Image Restoration: Mathematical Models, Algorithms, and MATLAB Implementation

This article focuses on methodologies for image restoration. Chapter 14 introduces fundamental concepts and techniques, covering degradation models and algebraic restoration principles alongside practical implementation approaches. Section 14.1 examines image degradation models, including continuous systems (modeled via convolution integrals) and discrete systems (represented by matrix operations). Section 14.2 explores algebraic restoration methods, featuring inverse filtering (direct inversion of degradation operators) and least squares filtering (optimization-based noise suppression). Section 14.3 emphasizes MATLAB implementation, demonstrating Wiener filtering (frequency-domain statistical optimization), regularized filtering (Tikhonov regularization for ill-posed problems), Lucy-Richardson algorithm (iterative maximum-likelihood deconvolution), blind deconvolution (joint estimation of image and PSF), and additional built-in functions like deconvwnr, deconvreg, and deconvblind. Through this comprehensive guide, readers will gain both theoretical understanding and practical skills to restore degraded images using computational approaches.