Image Processing Apparatus Resolving Polarization Angle Ambiguity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing apparatuses cannot combine images illuminated from arbitrary directions due to the indistinguishability of polarization angles, specifically 0° and 180° or 90° and 270°, which limits their ability to adjust illumination states effectively.
Innovation Solution
An image processing apparatus that acquires surface normal information and captures images while changing the polarization angle, extracting light intensity information and using it to generate images with adjusted illumination states by separating angle-dependent and angle-independent components, allowing for the combination of images from arbitrary directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If polarization angle information is used to change illumination state, then image processing capability is improved, but indistinguishability of 0° and 180° polarization angles prevents combining images from arbitrary directions
Solution Approach 1:
The patent introduces a new dimension by acquiring polarization information at multiple viewpoints (different camera positions) rather than relying solely on single-viewpoint polarization angle measurement. This multi-viewpoint approach adds spatial dimension information that resolves the 180° ambiguity in polarization angle, enabling distinction between directions that would otherwise be indistinguishable.
Solution Approach 2:
The patent uses surface normal information as an intermediary to resolve the polarization angle indistinguishability problem. By comparing the relationship between surface normals at different viewpoints and the light source direction, the system can determine the correct illumination direction even when polarization angles are ambiguous, effectively using surface geometry as a mediator to disambiguate the polarization information.
2Measurement precision
If light intensity information is extracted using surface normal information, then illumination state control is improved, but extraction accuracy depends on precise surface normal acquisition
Solution Approach 1:
The patent performs preliminary acquisition of surface normal information before extracting light intensity information. By pre-computing surface normals from the captured images and storing them for reference, the system prepares the necessary geometric data in advance, which simplifies the subsequent light intensity extraction process and improves measurement precision without significantly increasing overall system complexity.
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
Enables the combination of images illuminated from arbitrary directions, effectively eliminating the indistinguishability of polarization angles and allowing for intuitive adjustment of illumination states, resulting in more natural and controlled image illumination.
Implementation Method 1
acquire a plurality of captured images by capturing an object while changing a polarization angle of light from the object
Data Source
AI summary
An image processing apparatus includes a first processing task configured to acquire surface normal information of an object, a second processing task configured to acquire a plurality of first images acquired by capturing the object while changing a polarization angle of light from the object, a third processing task configured to acquire polarization information of the object, and a fourth processing task configured to generate a second image in which an illumination state of the object is different from that of each first image. The third processing task acquires light intensity information that changes according to the polarization angle, using the first images. The fourth processing task extracts a partial area of the first images using the surface normal information, and generates the second image using the partial area and the light intensity information.