基于matlab车牌的定位与分割识别程序.doc

基于Matlab的车牌定位与分割
经典算法
I=imread('car.jpg'); %读取图像
figure(); subplot(3,2,1),imshow(I), title('原始图像');

I1=rgb2gray(I);%转化为灰度图像
subplot(3,2,2),imshow(I1),title('灰度图像');

I2=edge(I1,'robert',0.09,'both');%采用robert算子进行边缘检测
subplot(3,2,3),imshow(I2),title('边缘检测后图像');

se=[1;1;1]; %线型结构元素
I3=imerode(I2,se); %腐蚀图像
subplot(3,2,4),imshow(I3),title('腐蚀后边缘图像');

se=strel('rectangle',[25,25]); 矩形结构元素
I4=imclose(I3,se);%图像聚类、填充图像
subplot(3,2,5),imshow(I4),title('填充后图像');

I5=bwareaopen(I4,2000);%去除聚团灰度值小于2000的部分
subplot(3,2,6),imshow(I5),title('形态滤波后图像');

[y,x,z]=size(I5);
I6=double(I5);
Y1=zeros(y,1);
for i=1:y
for j=1:x
if(I6(i,j,1)==1)
Y1(i,1)= Y1(i,1)+1;
end
end
end
[temp MaxY]=max(Y1);
figure();
subplot(3,2,1),plot(0:y-1,Y1),title('行方向像素点灰度值累计和'),xlabel('行值'),ylabel('像素');

%求的车牌的行起始位置和终止位置
PY1=MaxY;
while ((Y1(PY1,1)>=50)&&(PY1>1))
PY1=PY1-1;
end
PY2=MaxY;
while ((Y1(PY2,1)>=50)&&(PY2<y))
PY2=PY2+1;
end
IY=I(PY1:PY2,:,:);
X1=zeros(1,x);
for j=1:x
for i=PY1:PY2
if(I6(i,j,1)==1)
X1(1,j)= X1(1,j)+1;
end
end end
subplot(3,2,2),plot(0:x-1,X1),title('列方向像素点灰度值累计和'),xlabel('列值'),ylabel('像数');

%求的车牌的列起始位置和终止位置PX1=1;
while ((X1(1,PX1)<3)&&(PX1<x))
PX1=PX1+1;
end
PX2=x;
while ((X1(1,PX2)<3)&&(PX2>PX1))
PX2=PX2-1;
end
PX1=PX1-1;
PX2=PX2+1;
%分割出车牌图像%
dw=I(PY1:PY2,PX1:PX2,:);
subplot(3,2,3),imshow(dw),title('定位剪切后的彩色车牌图像')

4.2 车牌字符分割
确定车牌位置后下一步的任务就是进行字符切分分离出车牌号码的全部字符图像。
if isrgb(I)
I1 = rgb2gray(I); %将RGB图像转化为灰度图像
else I1=I； end
g_max=double(max(max(I1)));
g_min=double(min(min(I1)));
T=round(g_max-(g_max-g_min)/3); % T 为二值化的阈值
[m,n]=size(I1);% d:二值图像
%h=graythresh(I1);
I1=im2bw(I1,T/256);
subplot(3,2,4);
imshow(I1),title('二值化车牌图像');

I2=bwareaopen(I1,20);
subplot(3,2,5);
imshow(I2),title('形态学滤波后的二值化图像');

[y1,x1,z1]=size(I2);
I3=double(I2);
TT=1;
%%%%%%%去除图像顶端和底端的不感兴趣区域%%%%%
Y1=zeros(y1,1);
for i=1:y1
for j=1:x1
if(I3(i,j,1)==1)
Y1(i,1)= Y1(i,1)+1 ;
end
end
end
Py1=1;
Py0=1;
while ((Y1(Py0,1)<20)&&(Py0<y1))
Py0=Py0+1;
end
Py1=Py0;
while((Y1(Py1,1)>=20)&&(Py1<y1))
Py1=Py1+1;
end
I2=I2(Py0:Py1,:,:);
subplot(3,2,6);
imshow(I2),title('目标车牌区域');

% 分割字符按行积累量%
X1=zeros(1,x1);
for j=1:x1
for i=1:y1
if(I3(i,j,1)==1)
X1(1,j)= X1(1,j)+1;
end
end
end
figure(5);
plot(0:x1-1,X1),title('列方向像素点灰度值累计和'),xlabel('列值'),ylabel('累计像素量');

Px0=1;
Px1=1;
%分割字符
for i=1:7
while ((X1(1,Px0)<3)&&(Px0<x1))
Px0=Px0+1;
end
Px1=Px0;
while (((X1(1,Px1)>=3)&&(Px1<x1))||((Px1-Px0)<10))
Px1=Px1+1;
end
Z=I2(:,Px0:Px1,:);
switch strcat('Z',num2str(i))
case 'Z1'
PIN0=Z;
case 'Z2'
PIN1=Z;
case 'Z3'
PIN2=Z;
case 'Z4'
PIN3=Z;
case 'Z5'
PIN4=Z;
case 'Z6'
PIN5=Z;
otherwise
PIN6=Z;
end
figure(3);
subplot(1,7,i);
imshow(Z);

Px0=Px1;
End
function inpt = pretreatment(I)
%YUCHULI Summary of this function goes here
% Detailed explanation goes here
if isrgb(I)
I1 = rgb2gray(I);
else
I1=I;
end
I1=imresize(I1,[50 20]);%将图片统一划为50*20大小
I1=im2bw(I1,0.9);
[m,n]=size(I1);
inpt=zeros(1,m*n);
%将图像按列转换成一个行向量
for j=1:n
for i=1:m
inpt(1,m*(j-1)+i)=I1(i,j);
end
end

%神经网络进行识别。
close all;
clear all;
%%%%归一化训练样本%%%%%%
I0=pretreatment(imread('0.jpg'));
I1=pretreatment(imread('1.jpg'));
I2=pretreatment(imread('2.jpg'));
I3=pretreatment(imread('3.jpg'));
I4=pretreatment(imread('4.jpg'));
I5=pretreatment(imread('5.jpg'));
I6=pretreatment(imread('6.jpg'));
I7=pretreatment(imread('7.jpg'));
I8=pretreatment(imread('8.jpg'));
I9=pretreatment(imread('9.jpg'));
I10=pretreatment(imread('A.jpg'));
I11=pretreatment(imread('C.jpg'));
I12=pretreatment(imread('G.jpg'));
I13=pretreatment(imread('L.jpg'));
I14=pretreatment(imread('M.jpg'));
I15=pretreatment(imread('R.jpg'));
I16=pretreatment(imread('H.jpg'));
I17=pretreatment(imread('N.jpg'));
P=[I0',I1',I2',I3',I4',I5',I6',I7',I8',I9',I10',I11',I12',I13',I14',I15',I16',I17'];
%输出样本%%%
T=eye(18,18);
%%bp神经网络参数设置
net=newff(minmax(P),[1000,32,18],{'logsig','logsig','logsig'},'trainrp');
net.inputWeights{1,1}.initFcn ='randnr';
net.layerWeights{2,1}.initFcn ='randnr';
net.trainparam.epochs=5000;
net.trainparam.show=50;
%net.trainparam.lr=0.003;
net.trainparam.goal=0.0000000001;
net=init(net);
%%%训练样本%%%%
[net,tr]=train(net,P,T);
%%%%%%%测试%%%%%%%%%
%I=imread('DSC01323.jpg');
I=imread('DSC01344.jpg');
dw=location(I);%车牌定位
[PIN0,PIN1,PIN2,PIN3,PIN4,PIN5,PIN6]=StringSplit(dw);%字符分割及处理

%%%%%%%%%%%测试字符，得到识别数值%%%%
PIN0=pretreatment(PIN0);
PIN1=pretreatment(PIN1);
PIN2=pretreatment(PIN2);
PIN3=pretreatment(PIN3);
PIN4=pretreatment(PIN4);
PIN5=pretreatment(PIN5);
PIN6=pretreatment(PIN6);
P0=[PIN0',PIN1',PIN2',PIN3',PIN4',PIN5',PIN6'];
for i=2:7
T0= sim(net ,P0(:,i));
T1 = compet (T0) ;
d =find(T1 == 1) - 1
if (d==10)
str='A';
elseif (d==11)
str='C';
elseif (d==12)
str='G';
elseif (d==13)
str='L';
elseif (d==14)
str='M';
elseif (d==15)
str='R';
elseif (d==16)
str='H';
elseif (d==17)
str='N';
else
str=num2str(d);
end
switch i
case 2
str1=str;
case 3
str2=str;
case 4
str3=str;
case 5
str4=str;
case 6
str5=str;
otherwise
str6=str;
end
end
%显示定位后的分割出的车牌彩图，%

%%%%%%识别结果以标题形式显示在图上%%%
s=strcat('渝',str1,str2,str3,str4,str5,str6);
figure();
imshow(dw),title(s);
非诚勿扰 欢迎下载