Let's say we pick an image of space. Let's assume there is big star at 40 light years away and a smaller star 20 light years away. But when we look at the image both would look same size. How can we identify which is bigger and which is smaller by size and how far is it from us.
I want to do image processing on these images.
Keywords: Astronomy, space telescope, stars machine learning, opencv, python.
Short answer: It is impossible. Long answer: You need additional information, such as light emission from each star and received by the camera (if they are equal, the closer star will be brighter). Using only size in pixels you cannot determine even the relative distance between them as they may differ in the size and distance to the observer (the star with radius r and distance d is projected exactly the same as the star with radius 2r and distance 2d).
Unfortunately, it is not straightforward to infer the distance from the size by just looking at your image. You could make guesstimates on the stars' relative distance by comparing their colors (aka temperatures). However, this requires you to have accurate UBV measurements (quite some other filter systems also work).
If you know (or can identify) the stars in your image, you could look up any relevant information on public databases.
Types of measurements that allow you measure the distance to a star are their parallax, their color, or in some cases it can be inferred from their variability.
You can't tell from the image alone, you need to know the apparent and absolute brightness of each star, their temperature, spectrum, and other characteristics to know which star is the brightest in compare to the others.
