Minimal BRDF Sampling for Two-Shot Near-Field Reflectance Acquisition

摘要

We develop a method to acquire the BRDF of a homogeneous flatrnsample from only two images, taken by a near-field perspectiverncamera, and lit by a directional light source. Our method usesrnthe MERL BRDF database to determine the optimal set of lightviewrnpairs for data-driven reflectance acquisition. We develop arnmathematical framework to estimate error from a given set of measurements,rnincluding the use of multiple measurements in an imagernsimultaneously, as needed for acquisition from near-field setups.rnThe novel error metric is essential in the near-field case, where wernshow that using the condition-number alone performs poorly. Werndemonstrate practical near-field acquisition of BRDFs from onlyrnone or two input images. Our framework generalizes to configurationsrnlike a fixed camera setup, where we also develop a simplernextension to spatially-varying BRDFs by clustering the materials.