matlab LBP+SVM图像分类

clc; 
clear all; 
close all;
%%   使用SVM做识别，测试准确率

% Files = dir(fullfile('.\Car','*.jpg'));
file_path =  'E:/matlab_0808/0816/people1/';% 图像文件夹路径
img_path_list = dir(strcat(file_path,'*.jpg'));%获取该文件夹中所有jpg格式的图像
CarNum = length(img_path_list);%获取图像总数量
% Files = dir(fullfile('.\Car2','*.png'));
% CarNum = length(Files);
CarFea=[];
for i = 1:CarNum
    image_name = img_path_list(i).name;% 图像名
    Img =  imread(strcat(file_path,image_name)); 
    imggray = rgb2gray(Img);
    [m,n] = size(imggray);
    H2=lbp(imggray);
    H2fea = (H2-min(H2))/(max(H2)-min(H2));
    CarFea=[CarFea;H2fea];
end

%Files = dir(fullfile('.\Tank3','*.jpg'));
file_path =  'E:/matlab_0808/0816/other/';% 图像文件夹路径
img_path_list = dir(strcat(file_path,'*.jpg'));%获取该文件夹中所有jpg格式的图像
TankNum = length(img_path_list);%获取图像总数量
TankFea=[];
for i = 1:TankNum
    image_name = img_path_list(i).name;% 图像名
    Img =  imread(strcat(file_path,image_name)); 
    imggray = rgb2gray(Img);
    [m,n] = size(imggray);
    H2=lbp(imggray);
    H2fea = (H2-min(H2))/(max(H2)-min(H2));
    TankFea=[TankFea;H2fea];
end
 
dataFea = [CarFea;TankFea];
Group = [0*ones(CarNum,1); 1*ones(TankNum,1)];

%classifer = fitcecoc(dataFea,Group);  

% SVM train
svmStr = svmtrain(dataFea,Group,'kernel_function','linear','showplot',true);

k1=1;
for j=20:50
    name=num2str(j);
    name1=strcat('E:/matlab_0808/0816/test_people_other/',name);
    Img2 = imread(strcat(name1,'.jpg')); 
    imggray2 = rgb2gray(Img2);  
    [m,n] = size(imggray2);
    H21=lbp(imggray2);
    H2fea1 = (H21-min(H21))/(max(H21)-min(H21));
    pgroup(k1) = svmclassify(svmStr,H2fea1,'Showplot',true);     % svm预测
    k1=k1+1;
end

i=1;
for k=71:100
    name=num2str(k);
    name1=strcat('E:/matlab_0808/0816/test_people_other/',name);
    Img2 = imread(strcat(name1,'.jpg')); 
    imggray2 = rgb2gray(Img2);  
    [m,n] = size(imggray2);
    H21=lbp(imggray2);
    H2fea1 = (H21-min(H21))/(max(H21)-min(H21));
    pgroup1(i) = svmclassify(svmStr,H2fea1,'Showplot',true);     % svm预测
    i=i+1;
    %pgroup1(k) = svmclassify(svmStr,H2fea1,'Showplot',true);     % svm预测
end