BTF modelling using BRDF texels

Abstract

The highest fidelity representations of realistic real-world materials currently comprise Bidirectional Texture Functions (BTF). The BTF is a six-dimensional function depending on view and illumination directions as well as on planar texture coordinates. The huge size of such measurements, typically in the form of thousands of images covering all possible combinations of illumination and viewing angles, has prohibited their practical exploitation, and obviously some compression and modelling method of these enormous BTF data spaces is inevitable. The two proposed approaches combine BTF spatial clustering with cluster index modelling by means of efficient Markov random field models. The methods allow the generation of a seamless cluster index of arbitrary size to cover large virtual 3D object surfaces. Both methods represent original BTF data using a set of local spatially dependent Bidirectional Reflectance Distribution Function (BRDF) values which are combined according to the synthesized cluster index and illumination/viewing directions by means of two types of Markov random field models. BTF data compression using both methods is about 1:200 and their synthesis is very fast.

Keywords:

• Markov random fields
• Image modelling
• Pattern recognition
• Texture synthesis
• Virtual reality

Acknowledgements

This research was supported by EC project No. FP6-507752 MUSCLE and partially by grants Nos. A2075302 and 1ET400750407 of the Grant Agency of the Academy of Sciences, Czech Republic, and MSMT project No. 1M0572, 2C06019. BTF measurements courtesy of Bonn University. Mercedes Class-C 3D model courtesy of DaimlerChrysler.