An Implementation and Detailed Analysis of the K-SVD Image Denoising Algorithm

K-SVD is a signal representation method which, from a set of signals, can derive a dictionary able to approximate each signal with a sparse combination of the atoms. This paper focuses on the K-SVD-based image denoising algorithm. The implementation is described in detail and its parameters are anal...

Full description

Bibliographic Details
Main Authors: Marc Lebrun, Arthur Leclaire
Format: Article
Language:English
Published: Image Processing On Line 2012-05-01
Series:Image Processing On Line
Online Access:http://www.ipol.im/pub/art/2012/llm-ksvd/