Three-Dimensional Machine Vision PDF
Preview Three-Dimensional Machine Vision
THREE-DIMENSIONAL MACHINE VISION THE KLUWER INTERNATIONAL SERIES IN ENGINEERING AND COMPUTER SCIENCE ROBOTICS: VISION, MANIPULATION AND SENSORS Consulting Editor Takeo Kanade Other books in the series: Robotic Grasping andFine Manipulation. M. Cutkosky. ISBN 0-89838-200-9. Shadowsand Silhouettes in Computer Vision. S. Shafer. ISBN 0-89838-167-3. PerceptualOrganization and Visual Recognition. D. Lowe. ISBN 0-89838-172-X. THREE-DIMENSIONAL MACHINE VISION edited by Takeo Kanade Carnegie Mellon University ~ " KLUWER ACADEMIC PUBLISHERS Boston/Dordrecht/Lancaster Distributors for North America: Kluwer Academic Publishers 101 Philip Drive Assinippi Park Norwell, MA02061, USA Distributorsfor the UK and Ireland: Kluwer Academic Publishers MTP Press Limited Falcon House, Queen Square Lancaster, LAI IRN, UNITED KINGDOM Distributorsfor all othercountries: Kluwer Academic Publishers Group Distribution Centre Post Office Box 322 3300AH Dordrecht, THE NETHERLANDS Coverphotographby BradleyA. Hersch, Monroeville, Pennsylvania. LibraryofCongressCataloging-in-Publication Data Three-dimensionalmachine vision. (The Kluwer international seriesin engineeringand computerscience. Robotics) I. Robot vision. I. Kanade, Takeo. TJ211.3.T47 1987 629.8'92 86-27599 ISBN-13:978-1-4612-9184-8 e-ISBN-13:978-1-4613-1981-8 DOl: 10.1007/978-1-4613-1981-8 Copyright © 1987 by Kluwer Academic Publishers Softcover reprintof the hardcover 1st edition 1987 All rightsreserved. Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,or transmitted in any form or by any means, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher, Kluwer Academic Publishers, 101 Philip Drive, Assinippi Park, Norwell, MA02061. Table of Contents Preface vii Takeo Kanade, Editor PART I: 3-D SENSORS A Three-Dimensional Sensor Based on Structured Light 3 J. L. Mundy and G. B. Porter III 3-D Vision System Analysis and Design 63 Steven K. Case, Jeffrey A. Jalkio) and Richard C. Kim Robot Vision by Encoded Light Beams 97 Martin D. Altschuler) Kyongtae Bae) Bruce R. Altschuler) Jerome T. Dijak, Louis A. Tamburino, and Barbara Woolford A Noncontact Optical Proximity Sensor for Measuring Surface Shape 151 Takeo Kanade and Michael Fuhrman PART II: 3-D FEATURE EXTRACTIONS Toward a Surface Primal Sketch 195 Jean Ponce and Michael Brady 3-D Object Representation from Range Data Using Intrinsic Surface Properties 241 B. C. Vemun) A. Mitiche) and J. K. Aggarwal Use of Vertex-Type Knowledge for Range Data Analysis 267 Kokichi Sugihara PART ill: 3-D RECOGNITION ALGORITHMS The Representation, Recognition, and Positioning of 3-D Shapes from Range Data 301 O. D. Faugeras and M. Hebert vi CONTENTS An Object Recognition System Using Three-Dimensional Information 355 Masaki Oshima and Yoshiaki Shirai 3DPO: A Three-Dimensional Part Orientation System 399 Robert C. Bolles and Patrice Horaud Recognition and Localization of Overlapping Parts From Sparse Data 451 W. Eric L. Grimson and Tomas Lozano-Perez PART IV: SYSTEMS AND APPLICATIONS Producing Space Shuttle Tiles with a 3-D Non-Contact Measurement System 513 T. E. Beeler Three-Dimensional Vision Systems Using the Structured-Light Method for Inspecting Solder Joints and Assembly Robots 543 Yasuo Nakagawa and Takanori Ninomiya A Semantic-Free Approach to 3-D Robot Color Vision 565 R. A. Jarvis Preface A robot must perceive the three-dimensional world ifit is to be effective there. Yet recovering 3-D information from projected images is difficult, and still remains the subject of basic research. Alternatively, one can use sensors that can provide three-dimensional range information directly. The technique of projecting light-stripes started to be used in industrial object recognition systems as early as the 1970s, and time-of-flight laser-scanning range finders became available for outdoor mobile robot navigation in the mid-eighties. Once range data are obtained, a vision system must still describe the scene in terms of 3-D primitives such as edges, surfaces, and volumes, and recognize objeCts of interest. Today, the art of sensing, extracting features, and recognizing objects by means of three-dimensional range data is one ofthe mostexciting research areas in computervision. Three-Dimensional Machine Vision is a collection of papers dealing with three-dimensional range data. The authors are pioneering researchers: some are founders and others are bringing new excitements in the field. I have tried to select milestone papers, and my goal has been to make this book a reference workfor researchers in three-dimensionalvision. The book is organized into four parts: 3-D Sensors, 3-D Feature Extractions, Object Recognition Algorithms, and Systems and Applications. Part I includes four papers which describe the development of unique, capable 3-D range sensors, as well as discussions of optical, geometrical, electronic, and computational issues. Mundy and Porter describe a sensor system based on structured illumination for inspecting metallic castings. In order to achieve high-speed data acquisition, it uses multiple light stripes with wavelength multiplexing. Case, Jalkio, and Kim also present a multi stripe system and discuss various design issues in range sensing by triangulation. The numerical stereo camera developed by Altschuler, Bae, Altschuler, Dijak, Tamburino, and Woolford projects space-coded grid patterns which are generated by an electro-optical programmable spatial viii PREFACE light modulator. Kanade and Fuhrman present a proximity sensor using multiple LEDs which are conically arranged. It can measure both distance and orientation ofan object's surface. Having acquired range data, the next step is to analyze it and extract three-dimensional features. In Part II, Ponce and Bradypresent the Surface Primal Sketch. They detect, localize, and symbolically describe the significant surface changes: steps, roofs, joints, and shoulders. Vemuri, Mitiche, and Aggarwal represent 3-D object surfaces by patches which are homogeneous in intrinsic surface properties. Sugihara describes how the knowledge ofvertex types in polyhedra can be used to guide the extraction ofedges and vertices from light-stripe range data. An important goal of 3-D vision systems is to recognize and position objects in the scene. Part III deals with algorithms for this purpose. Mter describing object surfaces by curves and patches, Faugeras and Hebert use the paradigm of recognizing while positioning for object matching by exploiting rigidity constraints. Oshima and Shirai have long worked on a systemfor recognizing stacked objectswith planar and curved surfaces using range data. Their paper presents procedures for feature extraction and matching together with recent experimental results. The matching strategy used by Bolles and Horaud to locate parts in a jumble starts with a distinctive edge feature of an object and grows the match by adding compatible features. Grimson and Lozano-Perez discuss how sparse measurement of positions and surface normals may be used to identify and localize overlapping objects. Their search procedure efficiently discards inconsistent hypothetical matches between sensed data and object model usinglocal constraints. Finally, PartIVcontainsthree papers discussing applications and system issues of three-dimensional vision systems which use range sensing. Beeler presents a precision 3-D measurement system for replacing damaged or missing tiles on space shuttle vehicles. The acquired shape description of a PREFACE ix tile cavity 1S used to control a milling machine to produce a new tile. Nakagawa and Ninomiya describe two three-dimensional vision systems that use active lighting: one is for inspecting solder joints and the other for automatic assembly ofelectronic devices. The paper by Jarvis, arguingfor a semantic-free approach to three-dimensional robotic vision, discusses various system issues in a real-time sensory-based robotic system, such as a hand-eyesystem. In editing the book I have received great help from Richard Mickelsen and Magdalena M~ler. Richard read all of the papers and brought consistency into the book. Maggie spent untiring effort for typing and perfectingtheformat. Iam heartily grateful: without them, this bookwould nothave been completed. Takeo Kanade PART I: 3-D SENSORS
The list of books you might like
