MFA Image Processing Using Quincuncial Pattern Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multispectral filter array (MFA) technologies result in low-resolution images when simultaneously capturing visible and near-infrared images, due to the sampling of color channels and near-infrared channels, leading to limited resolution and artifacts.
Innovation Solution
An image processing apparatus and method that converts a low-resolution MFA pattern image into a high-resolution image by using an adaptive interpolation device to convert the MFA pattern into a quincuncial pattern, interpolating color and near-infrared channels to maximum resolution, and employing a frequency compensation device to enhance high-frequency components, while a channel interference suppression device removes color distortion using weighted averages and linear regression analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a multispectral filter array (MFA) is used to simultaneously obtain visible and near-infrared images, then both image types can be captured at the same time, but the resolution of each image becomes low due to sampling
Solution Approach 1:
The patent segments the image processing into multiple stages: first converting MFA pattern to quincuncial pattern, then performing adaptive interpolation to separate color and NIR channels, and finally applying frequency compensation. This segmentation allows each processing stage to optimize for its specific function, resolving the resolution loss from sampling while maintaining simultaneous acquisition capability.
Solution Approach 2:
The patent transforms the MFA pattern (which samples color and NIR channels at the same spatial locations) into a quincuncial pattern that effectively creates offset sampling grids. This dimensional transformation allows the system to recover high-frequency information that would otherwise be lost, improving resolution while maintaining the simultaneous capture of multiple spectral bands.
2Quantity of substance
If color channels and NIR channel are sampled using MFA, then multispectral information is obtained, but high-frequency information is lost and artifacts are generated
Solution Approach 1:
The patent applies preliminary frequency compensation before final image reconstruction. By compensating for high-frequency losses in the quincuncial pattern intermediate representation, the system recovers fine details and reduces artifacts before the final color and NIR channel separation, preserving information that would otherwise be lost in the sampling process.
Solution Approach 2:
The quincuncial pattern serves as an intermediary representation that bridges the MFA sampled data and the final high-resolution color/NIR images. This intermediate form allows frequency compensation to be applied in a way that recovers high-frequency information while the adaptive interpolation device then separates the channels, preserving spectral information without the artifacts that would result from direct interpolation.
3Ease of manufacture
If simple interpolation is used to upscale MFA images, then processing is simple and fast, but image quality deteriorates with artifacts and reduced high-frequency content
Solution Approach 1:
The patent employs adaptive interpolation that dynamically adjusts its behavior based on local image characteristics. The interpolation device analyzes gradient directions and magnitudes in different regions, applying appropriate filtering and compensation techniques adaptively. This dynamic approach maintains processing efficiency while significantly improving image quality compared to static interpolation methods, preserving edges and high-frequency content without excessive complexity.
Data Source
AI summary
An image processing apparatus includes an adaptive interpolation device which converts a MFA pattern image into a quincuncial pattern image based on difference values, and interpolates color channels and an NIR channel, based on difference values of the converted quincuncial pattern image in vertical and horizontal pixel directions; a frequency compensation device which obtains a high-resolution MFA image using high-frequency and medium-frequency components of a high-resolution base image, based on linear regression analysis and compared energy levels of MFA channel images to an energy level of a base image; and a channel interference suppression device which removes color distortion generated between each channel of the high-resolution MFA image, and another channel of the high-resolution MFA image and a base channel using a weighted average of pixel value differences between each channel of the high-resolution MFA image, and the other channel of the high-resolution MFA image and the base channel.