PCNN Matlab显微图像处理

%------Pulse Coupled Neural Network-----------

%------名称：脉冲耦合神经网络（PCNN）

function H = PCNN2(I)

I = imread(‘BanLanGen.png‘);

figure,imshow(I);

title(‘原始图像‘);

th = graythresh(I);

I = im2bw(I,th);

%H = Pcnn(I,30);

%************************************************************

% Initialize Function

%************************************************************

[m,n] = size (I);

figure,imshow(I);

title(‘二值图像‘);

X = im2double(I);

%************************************************************

% Initialize PCNN Parameters

%************************************************************

% P ：L E F VF VL VE beta

% V: 1.0 1.0 0.1 0.5 0.2 20 0.1

al = 1.0; ae = 1.0; af = 0.1; vf = 0.5; vl = 0.2; ve = 20; B = 0.1;

W =[0.5 1 0.5;...

1 0 1;...

0.5 1 0.5];

M = W; Y = zeros(m,n); F = Y; L = Y; U = Y; E = Y;

%************************************************************

% PCNN 点火过程

%************************************************************

for i = 1:30

wk = conv2(Y,M,‘same‘);

F = exp(-af).* F + vf.* wk + X;

L = exp(-al).* L + vl.* wk;

U = F.*(1 + B.* L);

Y = double(U>E);

E = exp(-ae).* E + ve.* Y;

end

figure,imshow(Y)

title(‘PCNN处理后图像‘);

fs = fspecial(‘sobel‘);

filter_Y = imfilter(Y, fs);

figure,imshow(filter_Y);

title(‘滤波图像‘);