Title :
Laplacian Based Non-Linear Diffusion Filtering
Author :
Nishiguchi, Haruhiko ; Imiya, Atsushi ; Sakai, Tomoya
Author_Institution :
Sch. of Sci. & Technol., Chiba Univ.
Abstract :
This paper aims to introduce a diffusion filtering based on the Laplacian map. Classical nonlinear diffusion filtering using the gradient-map-controlled local diffusivity. The Laplacian maps has similar geometric properties with the gradient map for the extraction of the region boundaries. Laplacian-based diffusion function has the same property with the Perona-Malik type and the Weickert type diffusion functions for the small scale. However, for the large scale, the diffusion operation has similar geometrical properties with the linear diffusion filtering. Therefore, the filtering operation in this paper provides a method for the combination of hierarchical expression based on linear and nonlinear diffusion filtering operations
Keywords :
Laplace equations; feature extraction; filtering theory; image segmentation; Laplacian map; Laplacian-based diffusion function; Perona-Malik type diffusion functions; Weickert type diffusion functions; classical nonlinear diffusion filtering; geometrical properties; gradient-map-controlled local diffusivity; region boundary extraction; Filtering; Image edge detection; Image segmentation; Kernel; Laplace equations; Large-scale systems; Nonlinear filters; Pattern recognition; Smoothing methods; Tensile stress;
Conference_Titel :
Pattern Recognition, 2006. ICPR 2006. 18th International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
0-7695-2521-0
DOI :
10.1109/ICPR.2006.742