cla.s.m

clc
srcFiles = dir('E:\traffic\ftl\al--3---round about\New folder\*.jpg');  % the folder in which ur images exists
classifier=[];hog1=[];
trainingLabels=[1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2];
%trainingLabels=trainingLabels';
for i = 1 : length(srcFiles)
    filename = strcat('E:\traffic\ftl\al--3---round about\New folder\',srcFiles(i).name);
    I = imread(filename);
    %figure, imshow(I);
    hog1(i,:)=extractHOGFeatures(I);

end
t = templateSVM('KernelFunction','linear'); %gaussian
classifier = fitcecoc(hog1,trainingLabels,'Learners',t);

%classifier = fitcecoc(hog1,trainingLabels,'Learners','naivebayes');


srcFiles = dir('E:\traffic\ta\*.jpg');  % the folder in which ur images exists
hog2=[];
for i = 1 : length(srcFiles)
    filename = strcat('E:\traffic\ta\',srcFiles(i).name);
    I2 = imread(filename);
    %figure, imshow(I);
    hog2(i,:)=extractHOGFeatures(I2);

end


%testFeatures = extractHOGFeatures(hog2);
 testLabels=[1,2,2,2,2,1,1,1,1,2];

% Make class predictions using the test features.
predictedLabels = predict(classifier, hog2);

% Tabulate the results using a confusion matrix.
confMat = confusionmat(testLabels, predictedLabels)

helperDisplayConfusionMatrix(confMat)