Best targets to calibrate camera with long focusing distance

Question

I am currently using a number of cameras with long focusing distance, i.e. >3m, and I want to calibrate them to find the intrinsic matrix.

My problem is that if I want to cover all or most of the image with the target AND the target to be in focus, then the size would be too big of anything that I can print or buy. As I see it, I have two choices:

1. use a reasonably sized target (e.g. A1) at focusing distance, which means that it will be in focus but it will cover only a small section of the image
2. use a reasonably sized target closer than focusing distance, which means that it will be out of focus but it can cover most of the image

I have a chessboard printed in A1 size (60 x 84 cm), so I'm wondering what is the best way to approach the calibration process.

Right now I am more inclined to use method 1, by taking multiple images and try to move the target to all the areas in the field of view. However I would like to know if someone has faced a similar problem and if one solution is theoretically more robust than the other. Would method 2 work as well for example, since the chessboard corners can still be detected and interpolated?

Answer 1

If you're in a professional environment where time saved will pay for capital equipment, consider an optics table and a collimating lens. You'll have to take great care to make sure that the resulting optics train doesn't add appreciable distortion, but you'll be able to do the work in a room.

If you're a student or amateur -- find a nice building, with regularly-spaced windows. Use that as your target. You'll probably need to write your own calibration code, or modify some that's designed for square spacing, but it might work.