Assessment of Effectiveness in Image deblurring using Blind Deconvolution Methods

DOI: https://doie.org/10.10399/JBSE.2025892966

Sudharshan Banakar, Chandrashekhar K

Keywords:

Blind Deconvolution, Image Deblurring, Convolutional Neural Networks, Point Spread Function

Abstract:

Blind deconvolution is a critical image restoration technique aiming to recover latent sharp images from blurred optical images without prior knowledge of the point spread function (PSF). With advances in deep learning, convolution-based blind deconvolution methods have shown significant promise in addressing challenges posed by complex optical degradations in various advanced imaging applications like natural images. This study assesses the effectiveness of these methods by reviewing state-of-the-art algorithms, analyzing their mathematical foundations, and demonstrating their performance through experimental metrics like PSNR and SSIM. This study further provides details of current challenges, future research directions.

PDF