does not able to recognize correct class of the image.

%% Taking an Image
[fname,path]=uigetfile('.jpg','Open a image as Input for Training');
c=strcat(path,fname);
im=imread(fname);
imshow(im);
title('Input Image');
c=input('Enter the Class (Number from 1-10)');
%% Feature Extraction
F=FeatureStatistical(im);
try
    S = load('db.mat');
    F=[F c];
    db=[db; F, c];
  catch ME
    disp(ME.message);
    % F  = reshape(F,[3,2]);  % Omit the duplicate reshaping - it has no effect
    % F  = reshape(F, [1,6]);
    db = [F c];
  end
  save('db.mat', 'db');
  
  
  
  function[F]=FeatureStatistical(im)
im=double(im);
m=mean(im(:));
s=std(im(:));
F=[m s];
end

Here I trained 2 calsses using 10 images for each. class -1 is a buildind and class -2 is a mountain.

%% Test Image
[fname,path]=uigetfile('.jpg','Provide a Image for Testing');
c=strcat(path,fname);
im=imread(fname);
imshow(im);
title('Test Image');
%% Fond out which Class it Belongs to
Ftest=FeatureStatistical(im);
%% Compare with Database
load db.mat
Ftrain=db(:,[1,2]);
ctrain=db(:,3);
for(i=1:size(Ftrain,1))
dist(1,:)=sum(abs(Ftrain(1,:)-Ftest));
end
Min=min(dist);
if(Min<50)
m=find(dist==Min,1);
det_class=ctrain(m);
msgbox(strcat('detected class',num2str(det_class)));
else
    msgbox('Not Exist');
end

When testing the images using this code it says "Not exist" for even exactly the same image uploaded for test.

[Walter Roberson]

why do you put c onto the end of F and then put F and c both on the end of db? the F created in the line above already includes c.

[Sam Upeshala]

Thank you for the answer. Can you please explain it please?

[Walter Roberson]

Suppose that after the line

    F=FeatureStatistical(im);

that F is a row vector of length 2, such as [93 5.83]. And suppose that db is the empty matrix [].

c=input('Enter the Class (Number from 1-10)');

so c is expected to be a scalar integer. Suppose the user entered 10.

You have the code

    F=[F c];
    db=[db; F, c];

So before the F=[F c] line, F is [93 5.83] and after that line, F = [93 5.83 10] because you put c on the end of F. Then you have the assignment to db; with empty db that would be

    db=[[]; [93 5.83 10], 10]
   

which would store db = [93 5.83 10 10]

Now my question for you is why you are storing two copies of c in db? The last two elements stored are going to be the same. What benefit does that provide ? Why not just store

db=[db; F];

which would get db=[93 5.83 10] ?

[Sam Upeshala]

When I remove that "c" from "db=[db; F, c];" it gives the message "Dimensions of matrices being concatenated are not consistent." and it only store one row in the database. (last trained image data). others are disapeard.