Frontiers in Optics: T,W,Th

One of the things that happens to me as the years go by is that I spend less time at meetings listening to talks and more time talking to friends and colleagues and planning new research collaborations.  From discussions with said colleagues, I get the feeling that this shift in emphasis is not unique to me.  (I suppose this is why young professionals make better conference bloggers.)

So for my discussion of the last three days of the conference, let me just point out a few general observations that I had while attending.

First, there were some unconventional and very interesting talks at the conference.  On Tuesday, I attended a session on “Rogue waves and related phenomena.”  Rogue waves, also known as “freak waves”, are highly dangerous waves which can arise in open ocean, often against the prevailing winds and currents and in the absence of storms, and can attain heights of 100 ft (extreme ocean storm waves typically are no higher than 50 ft).  These waves were only positively confirmed by science in 1995, though mariners had spoken of them for at least a hundred years.  Rogue waves can sink even the largest ship in minutes, and are now thought to occur with some regularity.

Peter Janssen of the European Center for Medium-Range Weather Forecasts discussed modeling used to estimate the likelihood of rogue waves.  He noted in his talk that very little photographic evidence exists of rogue waves; I had a chance to ask him about this during the meeting, and he pointed out that buoy readings provide most of the data relating to rogue wave behavior.

What is the connection to optics?  Rogue waves are modeled by the nonlinear Schrödinger equation, which also can be used to describe nonlinear effects in optical systems.  A study of one system therefore gives some insight into the other.

Plasmonics and metamaterials research remains quite popular; there were no fewer than 10 sessions on the topics.  Plasmonics sessions seemed to be much more applications oriented (“plasmonic emitters and resonators”, “plasmonic sensors”, “plasmonic waveguides and devices”), which suggested that the field has matured enough that we may start to see some really interesting technological output related to plasmons in the near future.

There were also a surprising number of sessions on X-ray generation and imaging, somewhat unusual for an “optics” meeting!  It seems that new and improved methods of doing imaging with X-rays is leading to a resurgence in popularity of the subject.

Two other imaging concepts seemed to be very “hot” at this meeting, and are worth saying a few more words about: compressive sensing and “ghost” imaging.  I knew relatively little about either topic before going to the meeting, an oversight I’m now working to correct!

Compressive sensing refers to the measurement of an image at a resolution much higher than the resolution of the measuring device.  For example, a “compressive imaging camera” might be able to record a 200 by 200 pixel image using only a 100 by 100 pixel detector.

The genesis of this idea comes from image compression, such as the jpeg compression done by digital cameras: my Kodak Z1012IS camera, for instance, has 10 million pixels, but produces an image which is only 2 million bytes in size.  Since a single color pixel requires at least three bytes of storage, this suggests that the stored image is a factor of ten smaller than the amount of data actually recorded by the camera.  How is this possible?  Most images contain a very large amount of redunancy in them: as a crude example, if I took a picture of a perfectly white wall (or a polar bear in a snowstorm), my image could be characterized by a single RGB color: the particular shade of white of the wall.  Since most scenes we photograph have some amount of redundancy (forests are green, the sky is blue, goth clubs are black), a standard camera is typically recording more information than it needs.

The philosophy of compressive sensing (or compressive imaging) is to design an optical system that is, in a sense, optimally efficient.  Instead of measuring too much data and throwing out the redundant information, one measures the minimal amount of data and uses signal processing techniques to reconstruct an image of much higher resolution.  Strategies for doing so seem to involve a combination of developing novel optical devices which collect data in unusual ways and computational techniques to analyze said data.

“Ghost” imaging is a technique that is in some sense an extreme version of compressive sensing: the main detector has only a single pixel!  The crux of the technique is the comparison of the intensities of two optical beams, one of which has interacted with the object to be imaged.  The original version of the experiment utilized the quantum correlations associated with entangled photons, as shown in the schematic below:


Light from a thermal source is split into two paths by a beam splitter, one of which illuminates the object to be imaged and the other of which illuminates a CCD camera.  It should be noted that the CCD camera does not have any view of the object.  Light scattered from the object is recorded by a “bucket” detector, and signals from this “bucket” are correlated with photons arriving at the CCD camera.  By only keeping CCD signals which are correlated with the “bucket” signals, one remarkably finds that an image of the object can be reconstructed on the CCD camera!

I’ll come back and describe “ghost” imaging in more detail in a future post.  It should be noted, however, that researchers have determined that quantum effects are not strictly necessary, and a classical version of ghost imaging has been demonstrated.  The consequences and applications of such imaging strategies are not immediately obvious to me, but it is a very clever idea.

The OSA meeting seemed rather quiet this year, overall.  I suspect that attendance was down due to the ongoing financial crisis.  I still had a great time and had a lot of productive discussions, but here’s hoping that next year’s meeting, in Rochester, will be back up to speed.

P.S. I should give a shout-out to Maceió Brazilian Steakhouse, which I can highly recommend if you happen to be in downtown San Jose!  There’s one thing on the menu: the rotational dinner, which involves the servers bringing around skewers of 14 different types of meat until you beg them to stop!  Nancy, the proprietor, is a very nice lady and made our group feel right at home.

This entry was posted in Optics, Science news. Bookmark the permalink.

4 Responses to Frontiers in Optics: T,W,Th

    • Hi Ori,

      Thanks for the comment! As it turns out, I was at your talk – very nice work, and very nicely presented! I’ll hopefully come back and take a closer look at the research and blog about it in the near future.

  1. Wade Walker says:

    Interesting! The October 2009 issue of Physics Today has a mini-article on ghost imaging, but I didn’t make the connection with the Hanbury Brown and Twiss effect ( until I saw your diagram, which looks similar to the usual HBT diagram.

    The two effects look very similar to my untrained eye, but there seems to be some controversy in the references I could find on the subject. Did the presenters have anything to say about the subject?

    • I didn’t hear of any explicit controversy from the talks I attended, but my impression is that there was an early argument about whether or not there was something inherently quantum-mechanical about the effect. It is now clear, and has been demonstrated, that one doesn’t need quantum mechanics to do a form of ghost imaging, but it is not clear (at least to me) whether quantum effects add something to the mix.

      This sort of controversy is very similar to that which appeared when the HBT experiment was first reported. Researchers attempted to reproduce the HBT experiment with laser light, with negative results. It turns out that natural light is necessary to get meaningful HBT data; it seems that the same is true for ghost imaging.

      I’ll try and sort through this in more detail in a future post; now I’m curious…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.