A Semi-implicit Image Denoising Algorithm in Matrix Form
Abstract：The image denoising is one of the essential steps and key links in the process of image pro-cessing. Firstly, the Crank-Nicolson semi-implicit difference scheme is applied to discrete the famous Ru- din-Osher-Fatemi model which was proposed by Rudin et al. in 1992, overcoming the shortcomings of in- stability and many iterative numbers that the explicit discrete scheme has. Secondly, a semi-implicit image denoising algorithm in the matrix form is proposed. In the numerical experiment, we adopt Dirichlet boundary conditions, Periodic boundary conditions and Neumann boundary conditions. The experimental results show that the Crank-Nicolson semi-implicit scheme in matrix form is efficient and the Neumann boundary conditions keep the continuity of boundary.
Keywords：denoisingthe Rudin-Osher-Fatemi modelthe Crank-Nicolson difference schemematrixboundary conditions
Publication Date：2012-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 43-48 )