The interaction between light and material leads to beautiful visual phenomena that greatly enrich our perception of the world. The ability to measure and model light scattering is central to almost any field of science. However, light transport in rich scenes is a complex process involving a long sequence of scattering events. Computationally modeling, reproducing and acquiring the processes generated so easily by Mother Nature is an extremely challenging task. While several computational models have been proposed, they are all making various simplifying assumptions that cannot capture the full complexity of light transport processes in nature. This research develops new measurement strategies and new inference algorithms allowing us to infer more information on light and material interaction.
Specifically, our research has focused at the following tasks:
(i) Acquiring subsurface scattering parameters. We develop algorithms which image light as it scatters through material and infer the volumetric structure of the material as well as its optical properties, such as the type and refractive properties of the particles composing the material, their size and density. As part of this research we have developed a through mathematical framework for simulating and modeling the statistical properties of coherently scattered light and the resulting speckle patterns.
(ii) Acquiring internal illumination: by capturing images of scattering materials such as tissue and analyzing speckle variations we illuminators that are not directly visible to the camera. In particular, fluorescent illumination sources deep inside biological tissue, which is an extremely central challenge in biomedical imaging.
(iii) Developing light sensitive displays, capable of presenting 3D scenes with spatially varying reflectance properties.
As light scattering is such a fundamental phenomenon, our tools have applications in almost any field of science, from astronomy to microscopy, and in medicine. In particular, the research has significantly advanced our ability to image deep inside biological tissue.
