Super-Resolution Reconstruction Demonstration

Project Overview

This project implements a demonstration of super-resolution reconstruction based on sub-pixel shifts. The method enhances image resolution by combining multiple low-resolution images with slight pixel shifts. Currently, the projects generates n*n low pixel Shepp-Logan phantoms with subpixel shifts in both directions, depending on the inputs.

Usage Instructions

Steps to Run:

1. Open twoD.mlx in MATLAB.
2. Set the following parameters:
   • larger_pixel_in_FoV: Number of pixels on each side before reconstruction.
   • compressing_factor: Defines how much finer the resolution should be.

Example Configuration:

larger_pixel_in_FoV=100;
compressing_factor=3;

It generates 9 images of size 100×100. Reconstructs a 300×300 super-resolved image.

Future updates

The project is currently under construction to include additional options: Noise configuration, input image selection, shifting direction control, etc.