Title :
A COMPARISON OF THE BILATERAL FILTERAND TV-NORM MINIMIZATION FOR IMAGE DENOISING
Author :
Danda, Swetha ; McGraw, Tim
Author_Institution :
Dept. of Comput. Sci. & Electr. Eng., West Virginia Univ., Morgantown, WV
Abstract :
In this paper we compare two non-linear filters: the bilateral filter (BF) and total variation norm (TV-Norm) minimization. Bilateral filter preserves edges while removing noise by means of non-linear combination of domain and range filters. TV-Norm minimization technique minimizes the total-variation of the input image subject to constraints. We use Mean squared error estimation and L1-norm to evaluate the performance of the two filters. Bilateral filters with different spatial windows and TV-Norm minimization are applied to 2D brain images with different levels of noise. The effect of the parameters and spatial window size on bilateral filter and performance of these two filters with respect to edge preservation is observed
Keywords :
biomedical MRI; brain; image denoising; medical image processing; minimisation; nonlinear filters; MRI images; bilateral filter; domain filter; edge preservation; image denoising; image noise levels; mean squared error estimation; noise removal; nonlinear filters; range filter; total variation-norm minimization; two-dimensional brain images; Anisotropic magnetoresistance; Biomedical imaging; Brain; Computer science; Error analysis; Filtering; Filters; Image denoising; Noise level; Pixel;
Conference_Titel :
Biomedical Imaging: From Nano to Macro, 2007. ISBI 2007. 4th IEEE International Symposium on
Conference_Location :
Arlington, VA
Print_ISBN :
1-4244-0672-2
Electronic_ISBN :
1-4244-0672-2
DOI :
10.1109/ISBI.2007.356942
