How to find the (pixel) dimensions of images on a white background

Hello there,

I have a folder of 60 images of irregular household items that are greyscale on a white 500x500 pixel canvas. I need to have the images rescaled (will do in photoshop) to match the average size of a different set of images. Using MATLAB, I found the references images have 1.51 times more non-white pixels (i.e. image pixels) than the items do and need to alter these images to match the size. However, because the images are of vastly different items (i.e. a paperclip vs a closet), they vary a lot by dark pixel count.

How can I find the dimensions (height and width) of these images in MATLAB. I've included an example of an image being

samplePic = imread(img_dir + "/hn_processed");

image_double = im2double(samplePic); % turn all white pixels to 1

% I tried the below code by it didn't do what I wanted
actual_image = find(image_double ~= 1);

[rows, columns, numberOfColorChannels] = size(actual_image);

used. Thank you!

[Matthew Peoples]

I realize in hindsight this is a bit of a confusing question. Basically I really need to get the dimensions (height and width) of the strawberry image within the white canvas (not the entire canvas itself). Hope someone can help! :)

[Adam Danz]

As long as the strawberry is oriented vertically or horizontally, the simple approach in my solution gives you to width and height.

Answer 1

[Adam Danz]

Assuming height and width are defined by the first and last non-white (or any background color) pixel from top-to-bottom and from left-to-right,

I = imread('image.bmp');

% Assumes the first pixel is the background color, 
% otherwise just use 255 or likewise.  
[row,col] = find(I~=I(1)); 

height = max(row)-min(row); 
width = max(col)-min(col); 

To visually confirm the measurements,

imshow(I)
hold on 
r = rectangle('position', [min(col), min(row), width, height],'EdgeColor', 'r');