting a returning pulse, if any, reflected from the object, and determining the distance to the object according to the time delay between the transmitted pulse and the reception of the reflected pulse

Methodology Applied
Scientific EffectTime of Flight: Time of Flight

Implementation Method 2

A LIDAR device can estimate a distance to an object while scanning through a scene to assemble a point cloud representing a reflective surface of the object

Methodology Applied
Scientific EffectLight Reflection: Reflection

Data Source

PatentEP3939005B1A detector for point cloud fusion
Publication Date: 2024.10.23 BAIDU COM TIMES TECH (BEIJING) CO LTD
  • EP3939005B1 patent drawingFigure 1
  • EP3939005B1 patent drawingFigure 2
  • EP3939005B1 patent drawingFigure 3A

AI summary

A method, apparatus, and system for determining whether all extrinsic matrices are accurate is disclosed. A plurality of post-LIDAR fusion point clouds that are based on simultaneous outputs from a plurality of LIDAR devices installed at one ADV are obtained (610). The obtained plurality of point clouds are filtered to obtain a first set of points comprising all points in the plurality of point clouds that fall within a region of interest (620). Each point in the first set of points corresponds to one coordinate value on the axis in the up-down direction (630). A distribution of the first plurality of coordinate values is obtained (640). A quantity of peaks in the distribution of the first plurality of coordinate values is determined (650). Whether all extrinsic matrices associated with the plurality of LIDAR devices are accurate is determined based on the quantity of peaks in the distribution of the first plurality of coordinate values (660).