can someone explain about this source code. its feature extraction for training back propagation. what is the input for training? whats mean by column 'number' through 'number' after run this. can someone explain details. thanks in advance

clc
close all
clear all
feature_extraction = [];
feature_extra = [];
j1=1;
j2=1;
% for j=65:70
for j=1:828
    ...for i=1:5
        RGB = imread(['frame',num2str(j),'.jpg']);
    img_crop = RGB;
    I= rgb2gray(RGB);
    I = imresize(I,[80 64],'bicubic');

      ...figure, imshow(I);
          BW = edge(I,'canny',0.25); %Finding the edges of the image
      [imx,imy]=size(BW);
      msk=[0 0 0 0 0;
          0 1 1 1 0;
          0 1 1 1 0;
          0 1 1 1 0;
          0 0 0 0 0;];

      B=conv2(double(BW),double(msk)); %smoothing the image to reduce the number of connected components
      ....figure,imshow(B);
          bw_resize=imresize(B,[70,50]);

      for fr=1:7
          for fc=1:5

              ...Atemp=sum(bw_resize((fr*10-9:fr*10),(fc*10-9:fc*10)))
                  temp=bw_resize((fr*10-9:fr*10),(fc*10-9:fc*10));
              Atemp = sum(temp);
              sum_feature((fr-1)*5+fc)=sum(Atemp)
          end
      end
      sum_feature=((100-sum_feature)/100);
      sum_feature=sum_feature;
      feature = [sum_feature];
      feature_extra = [feature_extra; feature];
      data_target(j1,j2) = 1;
      j2 = j2+1;
      ...end
          j1 = j1+1;
  end
  ...store these feature for training...

[Walter Roberson]

Please edit the source to have reasonable indentation. It is too hard to read the way it is.

[Image Analyst]

In MATLAB, type control-a (to select all the code), then type control-i to indent it properly. Then copy to clipboard and paste here. Then highlight the code you pasted and click the {}Code icon above the textbox to format it as code.

[Walter Roberson]

What are those "..." in the code at the beginning of some lines?

[meed negne]

after I run this code, it show like this

sum_feature =

    Columns 1 through 4

           0         0         0         0

    Columns 5 through 8

           0         0         0         0

    Columns 9 through 12

           0         0         0    1.0000

    Columns 13 through 16

      0.9963    0.7258    1.0000    1.0000

    Columns 17 through 20

      1.0000    0.0265   -1.2025    0.9410

    Columns 21 through 24

      1.0000    1.0000   -0.5532   -1.2016

    Columns 25 through 28

      0.6577    1.0000    1.0000   -0.3757

    Columns 29 through 32

     -0.3820    1.0000    1.0000    1.0000

    Columns 33 through 35

     -0.0382    0.5513    1.0000

[Image Analyst]

The posted code would not do that because it has a semicolon at the end of the line. It will only print that to the command window if there was no semicolon. But anyway,...why did you tell us this?

[meed negne]

because after I run that coding, the code like above print to my command window. so, I don't know what that happens

[Image Analyst]

The code you posted won't even run at all because of several syntax errors. You must be running a different version that does run.

[meed negne]

could you help me to fix it?

[Walter Roberson]

midah commented "its like took the lines as a comment" in response to my question about the "..." .

Yes, those lines would be interpreted as comments, but how were the lines originally written? Are those "..." part of the actual source you are using?

[meed negne]

no

[Walter Roberson]

It is difficult for us to debug source that isn't the same as the source you are using.

[meed negne]

and what should I do?

[Walter Roberson]

Well we don't care much, but if you don't post your actual source then chances are that we will lose interest and pay attention to something else instead.

[meed negne]

this my actual source. but it not working

feature_extraction = [];
feature_extra = [];
j1=1;
j2=1;

for j=1:450

RGB = imread(['frame',num2str(j),'.jpg']);
img_crop = RGB;
I= rgb2gray(RGB); 
I = imresize(I,[80 64],'bicubic');

BW = edge(I,'canny',0.25); %Finding the edges of the image
[imx,imy]=size(BW);
msk=[0 0 0 0 0;
0 1 1 1 0;
0 1 1 1 0;
0 1 1 1 0;
0 0 0 0 0;];

B=conv2(double(BW),double(msk)); %smoothing the image to reduce the number of connected components

bw_resize=imresize(B,[70,50]);

[Image Analyst]

What is the problem again? Is it that it prints stuff to the command window still? Because it does not look like it would do that since every line has a semicolon at the end of it. Is it that you're missing and "end" for your "for"? If there is any error, please paste all the red text here.

Answer 1

[Image Analyst]

It looks like the input is 5 jpeg images called Frame1.jpg through Frame5.jpg. There is no reason to read it 5 times during the i loop. It should be pulled out and be put inside the j look but above the i loop since the filename depends only on j, not i. I'm not exactly sure what it's doing but it looks like it's something with computing and saving the normalized intensity of a blurred version of the edge image.

I have no idea what you mean by "column 'number' through 'number'". I don't see anywhere in the code where column or number is mentioned