Understanding Bounding Box Modifications

Hi All,

This is regarding Bounding box explanation in Part 2 lesson 8 where Jeremey modifies the original bounding box annotation i.e. [columns, rows, height, width ] to [rows, columns, top left, bottom right].

What I am not able to understand is the mathematical intuition behind the conversion. How we would get top left by doing width + rows - 1 and the same for bottom right. Can somebody please explain with a example ?

Thanks folks in Advance!

Maybe a picture would be able explain this line of code
bb = np.array([bb[1], bb[0], bb[3]+bb[1]-1, bb[2]+bb[0]-1])


Thanks a lot. This helped and made sense.