I posted this question in stack overflow, but I think it is more appropriate here:

I have a reference image - 'image_01.png' of size 200x200 and resolution 72 px/in.

I have another image 'image_02.png' (with the same size and resolution) with several levels of zoom, of which, one is close to 'image_01.png', in the sense that the objects in that image are of "similar" size. They are NOT the same images, btw.

Here is an example with three levels of zoom, and clearly zoom_02 is "closer" to image_01 than zoom_01 and zoom_02. My question is, what mathematical metric (ratio, or any measure) do we choose to determine what zoom is appropriate to ensure such similarity? Let's say I even have a screen ruler to measure the distance in the images, in pixels.

image_02_zoom_03 zoom_03 image_02_zoom_01 zoom_01

image_02_zoom_02 zoom_02 image_01 image_01


Since you want the image size and resolution to remain constant, the term optical magnification seems apt there.

You could then say

Please use 20x optical magnification to identify said feature.

Please be careful that the term magnification is different from zoom.

Two different cameras with the same 5x optical zoom might produce images with different optical magnification depending on their optics.

Zoom refers to the ratio of the max to min focal length of your camera, whereas,
Magnification refers to the ratio of projection size to the real subject size.

Given a particular magnification, you could use your camera focal lengths and apertures to come up with at zoom metric for that particular camera to produce the desired results.

See this for a better explanation.

  • $\begingroup$ Thank you for the explanation. Yes, I found out there was a software that used the optical magnification to output the area. Your explanation is helpful. $\endgroup$
    – Kasthuri
    Dec 31 '18 at 18:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.