Crop images using coordinates and save them

I am cropping an image using some x, y coordinates. After the cropping i can save the cropped image. Is there a way i can save both cropped image as well as the rest?

(What i meant by rest is the remaining part of the image, what i got after removing the cropped image).

Please help me.

Thanks!

Answer 1

[Image Analyst]

Use imwrite()

imwrite(croppedImage, croppedFileName);

The original image is most likely already saved, if you read it in using imread().

[Anounymous85]

@Image Analyst

I meant something like this. I have the image mentioned below.

I am cropping that image using coordinate values. The result i am getting is the image mentioned below. And i can save that part using imwrite.

What i want is to get this part as well as the other half of the image separatly.

[Image Analyst]

OK, and my answer will let you save that cropped image you got. I believe I answered your problem - you need to use imwrite(). First on the image you showed, then on the "other half of the image". Do you still have a problem? If so, what is it?

[Image Analyst]

Since it appears that the problem you're trying to communicate is that your image is not being cropped correctly, I believe you're making the common beginner mistake of assuming that arrays are indexed (x,y) instead of (y, x). The first index is y, which is row, and the second index is x, which is column. So try this:

[rows, columns] = find(binaryImage); % or [y, x] = find(binaryImage);
row1 = min(rows);
row2 = max(rows);
col1 = min(columns);
col2 = max(columns);
imwrite(binaryImage(row1:row2, col1:col2), fileName);

Answer 2

[Mark Sherstan]

You will need to fill in the "hole" with white or black pixels. Give something like this a try:

I = imread('test.png');

J = imcrop(I,[200 200 400 600]);

I(200:600,200:800,:) = 0;

figure(1)
subplot(1,2,1); imshow(J)
subplot(1,2,2); imshow(I)

Or if you are splitting the image in half try cropping twice like so:

I = imread('test.png');

idx = size(I);

J = imcrop(I,[0 0 idx(2)/2 idx(1)]);
K = imcrop(I,[idx(2)/2 0 idx(2) idx(1)]);

figure(1)
subplot(1,2,1); imshow(J)
subplot(1,2,2); imshow(K)

[Anounymous85]

@Mark Sherstan

Can you please tell me how to select these values?

J = imcrop(I,[200 200 400 600]);

I(200:600,200:800,:) = 0;

[Mark Sherstan]

[200 200 400 600] was randomly selected but corresponds to [x y width height]. Using that information the next part I(200:600,200:800,:) can be rewritten as: I(200:(200+400),200:(200:600),:). Basically you are setting some refernce coordinate point and then selecting some width and height of data with (0,0) being the top left of the image.

[Anounymous85]

Mark Sherstan

I guess the first solution you have provided here can help me with what i want. Thank you!

But only problem I have is my x, y, width and height values are getting changed for different pictures. It is not easy to change those values time to time. I just want to get those values automatically, matching tto different images.

[Mark Sherstan]

How does your area change from picture to picture? Can you find the size of the image and then just take a percentage of that? E.g:

I = imread('test.png');
[height, width, ~] = size(I);
J = imcrop(I,[0 0 width/2 height]);
imshow(J)

Answer 3

[Aytaç Tok]

How do I crop a picture using border coordinates.I have 1695x2 size limit values. How do I crop the object inside these values ​​from the original picture?

[Image Analyst]

You can use imcrop() or indexing. If you need additional help, start a new question and attach your image, and a mat file with your "limit values" or "boundary coordinates" in a .zip file with the paper clip icon.

[Aytaç Tok]

I shared the image.I just want to crop the banana picture here, so the rgb banana picture in the nerve pixel values

clc;close all;clear all
e=imread('muzpp.jpeg');
    l=imbinarize(rgb2gray(e));
grayImage=rgb2gray(e);
gri=grayImage;
[r c]=size(l);
 sb=imbinarize(gri);
  sb=bwareaopen(sb,10);
   se=strel('disk',5);
 sb=imclose(sb,se);
  sb=imfill(sb,'holes');
 se2=strel('disk',30);
 sb2=imerode(sb,se2);
  [etiket nesnesayisi]=bwlabel(sb2,4);
   [sinirkordinatlari etiketler]=bwboundaries(sb2);

   [etiketler nesnesayisi]=bwlabel(sb2,4);
   for k=1:length(sinirkordinatlari)
       sinir=sinirkordinatlari{k};
   end
imshow(e)