RayTracingFromTheGroundUp-光线跟踪算法技术-KevinSuffern

Ray Tracing From The Ground Up-光线跟踪算法技术-Kevin Sufferneditorial, sales, and Customer Service OfficeAKIeters ltdksk worcester Street, suite 230)Wellesly, MA(2482w wlvakpeters((ImCopyright 2(10)7 by A K Peters, LtdAll rig hts reserved. No part of the material protected by this copyright notice may be repro-duc ed or utili.ed in any torm, electronic or mechanical, including photocopying, recording,tr by any information storage and retrieval system, without written permission from the('opyrig; ht ownerI ibrary of Congress Cataloging-in-Publication DataSutfern, Kevin gRay tracing from the ground up/Kevin G. SuffernCnIncludes bibliographical references and index.ISBN 978-1-56881-272-4(alk. paper)omputer graphIcsI. Title「、385.579952007)()6.6-(dc222007021706The idge on the cover is based on the fishbowl object described in Chapter 28. The gold-tish model, courtesy of James MeNess, is rendered using a triangle mesh, as described inpters 22 and 23ChaptePrinted in india11100908070987654321In memory ofSilvia gladys Suffern, 1910-2003Lucy Suffern 1977-1997To Elleen, I could not have written this without youContentsForewordXPrefaceAcknowledgmentsXIXRay Tracer Design and Programming2 Some Essential mathematics173 Bare-Bones Ray Tracing45Antialiasing815 Sampling techniques936 Mapping samples to a disk119125789Mapping Samples to a HemispherePerspective Viewing133A Practical Viewing System15110 Depth of Field1671 Nonlinear Projections18112Stereoscopy19713 Theoretical Foundations21714 Lights and materials245vContents15 Specular Reflection27916 Shadows29317 Ambient Occlusion30918 Area Lights32519 Ray-Object Intersections35320 Affine Transformations39721 Transforming objects41722 Regular Grids44323 Triangle Meshes47324 Mirror Reflection49325 Glossy Reflection52926 Globa|‖ amination54327 Simple Transparency56128Realistic Transparency59329Texture Mapping64330 Procedural Textures67131 Noise-Based Textures691Bibliography733Index745ForewordComputer graphics involves simulating the distribution of light in a 3Denvironment. There are only a few fundamentally different algorithms thathave survived the test of time. They can be loosely classified into projectivealgorithms and image-space algorithms. The former class projects each geo-metric primitive onto the image plane, with local shading taking care of theappearance of objects. This class of algorithm is still widely used because it isamenable to pipeline processing and therefore to hardware implementation asevidenced by all modern graphics cardsImage-space algorithms compute the color of each pixel by figuring outwhere the light came from for that pixel. Here, the basic operation is to determine the nearest object along a line of sight. Following light back along a linehas given this basic operation and the associated image-synthesis algorithmheir name: ray tracing.In 1980, ray tracing was at the forefront of science. The quality of theimages that can be computed with ray tracing was an eye opener, as it natu-rally includes light paths such as specular reflection and transmission, whichare difficult to compute with projective algorithms. Some shapes are easier tointersect rays with than others and in those early days spheres featured heavily in ray-traced images. Hence, old images often contained shiny spheres todemonstrate the power of ray tracinga vast amount of research was then expended to make ray tracing bothmore tractable and to include more features. variants were introduced, forinstance, that compute diffuse inter-reflection, caustics, andyor participatingmedia. To speed up image generation many data structures were developedthat spatially sort the 3d geometry. Spatial subdivision algorithms allowa very substantial reduction of the candidate set of objects that need to beXrewordintersected to find the nearest object for each ray. Ray tracing is also amen-able to parallel processing and has therefore attracted a substantial amount ofresearch in that areaAll of this work moved ray tracing from being barely tractable, to justabout doable for those who had state-of-the-art computers and plenty of timeto kill. High-quality rendering tends to take a whole night to complete, mostlybecause this allows artists to start a new rendering before going home, to findthe finished image ready when they arrive at work the next day. This, by theway, still holds true. For many practical applications, hardware and algorithmic improvements are used for rendering larger environments, or to includemore advanced shading, rather than to reduce the computation timeOn the other hand, more than 25 years after its introduction, ray tracinghas found a new lease on life in the form of interactive and real-time implementations. Such rendering speeds are obtained by using a combination ofsuper-fast modern hardware, parallel processing, state-of-the-art algorithms,and a healthy dose of old-fashioned low-level engineering. Recent advanceshave enabled ray tracing to be a useful alternative for real-time rendering ofanimated scenes, as well as huge scenes that do not fit into main memory. Inaddition, there is a trend towards the development of dedicated hardware forray tracingAll of this research has helped to push ray tracing from an interesting esoteric technique for image synthesis to a seriously viable algorithm for practicalapplications. If necessary, ray tracing can operate in real time. If desired,raytracing can be physically based and can therefore be used in predictive light-ing simulations. As a result, ray tracing is now used in earnest in the movieindustry, but also, for instance, in the automotive industry and in scientificvisualization. In addition, it forms the basis for several other graphics algorithms, including radiosity and photon mappingThe practical importance of ray tracing as a lighting- simulation techniquemeans that ray tracing needs to be taught to students, as well as to practitionersin industry. In addition, ray tracing is sufficiently multifaceted that teachingstudents all aspects of the algorithm will give them all manner of additionalbenefits: 3D modeling skills, mathematics skills, software engineering skills(writing a ray tracer is for many students the first time that they will have tomanage a sizeable chunk of code), hands-on experience in object-oriented programming, and deeper insights into the physics of light, as well as knowledgeof the behavior of materialsIt would be ideal to present a ray-tracing course to students at the under-graduate level for all of the above reasons, but also because a deep understanding of ray tracing will make it easier to grasp other image-synthesis algorithmsForewordXIFor this, a book is required that explains all facets of ray tracing at the rightpace, assuming only a very moderate amount of background knowledgeI' m positively delighted that such a book now exists. Ray Tracing from theGround up not only covers all aspects of ray tracing, but does so at a level thatallows both undergraduate and graduate students to appreciate the beautyand algorithmic elegance of ray tracing. At the same time, this book goes intomore than sufficient detail to deserve a place on the bookshelves of many professionals as a reference workKevin was gracious enough to let me read early drafts of several chapterswhen I was teaching a graduate-level ray-tracing course at the University ofCentral florida. This has certainly taught me many of the lesser-known intricacies of ray tracing. Kevin himself has taught ray tracing to undergraduatestudents for many years, and it shows. This book, which grew out of his coursenotes, is remarkably easy to follow, especially given the complexity of the subject matterAs such i can heartily recommend this book to both professionals aswell as students and teachers. Whether you are only interested in rendering acollection of shiny spheres or want to create stunning images of highly complicated and realistic environments, this book will show you how. Whetherits intended use is as a ray-tracing reference or as the basis of a course on raytracing this book is essential readinErik reinhardUniversity of BristolUniversity of Central Florida