NOISE REMOVAL APPARATUS USING AN ORIENTED LAPLACIAN CAPABLE OF REDUCING THE EDGE DISPERSION OF AN INPUT IMAGE BY APPLYING A TOTAL VARIATION (TV) MODEL TO THE INPUT IMAGE AND A METHOD THEREOF

摘要

PURPOSE: A noise removal apparatus using an oriented Laplacian and a method thereof are provided to apply the split Bregman algorithm whose solution converges within five repetitions, thereby removing noise precisely and quickly.;CONSTITUTION: A Laplacian calculation unit (110) calculates the area oriented Laplacian of an input image including noise which has a Gaussian density distribution by considering the orientation between each pixel of the input image and the neighboring pixels. A likelihood function calculation unit (120) calculates a likelihood function by applying the calculated area oriented Laplacian. A noise removal unit (130) removes noise by applying a Total Variation (TV) model which repeatedly calculates a solution which minimizes minus log likelihood using the split Bregman algorithm.;COPYRIGHT KIPO 2013;[Reference numerals] (110) Laplacian calculation unit; (120) Likelihood function calculation unit; (130) Noise removal unit