how to process images with different sizes for classification?

I've the following code for the image size (row = 320; col = 240;)

But i'm having images with variable sizes (116x110),(134x125), which need to be processed

function y = bounding_box(img)

[row col] = size(img);
top = 0;
bottom = 0;
left = 0;
right = 0;

for i = 1:row
    if sum(img(i,:)) > 0
        top = i;
        break
    end
end

for i = row:(-1):1
    if sum(img(i,:)) > 0
        bottom = i;
        break
    end
end

for i = 1:col
    if sum(img(:,i)) > 0
        left = i;
        break
    end
end

for i = col:(-1):1
    if sum(img(:,i)) > 0
        right = i;
        break
    end
end

y = img(top:bottom, left:right);

Answer 1

[Image Analyst]

Why not simply ask regionprops() to give you the bounding box?

[arun]

can you give me an example code for that?

[Image Analyst]

How many blobs do you have? And if you have more than 1, do you want the bounding box of each one, or the bounding box of all the individual bounding boxes. In short, it's

labeledImage = bwlabel(binaryImage);
props = regionprops(labeledImage, 'BoundingBox');

props is a structure array with measurements, like bounding box, for each region.

See a full example in my Image Segmentation Tutorial in my File Exchange