Camera Sensor Calibration Using Reliability Maps for Vehicle Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current sensor position calibration methods for infrastructure sensors in autonomous vehicle systems suffer from low accuracy due to deviations in position information collection, leading to inefficient vehicle control and increased errors in measurement reliability.
Innovation Solution
A sensor position calibration device that includes a moving object information acquisition unit, a moving object measurement unit, a calibration unit, a calibration error calculation unit, a reliability map generation unit, and a moving object control unit, which calculates and corrects position errors, generates reliability maps, and controls vehicle movement to minimize measurement errors by dividing the movement area into zones based on reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor position calibration is performed using moving object position measurement results, then position information can be collected in the control area, but the accuracy of position measurement becomes uneven due to deviations in collected position information
Solution Approach 1:
The movement area is divided into multiple zones based on reliability of position measurement. By segmenting the area into high-reliability and low-reliability zones, the system can identify and address accuracy deviations in specific regions without compromising the overall calibration process. This segmentation allows for targeted improvement of measurement accuracy in problematic areas.
Solution Approach 2:
Different reliability levels are assigned to different zones within the movement area based on local measurement characteristics. The system applies local quality assessment by evaluating position measurement reliability specifically in each zone rather than treating the entire area uniformly, enabling identification of areas with deviation issues.
2Extent of automation
If vehicle control is performed based on collected position information, then autonomous control can be achieved, but control efficiency decreases due to low accuracy and deviations in position information
Solution Approach 1:
The vehicle control system dynamically adjusts control strategies based on the reliability map that indicates position measurement accuracy in different zones. When the vehicle enters low-reliability zones, the control system adapts by using alternative control methods or reducing autonomous control dependency, thereby maintaining control efficiency despite accuracy deviations.
Solution Approach 2:
The system uses feedback from the reliability map to continuously adjust vehicle control decisions. By monitoring position measurement reliability in real-time and feeding this information back to the control system, the vehicle can optimize its control efficiency by avoiding reliance on inaccurate position data in low-reliability zones.
3Measurement precision
If position information is collected along the moving object's route, then calibration data can be obtained, but operation efficiency decreases due to route, speed, and other constraints
Solution Approach 1:
The system performs preliminary assessment of position measurement reliability across different zones before executing the calibration process. By pre-generating the reliability map and identifying high-reliability zones in advance, the moving object can plan its route to efficiently collect calibration data from optimal areas, reducing unnecessary travel and improving operation efficiency.
Data Source
AI summary
A sensor position calibration device includes a moving object information acquisition unit that acquires a self-position measured by a reference moving object moving in a movement area; a moving object measurement unit that measures a position of the reference moving object based on observation information of a camera; a calibration unit that calibrates position information of the camera by using the self-position of the moving object and an estimated position calculated based on the measured position and the position information of the camera; a calibration error calculation unit that calculates an error between a second estimated position and the self-position; a reliability map generation unit that, based on the error, generates a reliability map indicating reliability of position measurement in the movement area using the camera; and a moving object control unit that controls movement of the moving object by using the reliability map.