Hi Eliz, try this:
% Read your original image
Img1 = imread('http://www.mathworks.com/matlabcentral/answers/uploaded_files/50194/6.jpg');
% Make it black/white
BW1 = im2bw(Img1);
% It seems to have a border. Crop down to the corners
[toprow,leftcol] = find(~BW1,1);
[botrow,rightcol] = find(~BW1,1,'last');
BW_tight = BW1(toprow:botrow, leftcol:rightcol);
% Here, we COULD just divide the picture by its outer size.
sz = size(BW_tight);
szMid = ceil(sz/2);
% But the pattern isn't perfectly centered. So instead, let's find the
% centroid of ALL the white pattern parts and divide at that point
stats = regionprops(double(BW_tight));
szMid = fliplr(round(stats.Centroid)); % Centroid was XY, we want indices;
% Crop
ImgTopL = BW_tight(1:szMid(1),1:szMid(2));
ImgBotL = BW_tight(szMid(1):end,1:szMid(2));
ImgTopR = BW_tight(1:szMid(1),szMid(2):end);
ImgBotR = BW_tight(szMid(1):end,szMid(2):end);
% Display
figure
subplot(1,2,1), imshow(BW1), title('Original')
subplot(1,2,2), imshow(BW_tight), title('Cropped with centroid')
hold on, plot(szMid(2),szMid(1),'go')
figure
subplot(2,2,1), imshow(ImgTopL), title('Top left')
subplot(2,2,2), imshow(ImgTopR), title('Top right')
subplot(2,2,3), imshow(ImgBotL), title('Bottom left')
subplot(2,2,4), imshow(ImgBotR), title('Bottom right')
