Projects in VR Experimental Platforms for Computational Photography

The principles of photography have remained largely unchanged since its invention by Joseph Nicéphore Niépce in the 1820s. A lens focuses light from the scene onto a photosensitive plate, which records this information directly to form a picture. Because this picture is a simple copy of the optical image reaching the plate, improvements in image quality have been achieved primarily by refi ning the optics and the recording method. These refi nements have been dramatic over the past few decades, particularly with the switchover from fi lm to digital sensors, but they’ve been incremental. Computational photography challenges this view. It instead considers the image the sensor gathers to be intermediate data, and it uses computation to form the picture. Often, it requires multiple images to be captured, which are combined in some way to produce the fi nal picture. Representative techniques include high-dynamic-range (HDR) imaging, fl ash/no-fl ash imaging, coded-aperture and coded-exposure imaging, photography under structured illumination, multiperspective and panoramic stitching, digital photomontage, all-focus imaging, and light-fi eld imaging. In this article, I look at the lack of experimental platforms for computational photography (that is, cameras that are programmable), and I talk about one possible solution—the Frankencamera architecture designed in my laboratory at Stanford as part of our Camera 2.0 project (http://graphics. stanford.edu/projects/camera-2.0).