image dividing, padding, regions

How to divide an image up into say 10x10 regions and account for padding (an integer number of regions does not fit in the image)

I have tried complicated convolution methods, and now want to try a more simplified approach. The aim is to then locally threshold each image for SNR calculations

This was my attempt:

n=10;  %regions in each dimension
[x,y]=size(ROI)
nx=floor(x/n)
ny=floor(y/n)

xGrid=1:(nx+1):10*nx
yGrid=1:(ny+1):10*ny
[X,Y]=meshgrid(xGrid,yGrid)
mesh(X,Y)

Answer 1

[Image Analyst]

Attached is a demo where I use local thresholding to correct a shaded document to prepare it for OCR.

[Jason]

Thanks IA. I've also come across another example you have written: In particular to split the image up into 4:

[rows columns] = size(ROI);
% Taking care to handle odd-size dimensions:
col1 = 1;
col2 = floor(columns/2);
col3 = col2 + 1;
row1 = 1;
row2 = floor(rows/2);
row3 = row2 + 1;
% Now crop
upperLeft = imcrop(ROI, [col1 row1 col2 row2]);
upperRight = imcrop(ROI, [col3 row1 columns - col2 row2]);
lowerLeft = imcrop(ROI, [col1 row3 col2 row2]);
lowerRight = imcrop(ROI, [col3 row3 columns - col2 rows - row2]);
% Display the images.
subplot(2, 3, 2);
imshow(upperLeft);
subplot(2, 3, 3);
imshow(upperRight);
subplot(2, 3, 5);
imshow(lowerLeft);
subplot(2, 3, 6);
imshow(lowerRight);

If I want to split into 8 to see if local thresholding is more effective for smaller ROI's, do I handle fractional values by using floor? How do draw ontop of the raw image a line grid representing how the image is being split up?

Also how are boundaries between the subregions handled, are there any pixels that are ignored?

Thanks

[Image Analyst]

Try this:

rows = 1234 % Whatever....
tiles = 8  % Number of tiles or chunks you want it divide up into.
% Find out where the dividing lines are.  This may be a fractional value.
startingRows = linspace(1, rows, tiles+1)
% Now get the starting integer rows.
startingRows = int32(startingRows(1:end-1))
% Get the ending rows.
endingRows = [startingRows(2:end)-1, rows]

In the command window, see the proper starting and ending rows for each chunk:

startingRows =
  Columns 1 through 7
          1   155.125 309.25  463.375   617.5   771.625   925.75
  Columns 8 through 9
        1079.875   1234

startingRows =
 1    155    309    463    618   772   926   1080

endingRows =
 154  308    462    617    771   925  1079   1234

[Jason]

Many thanks, I've taken your suggestion and almost got my solution. I managed to add the dividing lines onto the image, and I have attempted to add each tile to an image stack so I can process further: Its the adding to image stack that I can't understand why my code isn't working.

[rows,cols]=size(ROI) % ROI is my Image
tiles = 4  % Number of tiles or chunks you want it divide up into.
% Find out where the dividing lines are.  This may be a fractional value.
startingRows = linspace(1, rows, tiles+1)
startingCols = linspace(1, cols, tiles+1)
% Now get the starting integer rows / cols.
startingRows = int32(startingRows(1:end-1))
startingCols = int32(startingCols(1:end-1))
% Get the ending rows.
endingRows = [startingRows(2:end)-1, rows]
endingCols = [startingCols(2:end)-1, cols]

%Determine number of pixels in each tile;
nx=endingRows(1)-startingRows(1)
ny=endingCols(1)-startingCols(1)

%PreDefine Image stack (IM)
nimages=tiles*tiles
IM = zeros(nx, ny, nimages, class(ROI));

subplot(2,4,[1,2,5,6])
hold on
for i=1:nimages
    plot([startingRows(i),startingRows(i)],[0,cols],'y-');
    plot([0,rows],[startingCols(i),startingCols(i)],'y-');
    IM(:,:,i)=ROI(startingRows(i):endingRows(i),startingCols(i),endingCols(i));
    %IM(:,:,i)=ROI(1:25,1:25);
end
hold off;

subplot(2,4,3)
imshow(IM(:,:,1))