Efficient Image Restoration Using Generalized Non-Convex Total Variation Regularization and Chebyshev-Accelerated ADMM
This repository contains the implementation of an efficient image restoration framework based on a generalized non-convex TV-$\ell_p$ regularization model combined with a Chebyshev-accelerated ADMM optimization scheme.
The proposed method is designed to achieve high-quality image deblurring while preserving important edge structures, offering improved performance over classical convex and non-convex approaches.
- MATLAB (R2018a or later recommended)
- Image Processing Toolbox
Ensure that all files are placed in a single directory before running the code: