3D Acquisition System Calibration via Projector Matrix Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Classical calibration methods for one-dimensional structured light active stereo 3D acquisition systems face challenges in accurately determining the relation between a non-calibrated projector and a pre-calibrated camera, particularly due to manufacturing imperfections and the sensitivity of the reconstruction process to angular misalignments, which complicates the calibration procedure and requires multiple views of a calibration object.
Innovation Solution
A novel calibration correction approach is introduced, which includes a coarse calibration stage followed by refinement, using a known planar calibration object to establish system parameters, and a modified technique to decompose the projector matrix into intrinsic and extrinsic parameters, allowing for refinement of these parameters to minimize geometric reconstruction error, even with limited views.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If classical calibration methods are used to determine the relation between projector and camera, then the calibration procedure can be performed, but the accuracy is reduced due to manufacturing imperfections and angular misalignments
Solution Approach 1:
The patent applies preliminary action by performing a coarse calibration stage before the fine calibration stage. The coarse calibration pre-establishes the projector-camera relation using a planar calibration object, which then serves as the foundation for the subsequent fine calibration that refines parameters to minimize geometric reconstruction error. This preliminary calibration reduces the sensitivity to manufacturing imperfections in the final precision calibration.
Solution Approach 2:
The patent employs parameter changes by decomposing the projector matrix into intrinsic and extrinsic parameters, then selectively refining these parameters in the fine calibration stage. By changing and optimizing specific parameters (focal length, principal point, distortion coefficients) separately from the overall calibration, the system achieves higher accuracy while compensating for manufacturing imperfections in projector-camera alignment.
2Measurement precision
If multiple views of a calibration object are used in classical calibration, then the calibration can be performed, but the calibration time and complexity increase
Solution Approach 1:
The patent extracts and utilizes a planar calibration object with known geometry to establish the projector-camera relation. By taking out the essential calibration information from a simplified planar object rather than requiring complex multi-view calibration of general objects, the method achieves accurate calibration with reduced time and complexity while maintaining measurement precision.
3Device complexity
If the projector matrix is not decomposed into intrinsic and extrinsic parameters, then the calibration is simpler, but the ability to correct manufacturing imperfections is reduced
Solution Approach 1:
The patent applies segmentation by decomposing the projector matrix into intrinsic parameters (focal length, principal point) and extrinsic parameters (rotation, translation). This segmentation allows independent refinement of each parameter component in the fine calibration stage, enabling targeted correction of manufacturing imperfections while keeping the overall procedure manageable through systematic parameter optimization.
Data Source
AI summary
A camera intrinsic calibration may be performed using an object geometry. An intrinsic camera matrix may then be recovered. A homography is fit between object and camera coordinate systems. View transformations are finally recovered.