Fiber Bundle Image Reconstruction for Faster Microscopy Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Structured illumination-based microscopes suffer from low image reconstruction efficiency, residual gratings, and low imaging quality due to the need for extensive pixel-by-pixel calculations and grating residues in reconstructed images.
Innovation Solution
An image reconstruction method that calculates gray values at fiber centers and uses spatial interpolation to determine gray values for other pixel points, forming a reconstructed image by using triangular structures based on fiber center positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the root mean square formula is used to calculate gray levels of all pixels in multiple images, then the reconstructed image can be obtained, but the calculation time is excessive and imaging quality is low
Solution Approach 1:
The patent segments the image reconstruction process into two stages: first calculating gray levels only at fiber center positions using the root mean square formula, then using spatial interpolation to obtain gray levels for all other pixels. This segmentation reduces the number of direct RMS calculations from all pixels to only fiber center pixels, significantly reducing calculation time while maintaining imaging quality through the interpolation step that fills in the remaining pixel values.
2Reliability
If the grating is moved to scan multiple images, then sectioning imaging function is achieved, but residual gratings appear in the reconstructed image
Solution Approach 1:
The patent extracts and removes the grating residue component from the reconstructed image through targeted pixel processing. By calculating gray levels at fiber center positions and interpolating to surrounding pixels, the method effectively separates and eliminates the periodic grating pattern artifacts while preserving the genuine biological structure information, thus removing the harmful grating residues while maintaining the sectioning imaging function.
3Measurement precision
If pixel-by-pixel calculation is performed for all pixels in the image, then complete image reconstruction is achieved, but the reconstruction efficiency is low
Solution Approach 1:
The patent divides the pixel set into two groups: fiber center pixels and non-center pixels. Only fiber center pixels undergo the computationally intensive root mean square calculation, while non-center pixels are generated through spatial interpolation. This segmentation maintains complete image reconstruction by eventually providing values for all pixels, but dramatically improves reconstruction efficiency by minimizing the number of expensive RMS operations.
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
Provided are an image reconstruction method, a device and a microscopic imaging device. The method includes calculating a gray value at each fiber center in a fiber bundle (04) in a reconstructed image according to a gray value at a center position of each fiber, determined in one or more sample images; performing a spatial interpolation using the gray value at the fiber center to obtain gray values of other pixel points in the fiber bundle (04) in the reconstructed image, so as to form the reconstructed image. This image reconstruction method only calculates the gray value of the pixel point at the fiber center position, and then obtains gray values of pixel points of the entire image based on the spatial interpolation, thereby reducing the calculation amount due to calculating the gray value of each pixel point, greatly accelerating the speed of image reconstruction, and the method is helpful to remove the grating (022) and fiber bundle (04) cellular grid residues in the reconstructed image and improve the imaging quality of the reconstructed image.