Super-Resolution Image Reconstruction Using Translation Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for super-resolution image reconstruction under total aliasing face high computational complexity due to large dimensional matrices, which degrades performance and requires significant resources, especially when dealing with entire band aliasing.
Innovation Solution
The method employs Fast Fourier Transformation to estimate translation values, Multiple Signal Classification to obtain optimized frequency spectrums, and Inverse Fourier Transformation to reconstruct super-resolution images, using a normalized cross power spectrum technique for updating translational values and gradient-based updates to handle smaller matrices independently of LR and SR image dimensions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional super-resolution image reconstruction methods are used to handle total aliasing, then image resolution is improved, but computational complexity becomes enormously high due to large dimensional matrices
Solution Approach 1:
The patent segments the aliasing problem into frequency domain components using Fast Fourier Transformation. By transforming the image into frequency components and processing them separately, the method avoids handling the entire large-dimensional matrix at once, thereby reducing computational complexity while maintaining resolution improvement
Solution Approach 2:
The patent introduces translation values as an intermediary parameter that captures the relative motion between low-resolution images. This intermediary allows the system to resolve aliasing without directly processing the full high-dimensional reconstruction matrix, reducing computational burden while achieving super-resolution
2Loss of information
If aliasing is included into an image model for super-resolution reconstruction, then high frequency information can be recovered, but the computational complexity becomes enormously high
Solution Approach 1:
The patent replaces the traditional mechanical approach of directly modeling and processing aliased images in the spatial domain with a frequency domain approach using Fast Fourier Transformation. This substitution allows high frequency information to be recovered through spectral analysis while avoiding the computational burden of direct spatial domain modeling
3Measurement precision
If large dimensional matrices are handled in iterative algorithms for super-resolution, then reconstruction accuracy is improved, but resource consumption and processing time increase significantly
Solution Approach 1:
The patent segments the large dimensional matrix into frequency domain components that can be processed independently and in parallel. This segmentation maintains reconstruction accuracy by preserving all frequency information while enabling faster processing through parallel computation of smaller frequency components
Solution Approach 2:
The patent employs periodic updates of translation values during the iterative reconstruction process. By periodically updating translation estimates and using them to guide the reconstruction, the method achieves accurate results with fewer iterations, thereby improving processing speed without sacrificing accuracy
Data Source
AI summary
Systems and methods for reconstructing super-resolution images under total aliasing based upon translation values. Traditional systems and methods provide for extracting super-resolution (SR) image/s from low-resolution (LR) image/s but include aliasing as a background noise which degrades the performance or include aliasing into an image model which leads to an enormously high computational complexity. Embodiments of the present disclosure provide for reconstructing super-resolution images based upon translation values by taking aliasing in consideration by capturing a set of LR images, estimating, using a Fast Fourier Transformation, a set of translation values based upon the set of LR images, obtaining, using a multi-signal classification technique, one or more optimized frequency spectrums based upon the set of translation values and reconstructing one or more SR images based upon the set of translation values.