Representing spectral functions by a composite model of smooth andspiky components for efficient full-spectrum photorealism

摘要

We propose a new model called the “composite model” torepresent spectral functions. This model is built on the idea ofdecomposing all spectral functions into smooth and spiky components,each with its own distinct representation. A smooth spectrum can beexpressed with coefficients of a set of given basis functions, and thediscrete spikes in a spiky spectrum with their locations and heights.For the smooth part, we propose re-sampling functions that arereconstructed from the coefficients in a linear combination to improveefficiency. Spectral multiplication is thus greatly reduced incomplexity. This new model shows remarkable advantages in representingspectral functions with aspect to accuracy, compactness, computationalefficiency, portability, and flexibility, and it has a great applicationpotential in color science, realistic image synthesis, and color imageanalysis. Here we apply it to rendering images involving real spikyilluminants as well as objects with light dispersion. The compositemodel is shown to surpass other models in these applications