/imaging

Image deblurring

Primary LanguageJupyter NotebookMIT LicenseMIT

Imaging

Introduction

Project for the Numerical Computing course at the University of Bologna (A.Y. 2021-2022).

Description

The assignment is to implement and analyze the behaviour of the image deblurring problem expressed as a minimum least square problem.
The Conjugate gradient and the Gradient descent method are used to minimize the loss function.
Tikhonov and Total variation regularization are used to improve the result.

Usage

The implementation and the experiments are in the progetto.ipynb notebook.
To install the dependencies run:

pip3 install -r requirements.txt