ArsTechnica

Scientific Method / Science & Exploration

Technique gets clear images from light reflected off blank paper

Taking pictures through translucent materials is also possible.

Air is reasonably transparent, but over long distances, things such as humidity, turbulence, and the like can distort images. This is because the density of air isn't the same everywhere, and it changes in time and space. The situation becomes even more complicated when trying to look through fog, biological tissues, or other inhomogeneous materials. Similarly, reflection off most (non-polished) surfaces doesn't produce coherent images, no matter how shiny the surface looks.

Despite the loss of information, researchers have developed a number of techniques to reconstruct the appearance of the original object.

Ori Katz, Eran Small, and Yaron Silberberg have now shown they can produce a fully three-dimensional image even after light has gone through thin, inhomogeneous layers. Known as turbid materials, these layers contain microscopic particles or density fluctuations that scatter light, preventing focusing. To accomplish this, they used wavefront shaping, whereby they pass the scattered light through a special modulator. This modulator produces constructive interference between light from two different wavefronts, allowing a coherent image to be produced. As a bonus, the image can be produced in real time, as opposed to related methods that require computer reconstruction.

When light passes through a turbid medium, the photons scatter off the inhomogeneities. If the source is incoherent, like an ordinary incandescent or fluorescent bulb, this results in a blurry image—if any image can be formed at all. If the light is coherent, such as a laser, scattering results is a speckled pattern. In either case, a clear view of the original object may not be possible. This spells doom for medical imaging, astronomy, and other applications. (The authors also suggested it gets in the way of peering through shower curtains. We at Ars condone such voyeuristic pursuits for consenting scientific partners only).

The researchers illuminated a printed letter "A"—the object—using an ordinary tungsten halogen lamp (a light bulb), which produces undirected incoherent white light. They used a thin polycarbonate film as the turbid medium. While a static medium like that doesn't change in time as a fog or other fluids do, the authors showed it was enough to keep the image of the "A" from forming.

By placing a spatial light modulator (SLM) immediately behind the polycarbonate film, the researchers manipulated the phase of the incoming light, changing it until light from the object interfered constructively. They also used a filter to eliminate some of the extra scattered light, leaving mostly only light from the object. Finally, they photographed the image using an ordinary digital camera. The setup looked like this:

Enlarge / The Spatial Light Modulator (SLM) changes the wavefront of the light, while the bandpass (BP) filter gets rid of any light that is still scattered.

Similarly, the researchers passed incoherent white light through a stenciled letter G. They bounced this light off a piece of white paper, using the paper like a mirror. While it's reflective, the paper certainly doesn't produce a clear image. Applying the SLM and filter as with the turbid medium, they were able to see the G with their eyes, as well as photograph it. While it isn't precisely the "seeing around corners" promised in the paper's title, it shows in principle that some non-polished surfaces can be used as mirrors, allowing coherent images to be formed using paths that might include things like walls.

Finally, the authors showed they could track moving objects in a limited way, shining light through a pinhole onto the white paper. They imaged the dot of light produced over several points in space in real time.

As they pointed out in the paper, there are two major limitations to this technique. First, the image produced isn't completely faithful to the color of the original object. After all, the method required using interference of light, which is wavelength-dependent. Second, the contrast between the resultant image and the background is very low: even if the light source was bright, the image produced was faint. This means the SLM technique is less attractive for astronomical observations, where the method known as adaptive optics (used for example at the Keck telescopes in Hawaii) preserves contrast and color.

Nevertheless, the method outlined in this paper has many potential uses in medicine, where forming real-time images through skin or thin bone is highly desirable. Since soft tissue imaging is often limited or impossible by ordinary means, wavefront reconstruction could possibly lead to significant advances in that area.

Nature Photonics, 2012. DOI: 10.1038/nphoton.2012.150  (About DOIs).

Listing image by André Mouraux

Reader Comments

There was a problem fetching the comments! Please try viewing them directly in the OpenForum

You May Also Like