CHS: Small: Developing and Validating a Physically Accurate Light-Scattering Model for the Rendering of Bird and Other Dinosaur Feathers

点击翻译按钮获取中文摘要

Although computer graphics has become a crucial tool for scientific visualization (and many other fields such as medicine, industry, education, and communication), more realistic, fine-scale representations of the natural world are still sorely needed. Rendering, a core component of computer graphics, uses models of light and material interaction to generate today's most photorealistic images, but there are many real-world materials yet to be fundamentally studied in this regard. In ornithology and paleontology, for example, researchers continue to investigate feather micro-structures in an effort to discover how these produce color and iridescence, and to reveal the similarities and differences between feathers of modern birds and those of extinct dinosaurs, but to date no sufficiently sophisticated model has been developed to represent the complex, multi-scale interaction of light with feather sub-structures. This interdisciplinary project will lead to the first computer-graphics appearance model specifically developed for feathers, by collecting and building on biophysical data to design a biologically realistic rendering method. Additional broad impacts will derive from the project outcomes contributing to educational activities related to birds and other dinosaurs while transforming rendering capabilities to create photorealistic images of feathers and feathered creatures for scientific visualization, illustration, documentaries, and interactive graphics. The appearance of real-world feather anatomy is the result of light interactions with complex, patterned structures of varying scale. Currently, fiber-shading models of mammalian hair and fur are often used for convenience, despite their quite different structural and mechanical properties; inherently, therefore, such models do not accurately recreate the scattering properties of feathers. Furthermore, structural coloration, a wave-optics property not represented at all in fiber-shading models, is present in many feathers. To address these deficiencies, novel material models will be developed by collecting and observing physical data of varying scales from feathers. Broad taxonomic sampling will make manifest important variation among birds and facilitate imaging of a wide variety of microscopic structures. Light-scattering data will be recorded to develop and validate the new physical material models. Greater understanding of these multi-scale light interactions as well as the complex phenomenon of structural coloration holds potential to advance perspectives in several fields. Visualizations and comparisons between living-bird feathers and non-avian dinosaur fossils will increase understanding of feather evolution, be applicable to studies of signaling and evolution of mate choice in ornithology, and improve sophistication of fiber and wave-optics models in computer graphics.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.