Roadside Unit Radar Calibration Using Point Cloud Alignment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for calibrating intelligent roadside units, such as GPS, fail to meet the high accuracy requirements for automatic driving, especially when combined with radar systems, as they do not effectively determine the precise location and orientation of sensors like radars and cameras.

Innovation Solution

A method involving an intelligent driving vehicle within a preset range of the roadside unit, using point clouds from radars to match and align the sensor positions through rotation and translation, determining the location relation between radars and cameras, and adjusting angles to achieve accurate positioning.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If GPS is used to obtain the location of the intelligent roadside unit, then the location can be obtained, but the accuracy does not meet the requirements of automatic driving

Engineering Contradiction:
Improvelocation accuracyVSAvoidsatisfaction of accuracy requirements
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent introduces point cloud data as an intermediary to establish a relationship between the roadside unit radar and the vehicle radar. By matching feature points in the point clouds obtained from both radars, the system determines the location and orientation of the roadside unit relative to the vehicle, thereby achieving high-precision positioning that satisfies automatic driving requirements.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If radar systems are added to improve active sensing ability, then sensing capability is enhanced, but the precise location and orientation of sensors become difficult to determine

Engineering Contradiction:
Improveactive sensing abilityVSAvoidsensor location and orientation determination
Core Design Contradiction:
Adaptability or versatilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent employs a feedback mechanism where the roadside unit radar detects the vehicle, obtains point cloud data, and matches feature points with the vehicle radar's point cloud. This feedback loop allows the system to iteratively refine the location and orientation information of the roadside unit sensors, making precise determination achievable despite the added complexity of multiple radar systems.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3617737B1Method, apparatus, device and medium for calibrating intelligent roadside unit
Publication Date: 2023.10.04 APOLLO INTELLIGENT DRIVING (BEIJING) TECHNOLOGY CO LTD
  • EP3617737B1 patent drawingFigure 1~2
  • EP3617737B1 patent drawingFigure 3~4
  • EP3617737B1 patent drawingFigure 5~6

AI summary

The present disclosure proposes a method, an apparatus, a device, and a medium for calibrating an intelligent roadside unit. The method includes: obtaining an intelligent driving vehicle within a preset range of the intelligent roadside unit; obtaining a first point cloud detected by a first radar in the intelligent driving vehicle, and obtaining location information of the intelligent driving vehicle; obtaining a second point cloud detected by a second radar in the intelligent roadside unit; and obtaining location information of the second radar based on the first point cloud, the second point cloud, and the location information of the intelligent driving vehicle.