3D Image De-Skewing Using Planar Surface Angle Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
3D imaging systems based on light triangulation often introduce a skewing effect, causing scanned objects to appear skewed in 3D images, which conventional methods using point reference features struggle to accurately correct.
Innovation Solution
Determine a transform based on detected planar imaged surfaces with known angular relationships, applying it to align the angular relationships between imaged surfaces to match those of the real surfaces, eliminating the need for point reference features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods using point reference features are used for de-skewing, then the method is simpler to implement, but the accuracy and robustness of de-skewing is insufficient
Solution Approach 1:
The patent transitions from using point reference features (0D) to using planar surface features (2D) for de-skewing calibration. By detecting planar surfaces and utilizing their angular relationships, the system achieves more robust and accurate de-skewing results. The planar surfaces provide richer geometric information that constrains the transformation more effectively than point features alone.
Solution Approach 2:
The patent creates a virtual reference model by detecting planar surfaces in the 3D image data and comparing their angular relationships against known angular relationships of real planar surfaces. This virtual reference approach allows the system to determine transformation parameters without requiring physical point reference features, thereby improving accuracy while maintaining computational efficiency.
2Reliability
If point reference features are used for calibration, then the calibration process is straightforward, but position and rotation invariance cannot be ensured
Solution Approach 1:
The patent elevates the calibration approach from point-based (0D) to planar surface-based (2D). Planar surfaces inherently possess orientation and angular properties that are invariant to position and rotation transformations. By detecting the angular relationships between planar surfaces and comparing them against known values, the system can reliably determine transformation parameters that maintain invariance under position and rotation changes.
Solution Approach 2:
The patent changes the calibration parameters from point coordinates to planar surface angular relationships. This parameter transformation enables the system to exploit geometric invariants that remain constant under position and rotation transformations, thereby ensuring reliability of the de-skewing process across different viewing conditions and object positions.
3Measurement precision
If planar surface detection is used for de-skewing, then robustness and accuracy improve, but the algorithm becomes more complex
Solution Approach 1:
The patent segments the calibration process into distinct steps: (1) detecting planar surfaces in the 3D image data, (2) extracting angular relationships between these surfaces, (3) comparing extracted angles against known angular relationships, and (4) determining transformation parameters. This segmentation allows each step to be optimized independently and makes the overall complex algorithm more manageable and implementable.
Solution Approach 2:
The patent introduces planar surface angular relationships as an intermediary between the raw 3D image data and the final transformation parameters. Instead of directly computing transformations from complex geometric data, the system uses angular relationships of planar surfaces as an intermediate representation that simplifies the measurement and comparison process, thereby reducing algorithmic complexity while maintaining high precision.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides more robust and accurate de-skewing of 3D image data, ensuring position and rotation invariance, scale invariance, and simpler algorithms, while avoiding the inaccuracies of conventional methods.
Implementation Method 1
3D machine vision systems or devices may be referred to as systems or devices for 3D imaging based on light, or light plane, triangulation, or simply laser triangulation when laser light is used
Implementation Method 2
the reflected light from the object is captured by an image sensor of a camera
Data Source
AI summary
Method and device(s) for determining a transform having a de-skewing effect on 3D image data resulting from scanning by a 3D imaging system. 3D image data is obtained from scanning by the 3D imaging system (605) of planar real surfaces that at least during the scanning were non-parallel with constant one or more angular relationships between each other, whereby said 3D image data comprises planar imaged surfaces corresponding to said planar real surfaces. The planar imaged surfaces are detected in the 3D image data. The transform is determined based on the detected planar imaged surfaces and said one or more angular relationships between the planar real surfaces, such that the transform, when applied to a description of the detected imaged surfaces in coordinates of said 3D data, results in that angular relationships between the detected imaged surfaces match said angular relationships between the planar real surfaces.