How to remove a mirror tilt

Hey

I am a PHD student and I am using for the first time Matlab in image processing. I am working on phase retrieval and using a michelson interferometer to apply the phase shifting interferometry.

For those who don't know, when using a Michelson interferometer we should tilt one mirror to obtain fringes. So after extracting the phase using this method I find myself with the good phase distribution but with the tilt of the mirror introduced in the bigenning. I need please a malab code to remove this tilt.( data is double and represent a phase distribution)

Thank you

[Jan]

This question cannot be answered yet. Of course you cannot remove a (physical) tilt just by Matlab code. It is more likely that you get a meaningful assistence here, when you post which data you have available in Matlab and explain the wanted result. The physical nature of the data are sometimes interesting, but for the Matlab code, it does not matter, what the values mean.

So please edit the original question and insert, which input data you have for the program.

[Pierre]

Thank you for your answer!

So i need to insert the data so my question can be answered?

[Neha]

how can you add tilt by mirror in matlab,,,, i mean how can you define spatial frequency cmponents in emergent orthogonal polarized components

[KALYAN ACHARJYA]

@Neha Since the question is over a decade old, you may consider posting a new fresh question with detailed information to receive better & faster responses.

Answer 1

[David Goodmanson]

Suppose the two mirrors, the source and the detector are as shown below. Assume the beam splitter does not have a wedge effect and is not birefringent, so that all waves pass straight through it with no angular deflection. If mirror 1 is rotated in the counterclockwise direction by theta1 and the same for mirror 2 and theta 2, then xfrin, the fringe spacing in x, is

xfrin = lambda / (2*(sin(theta1) + sin(theta2)))                     % lambda = wavelength

Since for a visible light interferometer the angles involved are very small, we can drop the sines and use

xfrin = lambda / (2*(theta1 + theta2))

If you know the mirror angles, it's clear how to change the fringe spacing. If one of the mirrors is unrotated, say m2, then of course

xfrin = lambda / (2*theta1)

If theta1 is known, great. If theta1 is unknown and all you have to go on is the fringe spacing xfrin, it's undetermined which direction the fringe diplacement is going, right or left. That means the angle could be +-theta1, so you would have to rotate the mirror by a small amount and see whether the fringe spacing increases or decreases.

      ---m1
       |                ^
   s---/---|m2          y  x >
       |                            
      ---d

Answer 2

[Neha]

@David Goodmanson can you help me in sagnac interferometers to generate interferogram by jones matrix