Weickert's coherence-enhancing filter

An anisotropic diffusion that enhances coherence of lines and flow-like textures

Description

Weickert's coherence-enhancing filter [1] is an anisotropic diffusion process that enhances the coherence of lines and flow-like textures. It is particularly effective in enhancing fingerprint images (see screenshot). The process removes noise and closes small gaps in the ridges while leaving the locations of minutiae points unchanged.

B = COHERENCEFILTER(A,T) applies the coherence-enhancing filter to a grayscale or color image A with diffusion time T. Larger T implies stronger enhancement.
 
 COHERENCEFILTER(...,'PARAM1',VALUE1,'PARAM2',VALUE2,...) can be used to specify additional method parameters.
 
For example, the screenshot is reproduced by
A = double(imread('fingerprint.png'))/255;
B = coherencefilter(A,4,'sigma',1.5,'rho',5);
image(min(max(B,0),1)); axis image
 
[1] Weickert, J. 1996. Anisotropic Diffusion in Image Processing. Ph.D. Thesis, Dept. of Mathematics, University of Kaiserslautern, Germany.

• MATLAB release: MATLAB 7.5 (R2007b)
• Other Files: coherencefilter.m, ; fingerprint.jpg, ; license.txt

[Chen yuan]

Good programs and works well.It helps me a lot,Thanks.