How to crop an image using matrices

I need to crop a certain portion of an image. But the problem is that I want the croped image in the same dimension of the orginal image. I am asked to do this with the help of matrices. How can I do it

Answer 1

[Ameer Hamza]

You can use imcrop(): https://www.mathworks.com/help/images/ref/imcrop.html and then apply imresize(): https://www.mathworks.com/help/images/ref/imresize.html to the cropped image.

[Arya Gopan]

No sir. I tried using imcrop. But the issue is that while doing so i get an image with different dimension.. I need the dimension of the cropped image same as that of the orginal one.

Can you please tell me whether it is possible to do like this.

A=imread('trial.jpg'); // reading the image

X=zeroes(size(A)); // Creating a matrix of the same size of A nd making all of the elements in it zero

X= ones(120:50,450;500);// Making the desired are of the image which has to be croped zero

Product = A*X

This code so that by multiplying X with A we will get the cropped image. But while I did this in Matlab The 3rd line(bold one) is error.

please help

[Ameer Hamza]

X= ones(120:50,450;500) is not correct MATLAB syntax. I guess you want to do this

A=im2double(imread('trial.jpg'));
X=zeros(size(A)); % Creating a matrix of the same size of A nd making all of the elements in it zero
X(50:120,450:500,:) = 1;

and then

Product = A.*X;

But this will give you white area around the cropped region.

Try something like this

A=im2double(imread('trial.jpg'));
rect = [50 450 70 50];
X = imcrop(A, rect);
X = imresize(X, size(A,[1 2]))

[Arya Gopan]

Yes sir I will try both and update you.

But what does this white are around the cropped region denote.. I mean the cropped portion would have the same dimension like that of the image right?

[Ameer Hamza]

Following your method, It will have same dimension, but looks like this

A=im2double(imread('pears.png'));
X=zeros(size(A)); % Creating a matrix of the same size of A nd making all of the elements in it zero
X(50:120,450:500,:) = 1;
Product = A.*X;

subplot(2,1,1)
imshow(A);
title('original')

subplot(2,1,2)
imshow(Product);
title('Cropped')

(* I meant to write black background)

Using my method

A=im2double(imread('pears.png'));
rect = [50 450 70 50];
X = imcrop(A, rect);
Product = imresize(X, size(A,[1 2]));

subplot(2,1,1)
imshow(A);
title('original')

subplot(2,1,2)
imshow(Product);
title('Cropped')

[Arya Gopan]

Sir I have to crop the left most dot from the image. When I did the first code I got the croped image so.

And when I tried the second one it says there is errorin the "imresize statement'

[Arya Gopan]

[Image Analyst]

Try

Product = imresize(X, size(A));

[Arya Gopan]

still

[Image Analyst]

It looks like you redefined size to be a variable in your program. What does this show in the command window.

>> which -all size

[Arya Gopan]

A=im2double(imread('trial.jpg'));

rect =[532 454 654 600];

X=imcrop(A,rect);

Product = imresize(X, size(A));

subplot(2,1,1)

imshow(A);

title ('Orginal')

subplot(2,1,2)

imshow(Product);

title('Cropped')

?? This is the program i wrote

[Ameer Hamza]

Which MATLAB release are you using? Try replacing the line with this

Product = imresize(X, [size(A,1) size(A,2)]);

[Arya Gopan]

Im using 2014 version.

[Arya Gopan]

Sir actuallyI am a Post Graduate student of Physics and this is for my final year projet.

Here I am given the image of a fringe, whih I have converted to its fourier transform. So the image that I have attatched above is its fourier transform. Now I am asked to crop one of those lobes so that rest of the calculations can be carried out. While doing imcrop funtion its dimension changes. So I was asked to do it using matrix method.

A=im2double(imread('pears.png'));

X=zeros(size(A)); % Creating a matrix of the same size of A nd making all of the elements in it zero

X(50:120,450:500,:) = 1;

Product = A.*X;

subplot(2,1,1)

imshow(A);

title('original')

subplot(2,1,2)

imshow(Product);

title('Cropped')

// I used similar code with diff variable names but I got that black screen as output.

[Ameer Hamza]

Have you changed this line?

X(50:120,450:500,:)

You need to specify the pixel coordinates for the left dot in your image.

[Arya Gopan]

Yes i used the data cursor to know the pixel co-ordinates and inlcluded it.

[Ameer Hamza]

Can you attach your image and also your code?

[Arya Gopan]

This is the fourier transform of the image. The left most dot is to be cropped. ill attatch the code below.

[Arya Gopan]

clear all

I = imread('a35.jpg');
I = rgb2gray(I);
[r , c] = size(I);
for i = 1:r
    X(i,:)=fft(I(i,:));
end
for j = 1:c
    Y(:,j)=fft(X(:,j));
end
figure(1)
imshow(I)
figure(2)
M=Y;
M=fftshift(M);
Ab=abs(M);
Ab=(Ab-min(min(Ab)))./(max(max(Ab))).*225;
imshow(Ab)



X=zeros(size(Ab));
X(563:475,639:556,:)=1;
Product = Ab.*X;

subplot(2,1,1)
imshow(Ab);
title('Orginal')

subplot(2,1,2)
imshow(Product);
title('Cropped')
////////////////////////////////// This is te code that I have used so far using matrices

[Ameer Hamza]

The indexes must be defined in increasing order. Change the line

X(563:475,639:556,:)=1;

to

X(475:563,556:639,:)=1;

[Arya Gopan]

Yes sir I got it.

Thankyou so much.

[Arya Gopan]

Sir an you help me with finding the inverse fourier transform of the croped image using ifftshift

. I get a blank black ffigure