Polarized Image Rendering for Smooth-Contour Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to detect abnormalities with small changes in luminance or color information, particularly when the contour of a deformed object is smooth, making it difficult to identify defects.
Innovation Solution
An information processing device and method that utilizes polarization rendering to generate a polarized rendered image by setting parameters such as light source, geometry, material, and camera parameters, and detects abnormalities based on differences between captured and rendered images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If luminance information or color information from captured images is used to detect abnormalities, then the detection method is simple and based on available image data, but abnormalities with small changes in luminance or color information cannot be detected
Solution Approach 1:
The invention changes the parameter used for abnormality detection from luminance/color information to polarization information. By capturing images with different polarization directions and comparing polarization states between captured and rendered images, the system can detect abnormalities that do not manifest in traditional luminance or color channels, thereby improving detection precision for subtle defects.
Solution Approach 2:
The invention adds a new dimension to image analysis by incorporating polarization direction as an additional parameter. Instead of relying solely on intensity (luminance) or wavelength (color) dimensions, the system utilizes the polarization dimension to extract abnormality information, enabling detection of defects that are invisible in conventional imaging modes.
2Reliability
If traditional luminance-based detection is used, then the processing is computationally simple, but it fails to detect abnormalities with smooth contours and minimal luminance variation
Solution Approach 1:
The invention creates a rendered image that copies the expected appearance of the target object under specific polarization conditions. By comparing the captured polarization image with this synthesized copy, the system can reliably detect deviations indicating abnormalities. The rendering process simulates how the object should appear given its geometric model and material properties under polarized light.
Solution Approach 2:
The polarization rendering process acts as an intermediary that translates geometric and material properties into expected polarization image characteristics. This intermediary step enables the system to bridge the gap between object properties and observed polarization patterns, improving detection reliability while managing system complexity through algorithmic processing.
3Measurement precision
If polarization rendering with multiple parameters is implemented, then detection capability for subtle abnormalities is improved, but the complexity of parameter measurement and optimization increases
Solution Approach 1:
The invention performs preliminary measurement and optimization of polarization parameters (light source characteristics, material polarization properties, camera parameters) before the actual abnormality detection process. By pre-characterizing these parameters, the system reduces the complexity during runtime detection, as the optimized parameters can be reused for multiple detection operations without repeated measurement and optimization cycles.
Solution Approach 2:
The system employs feedback mechanisms where the difference between captured and rendered polarization images is used to refine parameter optimization. The detection results feed back into the parameter optimization process, allowing the system to learn from detection outcomes and improve parameter accuracy over time, thereby managing complexity through iterative refinement.
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
Effectively identifies subtle abnormalities by generating a polarized rendered image that highlights differences, enabling precise detection of defects in objects with minimal luminance or color changes.
Implementation Method 1
a polarized captured image acquisition unit that acquires a polarized captured image by imaging the abnormality detection target
Data Source
AI summary
Provided is a polarization rendering setting unit that sets a plurality of parameters to be used for generating a polarized rendered image of an abnormality detection target. A polarized rendered image generation unit generates the polarized rendered image of the abnormality detection target on the basis of the parameters set by the polarization rendering setting unit. An abnormality detection unit detects an abnormal region of the abnormality detection target on the basis of a difference between a polarized captured image acquired by imaging the abnormality detection target and the polarized rendered image generated by the polarized rendered image generation unit. Abnormalities that are difficult to detect on the basis of luminance information and color information becomes able to be detected on the basis of polarized images.