How to divide an image in two blocks

How can I divide an image into two halves? Then, after dividing it make each half independent, say

A = top half and

B = the bottom half.

So that I can extract features from each half.

[Nourou]

Hi Jack, Azzi,

Please I need a clarification about the usage of the following method:

im=imread('yourimage')
n=fix(size(im,1)/2)
A=im(1:n,:,:);
B=im(n+1:end,:,:)
imshow(A)
figure
imshow(B)

I try to use it on a grayscale image, but the results are 2 blank images. Please how can I edit your code to make it divide my image. I know you use it with a rgb image.

Thank you!

[Image Analyst]

A and B are probably double for some reason. Try

imshow(A, []);
imshow(B, []);

Start your own question if you need more help.

And there is a FAQ for the original poster's question: http://matlab.wikia.com/wiki/FAQ#How_do_I_split_an_image_into_non-overlapping_blocks.3F

Answer 1

[Azzi Abdelmalek]

im=imread('yourimage')
n=fix(size(im,1)/2)
A=im(1:n,:,:);
B=im(n+1:end,:,:)
imshow(A)
figure
imshow(B)

[Jack Sparrow]

Thank you Azzi. .Is it also thesame procedure for more than two blocks?

[Azzi Abdelmalek]

I guess it's the same, unless there are tools for that in image processing toolbox.

[Jack Sparrow]

okay thanks

[Image Analyst]

There are tool in the Image Processing Toolbox. imcrop(). But Azzi, your method is simpler. imcrop() unfortunately it doesn't have an option (yet) where you can just pass in starting and ending rows and columns. It's based off a "rectangle" where you have to pass in [x, y, width, height] so you have to calculate width and height. It's a pain and I'm constantly telling the image processing team that we need a pixel-based option.

[Jack Sparrow]

Thanks image analyst. Although I have already used the solution provided by Azzi

[Explorer]

Please do necessary changes in below mentioned code for splitting image as left half and right half.

im=imread('yourimage')
n=fix(size(im,1)/2)
A=im(1:n,:,:);
B=im(n+1:end,:,:)
imshow(A)
figure
imshow(B)

[Image Analyst]

% Read in image and display it.
rgbImage=imread('peppers.png');
subplot(2, 2, 1);
imshow(rgbImage);
% Get sizes of the images.
[rows, columns, numberOfColorChannels] = size(rgbImage);
middleColumn = int32(columns/2)
title('Right Half', 'FontSize', 25);
% Split them up.
leftHalf = rgbImage(:, 1:middleColumn, :);
rightHalf=rgbImage(:, middleColumn+1:end, :);
% Display them.
subplot(2, 2, 3);
imshow(leftHalf);
title('Left Half', 'FontSize', 25);
subplot(2, 2, 4);
imshow(rightHalf);
title('Right Half', 'FontSize', 25);

% Enlarge figure to full screen.
set(gcf, 'Units', 'Normalized', 'OuterPosition', [0 0 1 1]);
% Give a name to the title bar.
set(gcf, 'Name', 'Demo by ImageAnalyst', 'NumberTitle', 'Off')

[Explorer]

Thank you Image Analyst.

[zohreh nezami]

thank you

[suma g]

sir u have given for left half and right half of the image... very helpful.i want to segment the face image into upper and bottom half of the face ... upper- forehead,eyes and eyebrows & lower- nose lips and chin CAN U GIVE ME SOME IDEA

Answer 2

[utkarsh kumar]

hi i am having a doubt in understanding the code.....in the below code

im=imread('yourimage')

n=fix(size(im,1)/2)

A=im(1:n,:,:);

B=im(n+1:end,:,:)

imshow(A)

figure

imshow(B)

What is the meaning of the highlighted one .

Thanks

[Image Analyst]

It says A is equal to rows 1 through n, all columns, all color rows. The colon means "all" if it's by itself, or "to" if it's in an expressions like 1:n.

So if n is the middle row, A will be the top half. Then B picks up from one row below the middle row. It says B equals all rows from one row below the middle row to the last row, and all columns and all color channels. If it's grayscale, having an extra : in there for the third index will not hurt - it will just ignore it.