The work of Ren Ng on digital light field photography is coming to fruition and has the potential to revolutionize picture-taking. Lytro Inc., the company founded by Ng, recently announced that it will offer a consumer-level digital light field camera by the end of this year. A light field camera allows digital images to be captured now and focused later. When the pictures are shared, the viewer can choose which part of the image is in focus: the foreground, the background, or something in between. It will allow people to capture a moment and dynamically relive it and share it in different ways, as determined by the eye of the beholder of the photograph instead of just the eye of the photographer. The images are described as “living pictures” that can be altered in real time with the click of a mouse. You can try it for yourself by viewingLytro’s picture gallery. In addition, pictures can converted for viewing in 3D. Dig out your red and blue 3D glasses and take a look at the demo on Lytro’s blog site.

These capabilities are made possible by using software and computing power to alter the image after it is acquired. In addition to making the novel features of the Lytro camera possible, digital post-processing can also be used to remove aberrations and improve image quality. Ng discusses digital correction of lens aberrations in light field photography is his Ph.D. thesis.

For more “standard” imaging methods, post processing for distortion correction is very important for some applications. In aerial photography, for example, obtaining an accurate representation of what is on the ground can be critical. ASE Optics has patented a methodology and optical system as well as an aerial camera that can characterize errors in an image or portion of an image and correct for them. A set of fiducials is projected onto the focal plane array as an image is being acquired. The difference between the ideal and actual fiducial locations is used to characterize the main optical system. Distortion coefficients are computed to back out aberrations. Since the fiducials are created along with every image, the effects of dynamic changes that cause errors during image acquisition can also be removed. The result is improved image quality and more precise geolocation information.

It will be interesting to see what the future has in store for photography as ever-increasing computing power leads to more and more post-processing capabilities.

Are you shifting to post-processing in any applications? If ASE's computational imaging capabilities can help, consider our discovery service.

I just caught an interesting interview with Dr. Henry Petroski over on the Groks Science Radio Show (a quirky show that I loved long before its co-producer, Charles Lee, moved to my undergraduate alma mater). Petroski is a professor of civil engineering over at Duke, and he's got a new book out entitled Essential Engineering, which deals with why engineering is essential in the modern world, and how engineering is different from basic science. The subtitle of the book is "Why Science Alone Will Not Solve Our Global Problems." That phrasing may sound critical of basic science, but that certainly isn't the intent. Rather Petroski is seeking to elevate the importance of engineering in the public eye so that science and engineering are seen as a partnership and not as a hierarchy. In particular he takes issue with the common misconception that engineering is "just applied science." On the contrary people often engineer a device before the scientific principles underlying it are understood. This is certainly true in optics. For example magnifying lenses have been used for thousands of years (for example Aristophanes mentions a "crystal lens" in his play The Clouds, circa 420 BCE), but the theory of refraction was not described mathematically until around 1000 CE. (As to who first devised the laws of refraction... I'll let the debate continue to rage on Wikipedia.)

It's been a really hectic few months here, what with SBIR proposals, a fewconferences, and, of course, actual engineering. We're not complaining about being busy, of course, and we recently hired a new scientist, Wade Cook, to act as Engineering Manager. You can expect to read posts from him here in the near future.