Fourier Domain Deconvolution Using NPOTF for Super-Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical systems are limited by diffraction, making it challenging to achieve high resolution images with small apertures, and existing deconvolution methods face issues with noise and computational complexity.
Innovation Solution
A deconvolution method using neighboring-pixel-optical-transfer-function (NPOTF) in the Fourier domain, which calculates initial correlation coefficients, applies discrete Fourier transforms, and generates modified Fourier components to enhance image resolution beyond the diffraction limit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a small aperture optical system is used, then the system cost and complexity are reduced, but the image resolution is limited by diffraction
Solution Approach 1:
The patent introduces an intermediary computational process (deconvolution algorithm using NPOTF in Fourier domain) that mediates between the diffraction-limited optical system and the final image output. This computational intermediary processes the blurred image data to reconstruct higher resolution information, effectively decoupling the optical hardware limitations from the final image quality.
Solution Approach 2:
The patent replaces the mechanical/optical approach of increasing aperture size with a computational approach. Instead of physically enlarging the aperture to improve resolution, the invention uses digital signal processing techniques (Fourier transform, deconvolution) to achieve super-resolution, substituting computational complexity for optical complexity.
2Measurement precision
If conventional deconvolution methods are used, then image resolution can be improved, but noise levels become unacceptable and computational complexity increases
Solution Approach 1:
The patent changes the parameter representation from spatial domain to frequency domain (Fourier domain). By performing deconvolution in the frequency domain using NPOTF, the method transforms the problem into a form where resolution enhancement can be achieved with better control over noise amplification, as the frequency domain operation allows for selective enhancement of different spatial frequencies.
3Measurement precision
If conventional deconvolution methods are used, then image resolution can be improved, but computational complexity and processing time increase
Solution Approach 1:
The patent replaces complex iterative deconvolution algorithms with a more efficient direct computation method in the frequency domain. By using the Fourier transform properties and NPOTF, the computationally intensive spatial domain deconvolution is substituted with simpler frequency domain operations, reducing processing time and computational complexity while maintaining resolution enhancement capabilities.
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
The method effectively increases image resolution beyond the diffraction limit by establishing a one-to-one relation between Fourier components of the original and detected images, resulting in sharper images with acceptable noise levels.
Implementation Method 1
applying discrete Fourier transform (FT) to the input image
Implementation Method 2
applying inversed discrete FT to these new Fourier components
Data Source
AI summary
The present disclosure includes an image processing technique that is capable to increase spatial resolution of single frame images beyond diffraction limit. If an image is taken by diffraction limited optical system with regularly spaced pixel detectors, and if the spacing of pixels of the detectors is much small than the diffraction pattern, then the spatial resolution of the image can be increased beyond the diffraction limit by using neighboring-pixel-optical-transfer-function (NPOTF) in Fourier Domain with periodical boundary conditions.