Polarization Imaging Parallax Correction via Image Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polarization imaging technologies face challenges in accurately imaging subjects with smooth surfaces due to parallax issues, which hinder the detection of foreign substances or surface irregularities on transparent objects.
Innovation Solution
An image forming device that uses multiple light emitting units with different polarization directions, capturing and processing polarized images from various viewpoints to calculate and correct for parallax, forming an image by averaging and subtracting parallel and crossed nicols images, thereby reducing parallax effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple light emitting units with different polarization directions are used to capture polarized images from various viewpoints, then the visibility of surface features is enhanced, but parallax effects are introduced that reduce imaging precision
Solution Approach 1:
The imaging system is segmented into multiple light emitting units, each with different polarization directions, that capture images from different viewpoints. This segmentation allows the system to gather comprehensive polarization information while enabling subsequent parallax correction through image processing that combines these segmented views.
Solution Approach 2:
The system changes the polarization parameter of illumination light by using light emitting units with different polarization directions. By capturing images with varying polarization states and processing them through averaging and subtraction operations, the system eliminates parallax effects while maintaining enhanced surface feature visibility.
2Measurement precision
If microlenses and polarizing filters are positioned precisely to correspond with image sensor areas, then polarization imaging quality is improved, but device complexity and manufacturing difficulty increase
Solution Approach 1:
Instead of using complex physical structures like microlenses and polarizing filters that require precise positioning, the invention uses software-based image processing to create virtual polarization images. Multiple standard images captured from different viewpoints are processed through averaging and subtraction to generate the equivalent of polarized images, eliminating the need for complex optical components.
3Measurement precision
If multiple polarized images are captured and processed through averaging and subtraction, then surface features are enhanced, but processing time and computational complexity increase
Solution Approach 1:
The system performs preliminary actions by capturing multiple polarized images simultaneously or in rapid succession from different viewpoints before parallax occurs. These pre-captured images are then processed through averaging and subtraction operations to enhance surface features, allowing the enhancement to be achieved without waiting for sequential capture of each polarization state.
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 enables precise imaging of transparent or semitransparent subjects by effectively canceling parallax and enhancing the visibility of surface features, such as foreign substances or surface irregularities, on smooth surfaces like the human eyeball or transparent objects.
Implementation Method 1
a plurality of first light emitting units that illuminate a subject with first illumination light beams polarized in a first direction, a plurality of second light emitting units that illuminate the subject with second illumination light beams polarized in a second direction crossing the first direction
Implementation Method 2
an imaging device having an imaging surface including a first area, which receives first reflection light beams polarized in the first direction, and a second area, which receives second reflection light beams polarized in the second direction
Data Source
AI summary
Images are formed from first and second polarized images acquired through illumination of the subject with first illumination light beams, and third and fourth polarized images acquired through illumination of the subject with second illumination light beams. Images of a subject are formed from translated first, second, third, and fourth polarized images, acquired from first and second luminous point images. The first bright spots appear on the first polarized image when the first illumination light beams are mirror-reflected off the subject. The second bright spots appear on the fourth polarized image when the second illumination light beams are mirror-reflected off the subject. The image forming circuit forms images of the subject by calculating a difference between a first average image, obtained by averaging the translated first and fourth polarized images, and a second average image, obtained by averaging the translated second and third polarized images.