Home > detect > radial_mean.m

radial_mean

PURPOSE ^

Detection using Radial Mean Kernel.

SYNOPSIS ^

function imConfidence = radial_mean(im,kernelSize)

DESCRIPTION ^

Detection using Radial Mean Kernel.

See my Master Thesis Detection part.

INPUT
 im               - image 
 kernelSize       - size of the radial mean kernel

OUTPUT
 imConfidence     - the score respond image

SEE ALSO imfilter

EXAMPLE
 imOut =  radial_mean(im,30);

Written by Rich Nguyen (rich.uncc@gmail.com)
Version 1.0, Feb 2010

CROSS-REFERENCE INFORMATION ^

This function calls: This function is called by:

SOURCE CODE ^

0001 function imConfidence = radial_mean(im,kernelSize)
0002 %Detection using Radial Mean Kernel.
0003 %
0004 %See my Master Thesis Detection part.
0005 %
0006 %INPUT
0007 % im               - image
0008 % kernelSize       - size of the radial mean kernel
0009 %
0010 %OUTPUT
0011 % imConfidence     - the score respond image
0012 %
0013 %SEE ALSO imfilter
0014 %
0015 %EXAMPLE
0016 % imOut =  radial_mean(im,30);
0017 %
0018 %Written by Rich Nguyen (rich.uncc@gmail.com)
0019 %Version 1.0, Feb 2010
0020 
0021 %% Construct filter kernels
0022 rm_k_max = 2; % number layer of radial kernel
0023 kh = floor(kernelSize/2);
0024 temp = zeros (kernelSize,kernelSize);
0025 temp(kh+1,kh+1) = 1;
0026 rm_k = floor(bwdist(temp));
0027 max_dist = max ( rm_k(:) );
0028 f = double(rm_k_max+1) / double ( floor(max_dist)+1 );
0029 rm_k = floor( rm_k * f );
0030 
0031 
0032 %% Compute radial mean kernels
0033 kernelInnerRing  = double(rm_k == 0);
0034 kernelMiddleRing = double(rm_k == 1);
0035 kernelOutterRing = double(rm_k == 2);
0036 
0037 
0038 %% Filter for all 3 rings
0039 im = double(im);
0040 r1 = imfilter(im,kernelInnerRing,'replicate');
0041 r2 = imfilter(im,kernelMiddleRing,'replicate');
0042 r3 = imfilter(im,kernelOutterRing,'replicate');
0043 
0044 
0045 %% Combine 3 responds
0046 imConfidence = r1 ./ (r1 + r2 + r3);
0047

Generated on Thu 17-Mar-2011 14:45:51 by m2html © 2005