Stereo Scene Flow for 3D Motion Analysis
The accurate and precise estimation of three-dimensional motion vector fields in real time remains one of the key targets for the discipline of computer vision.This important text/reference presents methods for estimating optical flow and scene flow motio
- PDF / 10,704,309 Bytes
- 133 Pages / 439.37 x 666.142 pts Page_size
- 86 Downloads / 214 Views
Andreas Wedel Daniel Cremers
Stereo Scene Flow for 3D Motion Analysis
Dr. Andreas Wedel Group Research Daimler AG HPC 050–G023 Sindelfingen 71059 Germany [email protected]
Prof. Daniel Cremers Department of Computer Science Technical University of Munich Boltzmannstraße 3 Garching 85748 Germany [email protected]
ISBN 978-0-85729-964-2 e-ISBN 978-0-85729-965-9 DOI 10.1007/978-0-85729-965-9 Springer London Dordrecht Heidelberg New York British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Control Number: 2011935628 © Springer-Verlag London Limited 2011 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licenses issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. The use of registered names, trademarks, etc., in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore free for general use. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. Cover design: VTeX UAB, Lithuania Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Preface
The estimation of geometry and motion of the world around us from images is at the heart of Computer Vision. The body of work described in this book arose in the context of video-based analysis of the scene in front of a vehicle from two frontfacing cameras located near the rear view mirror. The question examined of where things are in the world and how they move over time is an essential prerequisite for a higher-level analysis of the observed environment and for subsequent driver assistance. At the origin of this work is the combination of a strong interest in solving the real-world challenges of camera-based driver assistance and a scientific background in energy minimization methods. Yet, the methods we describe for estimating highly accurate optical flow and scene flow are a central prerequisite in other domains of computer vision where accurate and dense point correspondence between images or between geometric structures observed in stereo-videos is of importance. Step by step we introduce variational methods which allow us to enhance the image data acquired from two cameras by spatially dense information on the geometric structure and 3D motion of the observed structures. In particular, we introduce variational approaches to optic flow estimation and presen
Data Loading...
