Digital Modeling of Material Appearance

by Julie Dorsey, Holly Rushmeier, François Sillion

Overview

Computer graphics systems are capable of generating stunningly realistic images of objects that have never physically existed. In order for computers to create these accurately detailed images, digital models of appearance must include robust data to give viewers a credible visual impression of the depicted materials. In particular, digital models demonstrating the nuances of how materials interact with light are essential to this capability.

This is the first comprehensive work on the digital modeling of material appearance: it explains how models from physics and engineering are combined with keen observation skills for use in computer graphics rendering.

Written by the foremost experts in appearance modeling and rendering, this book is for practitioners who want a general framework for understanding material modeling tools, and also for researchers pursuing the development of new modeling techniques. The text is not a "how to" guide for a particular software system. Instead, it provides a thorough discussion of foundations and detailed coverage of key advances.

Practitioners and researchers in applications such as architecture, theater, product development, cultural heritage documentation, visual simulation and training, as well as traditional digital application areas such as feature film, television, and computer games, will benefit from this much needed resource.

ABOUT THE AUTHORS
Julie Dorsey and Holly Rushmeier are professors in the Computer Science Department at Yale University and co-directors of the Yale Computer Graphics Group. François Sillion is a senior researcher with INRIA (Institut National de Recherche en Informatique et Automatique), and director of its Grenoble Rhône-Alpes research center.

* First comprehensive treatment of the digital modeling of material appearance;
* Provides a foundation for modeling appearance, based on the physics of how light interacts with materials, how people perceive appearance, and the implications of rendering appearance on a digital computer;
* An invaluable, one-stop resource for practitioners and researchers in a variety of fields dealing with the digital modeling of material appearance.

Product Details

ISBN-13:	9780080556710
Publisher:	Elsevier Science
Publication date:	07/21/2010
Series:	Morgan Kaufmann Series in Computer Graphics
Sold by:	Barnes & Noble
Format:	eBook
Pages:	336
File size:	11 MB
Note:	This product may take a few minutes to download.

Read an Excerpt

Digital Modeling of Material Appearance

---

By JULIE DORSEY HOLLY RUSHMEIER FRANÇOIS SILLION

MORGAN KAUFMANN

Copyright © 2008 Elsevier Inc.
All right reserved.
ISBN: 978-0-08-055671-0

---

Chapter One

INTRODUCTION

Computer graphics systems are capable of generating imagery of stunning realism. For feature film and games, entire new visual worlds are created and synthetic scenes and characters are mixed seamlessly with recorded live action. Designers, architects, and engineers are able to visualize and evaluate product concepts in realistic settings. Historians and archaeologists are able to reconstruct a visual world of the past. The development of accurate digital models of material appearance has been an essential element in the building of this capability. This book presents the foundations of modeling material appearance.

Systems to generate realistic images are the result of efforts of three diverse communities: computer science researchers, software engineers, and artists and designers. Researchers have built mathematical models of materials based on many different disciplines. Computer scientists have adapted and expanded models of material light interaction that were originally developed in areas such as optics, biology, and various branches of engineering. Using these models, they have developed algorithms to efficiently capture different aspects of appearance. Software engineers have implemented these algorithms into systems that model an object's shape and illuminate its environment as well as material. Artists and designers use the various systems to produce visual output. The process is not just a one-way flow of information. To create what they envision, artists and designers often press systems and use them in ways not originally intended, as well as push software engineers and researchers to produce new and improved models and systems. Software engineers often enhance and expand the algorithms that come directly from researchers.

The result of this activity among the different communities involved in computer graphics has been a rich but disorganized body of work in the area of digital modeling of appearance. Many different classes of materials have been modeled, and many different classes of algorithms have been developed for computing the appearance of these materials. Some common material models are so complex that they require setting in excess of 50 parameters to obtain a particular look. Literally thousands of web sites directed at the full range of users; from hobbyists to professionals, are devoted to explaining how to achieve different effects. It is becoming increasingly difficult for individuals to learn what has been done, what can be done, and what needs to be done in the area of digital appearance modeling.

This book provides a common foundation for modeling appearance. This foundation is based on the physics of how light interacts with materials, how people perceive and specify appearance, and what the implications of encoding models are on a digital computer. While all communities are not going to explore each of these areas in equal depth, considering all of the issues will provide a basis for researchers and practitioners to more effectively interact.

Chapter 2 contains background information with a perspective on the major issues involved in materials modeling: the physics of light and human visual perception. These are broad topics and we describe the essentials that are needed to start modeling materials. First, we identify the three elements—shape, material, and incident light—that produce a visual image. Next, we identify the three major aspects of materials: spectral, directional, and spatial variations.

With this basic background outlined, we then consider different classes of materials in Chapter 3, Observation and Classification. For researchers and software engineers, it is essential to spend time looking at materials in the world before addressing mathematical, or any computational, considerations. We take as inspiration the training that artists have had in observation. However, in contrast to observation in art, the goal of the chapter is to observe, in a more technical framework, the way appearance is affected by how light rays reach the eye from a source.

Ultimately,toencodemodelsonadigitalcomputer,weneedmathematicaldescriptions. In Chapter 4, Mathematical Terms, we define the basic terms, concepts, and notations used in describing light and materials. For many years, computer graphics systems used ad hoc descriptions of shades and lights making it difficult to compare and combine methods across different vocabularies, units, and scales. Successful realistic modeling was advanced by adopting the standard terms used by the illumination and radiative transfer communities. While understanding the definitions of radiance and reflectance can be challenging, it is the key to being able to understand material models.

Using basic mathematical terms, Chapter 5, General Material Models, presents the basic models for how materials scatter light. The chapter traces the development of the various named reflectance models (e.g., Phong, Blinn, Cook–Torrance, and Oren–Nayar) that appear in most graphics systems. It also puts in context more complex models that account for interference, diffraction, and volumetric scattering.

While many materials can be defined as simple combinations of the general material models, other materials require specific models of small-scale geometric structure and multiple spatially varying layers. Chapter 6, Specialized Material Models, provides a guide to these specific models, which have been published in diverse, more specialized, conference documents and journals, as well as in major conference—ACM, IEEE, and Eurographics—proceedings.

Whichever model is used for materials, we need some means to determine the parameters for it. In a design setting, parameters may be selected to achieve a particular effect. In many other cases, a model of a particular existing material that matches suitable benchmarks is needed. Chapter 7, Measurement, discusses the different techniques available for obtaining data for existing materials.

Materials have a temporal, as well as a spatial, dimension. Material appearance depends on the history of usage and environment and the basic chemical components and surface finish. Materials may be purposefully treated to achieve particular end results or may change as the result of natural aging or weathering. Chapter 8, Aging and Weathering, presents methods that either simulate temporal evolution of changes in materials or directly model the visual results of processed materials. This is still an emerging area of materials modeling, and open questions and issues are identified.

Chapter 9, Specifying and Encoding Appearance Descriptions, considers how various models are efficiently encoded, and how these models are presented to the user for modification. Representation in terms of different basis functions expose different types of controls to the user. Further, the chapter considers how material models are associated with shapes to specify complete object descriptions.

Finally, Chapter 10, Rendering Appearance, looks at how material models are integrated into complete systems for generating images. Different representations of models are used for different rendering approaches. Different models can be appropriate for offline and real-time systems.

This is the first comprehensive work on digital modeling of the appearance of materials. It is not a "how-to" book in the sense of providing a series of screenshots showing how to set parameters in a particular software system, or pseudocode for rendering a particular material. The book is meant to be a guide of how to decompose the appearance of materials so that they can be represented digitally. It identifies the basic appearance attributes of materials. In addition, the book presents definitions and principles that have been applied to modeling general classes of materials. Finally, the book is a guide to the range of material appearance effects that have been considered in computer graphics to date. For researchers, we hope the book helps to focus future work in this area. For practitioners, we hope the book provides a perspective to help make sense of the wide array of tools currently available.

Chapter Two

BACKGROUND

"Material appearance" is the visual impression we have of a material. To model material appearance, we need to express, in the form of data and algorithms, everything needed to simulate how the material would look from any view, in any environment. Modeling materials is just one area of computer graphics, however, it is a complicated area that builds on models from a variety of other disciplines. In this chapter, we introduce some basic principles from other disciplines that we use as a starting point for modeling.

Modeling the appearance of an object's material gives us the capability to simulate what an object would look like. How the object will look depends on both physical and psychological phenomena. Light leaving an object that reaches our eye is the physical stimulus that allows us to see it. The connection between that stimulus and the idea that we form about the object is a complex result of the physiology of our eyes and the processing in our brains. The vehicle we use to present the simulation of an object is a two-dimensional image. In generating images from digital models, we want the display to produce a physical stimulus that will generate the same idea of how an object would appear if we had viewed it directly (Figure 2.1). To do this, we build on knowledge of the physics of light, human perception, and image formation.

The physics of light has been intensely studied for centuries. Our understanding of light has moved from the Greek's corpuscular theory, to Huygen's waves, to the theory of quantum electrodynamics, and continues to expand. In modeling appearance, our goal is not to incorporate the entire body of knowledge that has been accumulated.

It is generally not appropriate, or useful, to use the most low-level, detailed models of light that have been developed in physics. What we want to do is extract from what is known about light the models that are relevant to the human process of seeing an object.

Perception and the human mind have also been studied for centuries. Our understanding of how we form ideas about what we see is in a far more primitive state than our understanding of physical light. There have been a great deal of data collected for specific aspects of human vision and cognition, and many different models have been developed to explain these observations. As in the case of the physics of light, our goal is not to try to incorporate all of the models that have been formed for vision and cognition. Unlike physics, however, this is not because these models are not relevant to our goal. The problem is that in the field of psychology there is no comprehensive model that can reliably combine even a substantial subset of these individual observations.

We can observe, however, how we describe our impressions of objects. We refer to objects as being, for instance, "shiny red," "mottled gray," or "bumpy." We can separate out these various impressions into different types. Further, for different types of impressions, we can use models of the relationship of an impression, depending on the type of impression, to the physical light we encounter at different levels of detail. In many cases, the statements we can reliably make about the relationship between physical light and the impression we form will be very crude. However, even simple approximate relationships provide us with significant insight into how we should model light interactions to produce the appearance of an object.

In computer graphics, we combine descriptions of physical light and human perception to form images. To form an image, we need to model each object's shape, material, and the light incident on the object. Given these ingredients, we can generate an image for a human observer at a specific location and view direction.

Light, perception, and image formation are each topics that could be the subjects of a book, or series of books, on their own. Here, we briefly discuss some basic ideas from these areas that are needed as background for the detailed material models we will present in the following chapters.

2.1 LIGHT

Light is basic to our existence, and we think of it in many different ways. Light is the product that comes from the sun or a lamp that allows us to see things. Light leaves a source, strikes objects, and eventually hits our eyes. We only see objects for which there is a direct, unobstructed path for light to follow from the object to our eyes. If there is an unobstructed path from a light source to the object, the object will appear bright to us. If the path from the light source to the object is blocked, the object will look dark, or shadowed, to us.

(Continues...)

---

Excerpted from Digital Modeling of Material Appearance by JULIE DORSEY HOLLY RUSHMEIER FRANÇOIS SILLION Copyright © 2008 by Elsevier Inc. . Excerpted by permission of MORGAN KAUFMANN. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---

Table of Contents

'Chapter 1 Introduction
Chapter 2 Background
Chapter 3 Observation and Classification
Chapter 4 Mathematical Terms
Chapter 5 General Material Models
Chapter 6 Specialized Material Models
Chapter 7 Measurement
Chapter 8 Aging and Weathering
Chapter 9 Specifying and Encoding Appearance Descriptions
Chapter 10 Rendering Appearance'