On-Board Sensor Extrinsic Calibration Using Map Feature Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The reliability of automatic driving systems in autonomous vehicles is compromised due to the need for frequent depot recalibration of on-board sensors, which is costly and negatively impacts user experience, as external factors like vibration and collisions cause offsets in sensor spatial relationships.
Innovation Solution
A method and device for automatically calibrating extrinsic sensor parameters using first and second feature point information, determined through a position and orientation system, allowing calibration without reliance on fixed check places or laboratories, by establishing a mapping parameter from the sensor to the vehicle body based on geographic locations and coordinates within a world coordinate system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual calibration with preset markers in fixed areas is used, then initial sensor spatial relationship accuracy is achieved, but frequent depot recalibration is required due to vibration and collision offsets
Solution Approach 1:
The system performs automatic self-calibration by utilizing the vehicle's own sensors (camera, GPS, inertial sensor) to detect and correct spatial relationship offsets between sensors without requiring external calibration equipment or depot intervention. The calibration is executed automatically during vehicle operation based on sensed environmental features.
Solution Approach 2:
The calibration system dynamically adapts to changing sensor spatial relationships caused by vibration and collision by continuously monitoring and adjusting calibration parameters in real-time, rather than relying on static initial calibration that degrades over time.
2Measurement precision
If manual calibration with preset markers in fixed areas is used, then initial sensor spatial relationship accuracy is achieved, but transportation and time costs increase due to frequent depot visits
Solution Approach 1:
The system performs automatic self-calibration by utilizing the vehicle's own sensors (camera, GPS, inertial sensor) to detect and correct spatial relationship offsets between sensors without requiring external calibration equipment or depot intervention. The calibration is executed automatically during vehicle operation based on sensed environmental features.
Solution Approach 2:
The calibration system dynamically adapts to changing sensor spatial relationships caused by vibration and collision by continuously monitoring and adjusting calibration parameters in real-time, rather than relying on static initial calibration that degrades over time.
3Measurement precision
If manual calibration with preset markers is used, then calibration accuracy is achieved, but user experience deteriorates due to frequent depot recalibration requirements
Solution Approach 1:
The system performs automatic self-calibration by utilizing the vehicle's own sensors (camera, GPS, inertial sensor) to detect and correct spatial relationship offsets between sensors without requiring external calibration equipment or depot intervention. The calibration is executed automatically during vehicle operation based on sensed environmental features.
Solution Approach 2:
The calibration system dynamically adapts to changing sensor spatial relationships caused by vibration and collision by continuously monitoring and adjusting calibration parameters in real-time, rather than relying on static initial calibration that degrades over time.
4Ease of operation
If automatic calibration using environmental features is implemented, then depot recalibration is eliminated, but calibration reliability must be maintained without controlled check field conditions
Solution Approach 1:
The system uses environmental features (buildings, trees, poles, road markings) as intermediary reference objects to establish spatial relationships between sensors. These naturally occurring features serve as calibration targets, replacing the need for artificial check field markers while enabling calibration in real-world operating conditions.
Solution Approach 2:
The calibration system dynamically adapts to changing sensor spatial relationships caused by vibration and collision by continuously monitoring and adjusting calibration parameters in real-time, rather than relying on static initial calibration that degrades over time.
Data Source
AI summary
A method for calibrating an extrinsic parameter of an on-board sensor includes obtaining first feature point information collected by a first on-board sensor; determining a geographic location of a vehicle body in which the first on-board sensor is located in a world coordinate system; determining second feature point information within a collection range of the first sensor in a map database based on the geographic location of the vehicle body in the world coordinate system; determining a plurality of first feature points and a plurality of second feature points based on the first feature point information and the second feature point information.