In image processing, why is it recommended to loop over Y first and then X second?

Question

I've read in a few places that when processing the pixels of an image, it is recommended that you loop over Y and then X pixels because it is more likely to be memory efficient. Why is this so?

Answer 1

A1 A2 A3 A4 
B1 B2 B3 B4 
C1 C2 C3 C4 
D1 D2 D3 D4 

Supposing that's your image (and some sort of coordinates to it's points), it is stored in memory in a string, just like:

A1 A2 A3 A4 B1 B2 B3 B4 C1 C2 C3 C4 D1 D2 D3 D4.

Let's simulate those options:

For X and then by Y:

A1 B1 C1 D1 A2 B2 C2 D2 A3 B3 C3 D3 A4 B4 C4 D4 

Now check on the string, how messy those access would be (just like random reads, right?)

Nor, For Y and then by X

A1 A2 A3 A4 B1 B2 B3 B4 C1 C2 C3 C4 D1 D2 D3 D4

See? Straight forward read!

That's why.