CS 348B - Computer Graphics: Image Synthesis Techniques

Description

This course provides a broad overview of the theory and practice of rendering. Classic rendering algorithms will be covered, however, most of the course will cover current results in physically-based rendering algorithms.

Prerequisites

Attendees should have a working knowledge of computer graphics (at the level of CS248 and CS348A). In particular, you should fully understand the basic 3D display pipeline, viewing and modeling transformations, simple geometric modeling using polygons and quadrics, and hidden surface algorithms like the z-buffer algorithm.

Solid knowledge of integral calculus and basic geometric algorithms is an absolute must. Some exposure to signal processing and probability is also assumed.

Information

Personnel

Instructor

Pat Hanrahan
hanrahan@cs.stanford.edu
Rm 370 Gates Computer Science Building
(650) 723-8530
Office hours: 11:00-12:00noon Thu

Teaching assistant

Greg Humphreys
humper@graphics.stanford.edu
Rm. 381 Gates Computer Science Building
(650) 723-0618
Office hours: 2:00-4:00 Tuesday and Thursday

Secretary

Ada Glucksman
ada@graphics.stanford.edu
Room 368 Gates Computer Science Building
(next to Pat Hanrahan's office)
(650) 725-3724
Office hours: Mon-Fri, 9:00 to 4:30

Texts and Readings

Required

Andrew Glassner,
An Introduction to Ray Tracing,
Academic Press.

This book contains a collection of chapters written by many of the original inventors of major ray tracing algorithms. It is filled with both practical and theoretical information.

Matt Pharr, Greg Humphreys, Pat Hanrahan,
A Literate Ray Tracer,
Unpublished manuscript.

A work in progress. Chapters will be made available from the class directories.

Optional texts

Michael Cohen and John Wallace,
Radiosity and Realistic Image Synthesis,
Academic Press.
James Foley, Andries van Dam, Steven Feiner, and John Hughes,
Computer Graphics: Principles and Practice,
second edition in C, Addison-Wesley.
Andrew Glassner,
Principles of Digital Image Synthesis,
Morgan Kaufmann, 1995.
Peter Shirley,
Realistic Ray Tracing,
A. K. Peters, 2001.
Francois Sillion, Claude Puech,
Radiosity and Global Illumination,
Morgan Kaufmann, 1994.

The textbook by Foley, van Dam, Feiner, and Hughes is the most popular introductory text for computer graphics. CS248 covers roughly the first half of the book. In this course, we will cover material contained in chapters 13 through 20.

The ray tracing book by Peter Shirley is highly recommended. It covers much of the material in the class at an introductory level.

The other textbooks are good additional resources that provide more depth than the required text. The book by Cohen and Wallace and by Sillion and Puech are good introductions to physically-based rendering, although they emphasize radiosity. The book by Glassner is an encyclopedic overview of rendering.

Additional texts

Anthony Apodaca and Larry Gritz,
Advanced Renderman: Creating CGI for the Motion Pictures,
Morgan Kaufmann, 1999.
David Ebert, F. Kenton Musgrave, Darwyn Peachey, Steven Worley, Ken Perlin,
Texturing and Modeling,
second edition, Morgan Kaufmann, 1998.
Steve Upstill,
The RenderMan Companion: A Programmers Guide to Realistic Computer Graphics,
Addison-Wesley, 1989.

The book by Ebert et al. is a excellent overview of procedural modeling and texturing. The book by Upstill is the standard reference on the RenderMan interface; the other book on RenderMan by Apadoca and Gritz is the best current overview of advanced rendering from the user's point of view.

All books are on reserve in the Math and Computer Science Library.

Additional readings

Additional readings will be assigned from research monographs, papers from journals and conference proceedings, and excerpts from conference tutorials. Only papers NOT available online will be handed out in class.

Topics and Schedule


Assignments and Grading

The project for this quarter is to build a working a ray tracer. To help focus your efforts on the most interesting parts, we have built a ray tracing systems called lrt (for literate ray tracer). lrt is is written using noweb, a system originally developed by Don Knuth and reimplemented and improved by Norman Ramsey for describing large software systems. lrt is a very modular rendering system and provides an architectural framework for exploring different rendering algorithms. reads models that use PIXAR's RenderMan Interface.

The first part of the course involves three progamming assignments:

Stage 1: Implement ray-surface intersection methods for terrains.
Stage 2: Implement a camera model capable of generating depth-of-field.
Stage 3: Use Monte Carlo techniques to compute reflections from glossy surfaces illuminated by area light source.
Finally, in the last part of the course you will enhance your system so that it is capable of reproducing an image of a real object, for example, a gemstone, a puff of smoke, a candle flame, etc. Check out the results produced by previous students.

Evaluation criteria: The first three programming assignments will each count as 10% of your grade, and the last programming project will count as 50%. There will also be two written homeworks that each will count 10% of your final grade. There will be no exams.

Collaboration: For the first three programming and two written assignments, you may discuss the assignment with friends, but you are expected to implement your own solutions. On the last programming project, you are permitted (and encouraged) to form teams of two people and partition your planned extensions among the team members. Teams may discuss their project with other teams, but may not share code.

Late assignments: Since each assignment builds on the previous one, it is important that assignments be completed on time. To allow for unforeseeable circumstances, you will be allowed three weekdays of grace during the quarter. Beyond this, late assignments will be penalized by 10% per weekday that they are late. On the last programming project, neither the demo nor the writeup may be late. Incompletes in this course are given only in exceptional circumstances.

Hardware and software

Hardware: To do the assignments you will have access to the Stanford Graphics Laboratory. The lab is located in two rooms in the basement of Sweet Hall. One room contains 15 Silicon Graphics Octanes (175 Mhz R10000, 400K polygons/sec, 128 MB main memory); The SGI machines are named raptor1 - raptor15 and run Irix 6.5. The other room contains 15 PCs (Dell Precision 420, 730 Mhz PIII, 256MB) with NVIDIA Quadro Graphics Cards. The PCs are named firebird1 - firebird15 and run RedHat Linux 6.2. Use ssh for remote access. Students in CS 348B have non-exclusive priority access to this laboratory, whose door is opened using your Stanford ID.

All students with leland accounts automatically have accounts on these machines. Home directories on these machines are shared with other Stanford Computing Clusters using AFS. If you do not have a leland account, follow these instructions. Registered students will get an extra 60MB of disk quota and 200 pages of print quota for the quarter.

Using other machines: Our software will be installed on the Stanford Graphics Laboratory machines in Sweet Hall. We may also provide one or more 3D modeling programs. If you prefer working in your dorm room or workplace, and have access to a machine there, you are welcome to do the assignments on your own machines. We will make as many of these tools available via ftp as we are permitted by the terms of our licensing agreements, but we will not support them on any other platform. Moreover, your code must run on either the SGI or PCs in Sweet Hall. In particular, we will grade your projects by running your programs on an SGI using our own test data files.

Rendering competition

In case the delight of learning does not sufficiently motivate you to exert yourselves heroically on the programming assignments, there will also be a rendering competition. At 3:00pm on Friday, June 8, a judging will be held to select the best rendering made using the ray tracer you have written in the course. While grades for the projects are based solely on "technical merit", the competition will be judged on both "technical merit" and "artistic impression". The jury, to be named later, will consist of computer graphics experts from both industry and academia. There will be several awards and one grand prize - an all-expenses-paid trip to Siggraph 2001 in Los Angeles. Total value: about $1,000.

Copyright © 2001 Pat Hanrahan