Ego-Vessel Camera Calibration Using Solid-Object Interfaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera calibration methods are time-consuming, labor-intensive, and lack accuracy, particularly in determining extrinsic parameters, and current automatic methods fail to account for the extent of other vessels or accurate GPS positions, leading to imprecise calibrations that affect navigation systems.
Innovation Solution
A method and system for calibrating cameras on ego vessels using image data of solid objects and their interfaces with the sky or water surface, combined with reference data, to determine calibration values by minimizing differences between imaged and real-world object interfaces, enabling accurate and cost-efficient calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual camera calibration is performed, then calibration accuracy is improved, but time consumption and labor intensity increase
Solution Approach 1:
The system performs automatic calibration using onboard sensors (GPS, IMU, barometer) and image processing algorithms to determine camera parameters without human intervention. The calibration process serves itself by automatically detecting features, computing extrinsic parameters, and updating calibration data, eliminating the need for manual operation while maintaining high accuracy through multi-sensor fusion.
2Loss of time
If automatic calibration using horizon detection is used, then time consumption is reduced, but calibration accuracy deteriorates due to inability to determine position
Solution Approach 1:
The system merges data from multiple independent sources: GPS provides absolute position, IMU provides orientation angles, barometer provides altitude, and image processing provides visual verification. By combining these diverse measurement sources through sensor fusion, the system achieves complete calibration (position + orientation) that is more accurate than any single method alone, while maintaining automatic operation.
Solution Approach 2:
The system uses intermediate reference objects (other vessels, land features, water surface) detected in images as mediators to establish geometric relationships. These intermediaries provide measurable reference points that, when combined with sensor data, enable accurate determination of camera extrinsic parameters without requiring direct horizon detection.
3Productivity
If AIS data is used for calibration, then calibration speed is improved, but accuracy deteriorates due to not accounting for vessel extent and GPS position errors
Solution Approach 1:
The system uses image processing to detect actual positions and orientations of reference objects, then compares these with positions predicted from AIS data and sensor measurements. This feedback loop identifies and corrects errors in AIS data (such as GPS position offsets and vessel orientation deviations), continuously refining the calibration accuracy while maintaining the speed benefits of automatic operation.
Data Source
AI summary
A method and system for calibrating a camera which is arranged on an ego vessel are disclosed. The method comprises receiving image data captured by the camera, the image showing at least one imaged solid object and at least one imaged solid object interface between the imaged solid object and the imaged sky, or between the imaged solid object and an imaged water surface of a waterbody. The method further comprises extracting the imaged solid object interface from the image and determining a pose of the ego vessel at the time at which the image is captured. The method additionally comprises extracting a real solid object interface from received reference data. Furthermore, the method comprises determining a difference between the extracted imaged solid object interface and the extracted real solid object interface, and determining calibration values for the camera depending on the difference such that the difference is reduced.