Camera Extrinsic Calibration via Runtime Image Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera calibration methods rely on static intrinsic parameters and offline calibration, which are inadequate for dynamic extrinsic parameter changes due to mechanical influences, leading to performance issues in driving-assistance systems.
Innovation Solution
A method and device for determining extrinsic calibration parameters of a camera using image transformation matrices based on feature point correspondence and epipolar geometry, allowing for self-calibration during runtime by analyzing changes in camera position and orientation relative to a world coordinate system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If offline calibration with known calibration objects is used, then initial calibration parameters can be obtained, but the extrinsic parameters cannot adapt to mechanical influences during runtime
Solution Approach 1:
The patent implements dynamic self-calibration of extrinsic parameters during runtime by continuously tracking feature points in image sequences and updating calibration parameters based on detected camera motion, allowing the system to adapt to mechanical influences such as vibrations and mounting shifts that occur during vehicle operation
Solution Approach 2:
The system performs automatic self-calibration without requiring external calibration objects or manual intervention. The camera system uses its own captured image sequences and internal feature point detection algorithms to autonomously determine and update its extrinsic parameters, making the calibration process self-sufficient and continuously adaptive
2Device complexity
If static intrinsic parameters are used, then the camera model is simple, but it cannot account for extrinsic parameter changes during use
Solution Approach 1:
The patent introduces dynamic updating of extrinsic parameters while maintaining static intrinsic parameters. The system selectively updates only the extrinsic parameters (camera position and orientation) based on runtime image analysis, keeping the intrinsic parameters (focal length, optical center) fixed to preserve model simplicity while adapting to mechanical changes
3Measurement precision
If conventional offline calibration is performed, then initial calibration parameters are established, but continuous adaptation to camera position changes is not possible
Solution Approach 1:
The patent implements continuous calibration updating during runtime by processing image sequences in real-time. The system continuously detects feature points, computes camera motion, and updates extrinsic parameters without interruption, ensuring the calibration remains current throughout vehicle operation rather than being a one-time offline process
Data Source
AI summary
A method of determining calibration parameters of a camera comprises taking a first image of an object, taking a second image of the object, wherein the position of the camera with respect to the object is changed between the first and the second image, the calibration parameters of the camera being fixed between the first and the second image, determining a transformation that is adapted to transform a portion of the first into a corresponding portion of the second image, and determining the calibration parameters from the transformation.