Vehicle Camera Online Calibration Using Segmented Geometric Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online camera calibration systems for vehicles require significant processing resources and may suffer from accuracy issues due to the simultaneous calibration of rotational and translational parameters, leading to data ambiguity.
Innovation Solution
A method that separates the calibration of camera rotational parameters (ϕ, θ, ψ) and translational parameters (x, y, z) using a parallel geometrical method, calibrating them independently to avoid data ambiguity and reduce processing resource requirements, with the translational parameters calibrated while the vehicle is moving along a curved path and rotational parameters calibrated while traveling in a straight line.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If rotational and translational parameters are calibrated simultaneously, then calibration can be performed in a single process, but data ambiguity occurs and accuracy decreases
Solution Approach 1:
The calibration process is segmented into two distinct stages: first calibrating rotational parameters (pitch, roll, yaw) using parallel line features, then calibrating translational parameters (x, y, z positions) using the already-calibrated rotational parameters. This segmentation eliminates data ambiguity while maintaining process efficiency.
2Device complexity
If all calibration parameters are determined using a single method, then the process is simplified, but processing resources are excessively consumed
Solution Approach 1:
The calibration methodology is segmented into two distinct approaches: parallel geometrical method for rotational parameters and triangular geometry method for translational parameters. Each method is optimized for its specific parameters, reducing overall processing resource consumption while maintaining methodological simplicity.
Solution Approach 2:
The parallel line features in the calibration pattern serve as intermediaries that first establish rotational parameter relationships, which then serve as the foundation for determining translational parameters. This intermediary approach reduces the computational complexity of the overall calibration process.
3Adaptability or versatility
If calibration is performed whilst the vehicle is moving, then online calibration is achieved, but accuracy may be compromised by vehicle motion
Solution Approach 1:
The calibration system uses the vehicle's own motion and the calibration pattern's geometric features (parallel lines) to automatically determine both rotational and translational parameters. The system self-calibrates by exploiting the inherent geometric relationships in the calibration pattern without requiring external reference systems.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for online calibration of a camera (2, 3, 4) on a vehicle (1) is provided. The camera (2, 3, 4) has camera rotational parameters, ϕ, θ, ψ, and camera translational parameters xc, yc, zc in a camera image coordinate system and the vehicle has a vehicle coordinate system. The method comprises, whilst the vehicle (1) is moving relative to the ground, calibrating the camera rotational parameters, ϕ, θ, ψ, using a parallel geometrical method and calibrating the camera translational parameters xc, yc independently of the camera rotational parameters ϕ, θ, ψ.