How to detect the phase of a laser beam directly?

[xhtp2000]

Holography is based on the simultaneous record of the phase and intensity of the interference pattern. I'm wondering if one can figure out a way to record the phase of the reflected laser beam from the 3D object without the reference laser beam?

 

[Vanadium 50]

Phase with respect to what?

 

[xhtp2000]

with respect to time and space, i.e. time and space resolved record of phase

 

[sophiecentaur]

There is no absolute phase. For any diffraction, you are always dealing with the phase of one wave relative to another. With laser light, the coherence of the beams is so enormous that you can have very large path length differences and still get a diffraction pattern. This is why we had to wait for lasers to be invented before optical holograms could be produced.

 

[xhtp2000]

thanks a lot and I agree with that. One can only indirectly record the phase by comparison with another reference beam. But as is known, the stereo perception of human being never depends on a reference beam. So I believe in that there should be some measuring method to directly record the phase of a laser beam. Such as the impedance spectroscopy, one can record the imaginary and real parts of the impedance of a ultra sonic wave respectively and simultaneously.

 

[f95toli]

thanks a lot and I agree with that. One can only indirectly record the phase by comparison with another reference beam. But as is known, the stereo perception of human being never depends on a reference beam. So I believe in that there should be some measuring method to directly record the phase of a laser beam. Such as the impedance spectroscopy, one can record the imaginary and real parts of the impedance of a ultra sonic wave respectively and simultaneously.

But again, there is no such thing as absolute phase. Whenever we use phase information we are comparing two signals to each other, i.e. we are looking at the difference in phase.

And in impedance spectroscopy one is usually measuring the signal using say a lock-in amplifier (with the x and the y outputs corresponding to the real and the imaginary part of the signal); meaning you are measuring the phase with respect to the reference signal of the lock-in (which in turn allow you to calculate the phase difference).

 

[xhtp2000]

It's good to know that. But at least, if one can pre-set the reference as a lock-in amplifier in the impedance spectroscopy, then one can bypass the difficulty in finding a highly coherent laser beam. Ideally, if one can set such kind of imaginary reference beam, even one single laser spot can include all the 3D image information of a object.

 

[sophiecentaur]

thanks a lot and I agree with that. One can only indirectly record the phase by comparison with another reference beam. But as is known, the stereo perception of human being never depends on a reference beam. So I believe in that there should be some measuring method to directly record the phase of a laser beam. Such as the impedance spectroscopy, one can record the imaginary and real parts of the impedance of a ultra sonic wave respectively and simultaneously.

Woah there. Stereoscopy is not an interference phenomenon. It does, however, require two images of the given object for mutual comparison (each referred to the other).

"Real and Imaginary" are terms that imply a phase reference in any case. You still need an origin to measure the phase against and that origin is totally arbitrary. There has to be a reference somewhere. Look into "impedance spectroscopy' (whatever that may be - )and you will find that there is a reference in there somewhere; I can guarantee.

 

[sophiecentaur]

even one single laser spot can include all the 3D image information of a object.

Not strictly true. The smaller the spot, the more fuzzy will be the reconstructed image.

But you are jumping from lock in amplifiers, phase references to lasers. There seems to be a bit of a muddle here?

 

[xhtp2000]

But you are jumping from lock in amplifiers, phase references to lasers. There seems to be a bit of a muddle here?

Because I have been up to real-time 3D monitoring lately and want to make the setup as simple as possible. So far, very few techniques are feasible for quick real-time 3D imaging from a little fraction of sampling except for holograph. Thanks so much anyway!

 

[sophiecentaur]

It sounds like you may be more interested in microwave or ultrasound holography than using lasers (more appropriate wavelength plus achievable signal processing).

 

[Andy Resnick]

thanks a lot and I agree with that. One can only indirectly record the phase by comparison with another reference beam. But as is known, the stereo perception of human being never depends on a reference beam. So I believe in that there should be some measuring method to directly record the phase of a laser beam. Such as the impedance spectroscopy, one can record the imaginary and real parts of the impedance of a ultra sonic wave respectively and simultaneously.

That's correct- if you had a way to coherently detect visible radiation, you could simultaneously record the amplitude and phase. At this time, there are no devices that can do that- the best we can do now is to use a reference oscillator and perform heterodyne detection (and I'm sure that can even be done at visible frequencies).

AFAIK, coherent detection can only go up to THz right now.

 

[sophiecentaur]

Actually, phase is not the only way of approaching your problem. XRay tomography just works on 'ray' treatment and rotating the subject - calculating the layout from the density patterns of a number of slices. MRI scans, likewise, do not use phase information yet are very high res.

Depending on how complex your system could be, you could use multiple images of the scene, from several different directions to produce a stereo type of image but more so. It would depend upon the actual scene that you need to get the positional information out of. Would it be a general scene or could it be a purpose built enclosure with suitable guide markings on the walls? The latter would make it easier to compute the size and location of an object. I guess what I'm describing could be called a parallax based system.

 

[Andy Resnick]

Because I have been up to real-time 3D monitoring lately and want to make the setup as simple as possible. So far, very few techniques are feasible for quick real-time 3D imaging from a little fraction of sampling except for holograph. Thanks so much anyway!

What spatial resolution do you need? I worked with systems that can 3-D map a scene at <1cm accuracy with a maximum range of 1 km, but ththere has been major improvements since:

http://en.wikipedia.org/wiki/3D_scanner#Non-contact_active

 

[xhtp2000]

I agree with that. Actually there's no measuring method to record the phase directly. If you take a took at the Euler's formula, exp(ix)=cosx+isinx, x=arctan(sinx/cosx), the definition of phase is given by a value with respect to a reference. In practice, e.g. for RADAR transmitter, two real waves are emitted, i.e. cosx as the real part and sinx as the imaginary part of the signal of complex number, that's just a convenience for the maths deduce. By comparison of these two real waves (one as the reference), the positioning of the object can be realized. Is that really a simplification or a misleading? I mean, may be, there's a direct expression for the phase instead of the complex number. By the way, how do you understand "comparison"? The detection wave can be compared with the reference wave, but one photon can still interact with itself in a double-slit experiment Young's experiment as diffraction, if photons pass the slits one by one. In the micro-world, each comparison is actually a interaction. I mean our world is existent because of comparison or interaction. Even within one photon, there's another world there and it is self-contained, and it can interact with itself, there might be another reference.

 

[xhtp2000]

Actually, phase is not the only way of approaching your problem. XRay tomography just works on 'ray' treatment and rotating the subject - calculating the layout from the density patterns of a number of slices. MRI scans, likewise, do not use phase information yet are very high res.

Depending on how complex your system could be, you could use multiple images of the scene, from several different directions to produce a stereo type of image but more so. It would depend upon the actual scene that you need to get the positional information out of. Would it be a general scene or could it be a purpose built enclosure with suitable guide markings on the walls? The latter would make it easier to compute the size and location of an object. I guess what I'm describing could be called a parallax based system.

Thanks a lot! It deals with the on-line quality control of the streams on belt conveyor

 

[xhtp2000]

What spatial resolution do you need? I worked with systems that can 3-D map a scene at <1cm accuracy with a maximum range of 1 km, but ththere has been major improvements since:

http://en.wikipedia.org/wiki/3D_scanner#Non-contact_active

Thanks a lot! For my setup, I think 1mm resolution and a range of 100 m should be enough.

 

[sophiecentaur]

Does your measurement need to be from so far away? Actual position wouldn't' be as important as shape, presumably? How about using comparison pictures / outlines an then using photo imaging and corelation? The best solution will be very case specific.