字母识别程序，采用了bp网络，具有一定的模糊识别能力：
%recognize.m用于字母识别
[alphabet,targets] = prodat;
P =alphabet; % alphabet = [ letterA, letterB, ..., letterZ ];
T = targets; % targets = eye(26);
[R,Q] = size(P);
% 神经元个数
S1 = 10;
% 求T的大小，此时他的列数与P的列数为一致
[S2,Q] = size(T);
%创建网络
net = newff(minmax(alphabet),[S1 S2],{'logsig' 'logsig'},'traingdx');
net.LW{2,1} = net.LW{2,1}*0.01;
net.b{2} = net.b{2}*0.01;
% 无噪声字母识别网络的训练
net.performFcn = 'sse'; %执行函数为误差平方和函数
net.trainParam.epochs=800; %最大训练步长
net.trainParam.lr=0.01; %学习率
net.trainParam.lr_inc =1.05; %增长的学习率
net.trainParam.lr_dec =0.7;
net.trainParam.goal=0.1; %执行函数目标值
net.trainParam.mc =0.9; %附加动量因子
net.trainParam.min_grad=1e-10; %最小执行梯度
net.trainParam.show = 50;
% P = alphabet;
% T = targets;
[net,tr] = train(net,P,T);
noisyJ = alphabet(:,10)+randn(35,1) * 0.2;
A2 = sim(net,noisyJ);
A2 = compet(A2);
% compet is a transfer function. Transfer functions calculate a layer's output from its net input.
% compet(N) takes one input argument,
% N - S x Q matrix of net input (column) vectors.
% and returns output vectors with 1 where each net input vector has its maximum value, and 0 elsewhere
answer = find(compet(A2) == 1);
% indices = find(X) returns the linear indices corresponding to the nonzero entries of the array X.
% If none are found, find returns an empty, 0-by-1 matrix.
% In general, find(X) regards X as X(:),
% which is the long column vector formed by concatenating the columns of X.
% Some operations on a vector x = [11 0 33 0 55]';
% find(x)
%
% ans =
%
% 1
% 3
% 5
%
% find(x == 0)
%
% ans =
%
% 2
% 4
%
% find(0 < x & x < 10*pi)
%
% ans =
%
% 1
figure(2);
plotchar(alphabet(:,answer));
figure(3);
plotchar(noisyJ);
% %%%%%%%%%%%%%%%=========================================================
%%%%%%%%%%% 具有噪声的输入识别网络的训练，共循环10次，开始
netn = net;
netn.trainParam.goal = 0.1; %目标误差.
netn.trainParam.epochs = 300; %最大训练步长.
for pass=1:10
fprintf('Pass = %.0f/n',pass);
T = [targets targets targets targets];
P = [alphabet, alphabet, ...
(alphabet + randn(R,Q)*0.1), ...
(alphabet + randn(R,Q)*0.2)];
[netn,tr] = train(netn,P,T);
end
%%%%%%%%%%% 具有噪声的输入识别网络的训练，共循环10次，结束
noisyJ = alphabet(:,10)+randn(35,1) * 0.2;
B2 = sim(netn,noisyJ);
B2 = compet(B2);
answer = find(compet(B2) == 1);
figure(4);
plotchar(alphabet(:,answer));
figure(5);
plotchar(noisyJ);
%%%%%%% 再次用无噪声字母训练识别网络，开始
netn.trainParam.goal = 0.1; % 均方误差目标.
netn.trainParam.epochs = 500; % 最大训练步长
net.trainParam.show = 5; % 训练中显示的频率
P = alphabet;
T = targets;
[netn,tr] = train(netn,P,T);
%%%%%%% 再次用无噪声字母训练识别网络，结束
prodat.m程序用来产生26个字母的矩阵形式，及其所对应的目标矩阵表示形式，具体如下：
% prodat.m recognize.m调用的字母产生程序
function[alphabet,targets] = prodat()
% Returns:
% ALPHABET - 35x26 matrix of 5x7 bit maps for each letter.
% TARGETS - 26x26 target vectors.
letterA = [0 0 1 0 0 ...
0 1 0 1 0 ...
0 1 0 1 0 ...
1 0 0 0 1 ...
1 1 1 1 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ]';
letterB = [1 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 0 ]';
letterC = [0 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 1 ...
0 1 1 1 0 ]';
letterD = [1 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 0 ]';
letterE = [1 1 1 1 1 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 1 1 1 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 1 1 1 1 ]';
letterF = [1 1 1 1 1 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 1 1 1 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ]';
letterG = [0 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 1 1 ...
1 0 0 0 1 ...
0 1 1 1 0 ]';
letterH = [1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ]';
letterI = [0 1 1 1 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 1 1 1 0 ]';
letterJ = [1 1 1 1 1 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
1 0 1 0 0 ...
0 1 0 0 0 ]';
letterK = [1 0 0 0 1 ...
1 0 0 1 0 ...
1 0 1 0 0 ...
1 1 0 0 0 ...
1 0 1 0 0 ...
1 0 0 1 0 ...
1 0 0 0 1 ]';
letterL = [1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 1 1 1 1 ]';
letterM = [1 0 0 0 1 ...
1 1 0 1 1 ...
1 0 1 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ]';
letterN = [1 0 0 0 1 ...
1 1 0 0 1 ...
1 1 0 0 1 ...
1 0 1 0 1 ...
1 0 0 1 1 ...
1 0 0 1 1 ...
1 0 0 0 1 ]';
letterO = [0 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
0 1 1 1 0 ]';
letterP = [1 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ...
1 0 0 0 0 ]';
letterQ = [0 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 1 0 1 ...
1 0 0 1 0 ...
0 1 1 0 1 ]';
letterR = [1 1 1 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 1 1 1 0 ...
1 0 1 0 0 ...
1 0 0 1 0 ...
1 0 0 0 1 ]';
letterS = [0 1 1 1 0 ...
1 0 0 0 1 ...
0 1 0 0 0 ...
0 0 1 0 0 ...
0 0 0 1 0 ...
1 0 0 0 1 ...
0 1 1 1 0 ]';
letterT = [1 1 1 1 1 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ]';
letterU = [1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
0 1 1 1 0 ]';
letterV = [1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
0 1 0 1 0 ...
0 0 1 0 0 ]';
letterW = [1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 0 0 1 ...
1 0 1 0 1 ...
1 1 0 1 1 ...
1 0 0 0 1 ]';
letterX = [1 0 0 0 1 ...
1 0 0 0 1 ...
0 1 0 1 0 ...
0 0 1 0 0 ...
0 1 0 1 0 ...
1 0 0 0 1 ...
1 0 0 0 1 ]';
letterY = [1 0 0 0 1 ...
1 0 0 0 1 ...
0 1 0 1 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ...
0 0 1 0 0 ]';
letterZ = [1 1 1 1 1 ...
0 0 0 0 1 ...
0 0 0 1 0 ...
0 0 1 0 0 ...
0 1 0 0 0 ...
1 0 0 0 0 ...
1 1 1 1 1 ]';
alphabet = [letterA,letterB,letterC,letterD,letterE,letterF,letterG,letterH,...
letterI,letterJ,letterK,letterL,letterM,letterN,letterO,letterP,...
letterQ,letterR,letterS,letterT,letterU,letterV,letterW,letterX,...
letterY,letterZ];
targets = eye(26);
% Y = eye(n) returns the n-by-n identity matrix.
% Y = eye(m,n) or eye([m n]) returns an m-by-n matrix with 1's on the diagonal and 0's elsewhere.
% Y = eye(size(A)) returns an identity matrix the same size as A.
% eye(m, n, classname) or eye([m,n],classname) is an m-by-n matrix
% with 1's of class classname on the diagonal and zeros of class classname elsewhere.
% classname is a string specifying the data type of the output.
% classname can have the following values:
% 'double', 'single', 'int8', 'uint8', 'int16', 'uint16', 'int32', or 'uint32'.
