Judith Radloff

Obtaining the Bidirectional Texture Reflectance of Real-World Surfaces by means of a Kaleidoscope

Abstract

This thesis investigates the use of a kaleidoscope in measuring the reflectance properties of realworld

surfaces, in particular to obtain the Bidirectional Texture Function (BTF) of the surfaces.

The kaleidoscope technique for reflectance measurement represents an innovative economical

and simple approach to obtaining reflectance functions of real-world surfaces. While theoretically

the technique has been shown to be of merit, there has been no prior research conducted to

investigate practical issues arising from its implementation; neither have optimal kaleidoscope

or other design configurations been derived.

This research is an investigation of the kaleidoscope technique as an economical and simple

means of obtaining BTF information of surfaces. The investigation is conducted in two phases.

The first phase involves a simulation of kaleidoscope architectures in order to study the effects

of the structural parameters of kaleidoscopes, using this study to derive optimal kaleidoscope

configurations for use in surface reflectance measurement. The second phase investigates the

technique from a practical perspective through a complete working physical implementation of

the system, highlighting pertinent practical issues as well as potential optimisations.

The investigation reveals design features that would yield optimal kaleidoscope con.gurations

for reflectance measurement. These features have been used to design five kaleidoscope

configurations, the first of which has successfully been used to obtain an extensive BTF database

consisting of 585 view/illumination direction pairs. This makes it one of the largest recorded BTF

databases to date. The results show that the technique has significant merit in its efficiency and

accuracy at a fraction of the overhead required for current alternative techniques. Optimisations

are proposed in the thesis which could lead to the acquisition of even richer BTF information.