Sensor Self-Calibration via Object Detection and Analysis-by-Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sensors on apparatuses like autonomous vehicles often become misaligned due to vibrations and environmental factors, making frequent controlled environment calibrations impractical, necessitating a method for self-calibration in real-world conditions.
Innovation Solution
A system comprising an electronic control unit and multiple sensors (image and depth sensors) that capture overlapping environmental data, identify objects with known dimensions, and calibrate sensors based on comparisons with database parameters, adjusting intrinsic and extrinsic parameters to correct misalignments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensors are calibrated in a controlled environment with calibration objects, then calibration accuracy is improved, but the complexity and impracticality increase when calibration facilities are not available in the wild
Solution Approach 1:
The system performs self-calibration by using its own sensors to capture images of objects in the environment, automatically identifying objects through object detection, retrieving their ground truth dimensions from a database, and computing calibration parameters without requiring external calibration facilities or manual intervention
Solution Approach 2:
Instead of using specially designed calibration objects in controlled environments, the system inverts the approach by using ordinary objects found in the wild environment as calibration targets, leveraging their known dimensions from databases to perform calibration in practical operating conditions
2Reliability
If controlled environment calibration is performed periodically, then sensor alignment is maintained, but the time loss and operational interruption increase
Solution Approach 1:
The system performs calibration actions in advance or continuously in the background during normal operation, rather than interrupting operations for periodic calibration, by automatically detecting objects and computing calibration parameters whenever suitable objects are captured in sensor images
Solution Approach 2:
The calibration process continues continuously or repeatedly during normal system operation without interruption, allowing the system to maintain sensor alignment while constantly capturing images and performing calibration computations in the background
3Measurement precision
If frequent calibration is performed to compensate for vibrations and environmental factors, then sensor accuracy is maintained, but the productivity and operational efficiency decrease
Solution Approach 1:
The system performs calibration periodically or repeatedly at intervals during operation rather than continuously, triggering calibration computations only when suitable objects are detected in the environment, thus balancing accuracy maintenance with operational efficiency
Data Source
AI summary
A system for self-calibrating sensors includes an electronic control unit, a first image sensor and a second image sensor communicatively coupled to the electronic control unit. The electronic control unit is configured to obtain a first image and a second image, where the first image and the second image contain an overlapping portion, determine an identity of an object present within the overlapping portion, obtain parameters of the identified object, determine a miscalibration of the first image sensor or the second image sensor based on a comparison of the identified object in the overlapping portions and the parameters of the identified object, in response to determining a miscalibration of the first image sensor or the second image sensor, calibrate the first image sensor or the second image sensor based on the parameters of the identified object and the second image or the first image, respectively.