Volume 29, Issue 5, 2020
DOI: 10.24205/03276716.2020.1029
A semi-supervised feature selection method for image inpainting in Fourier transform domain
Abstract
This paper deals with recovering an image some of which Fourier transform coefficients are lost. Based on a semi-supervised feature selection method, we first present a truncated singular value decomposition approach and second present a new minimization algorithm for image inpainting in the Fourier transform domain. The convergence of this algorithm is also proved. Finally, some comparisons of the given methods and some known algorithms are presented to show the efficiency of our approaches.
Keywords
Image Inpainting, Fourier transform domain, Semi-supervised selection method