Wöhler 3D Computer Vision

Efficient Methods and Applications
1. Auflage 2009
ISBN: 978-3-642-01732-2
Verlag: Springer
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 385 Seiten

This work provides an introduction to the foundations of three-dimensional c- puter vision and describes recent contributions to the ?eld, which are of methodical and application-speci?c nature. Each chapter of this work provides an extensive overview of the corresponding state of the art, into which a detailed description of new methods or evaluation results in application-speci?c systems is embedded. Geometric approaches to three-dimensional scene reconstruction (cf. Chapter 1) are primarily based on the concept of bundle adjustment, which has been developed more than 100 years ago in the domain of photogrammetry. The three-dimensional scene structure and the intrinsic and extrinsic camera parameters are determined such that the Euclidean backprojection error in the image plane is minimised, u- ally relying on a nonlinear optimisation procedure. In the ?eld of computer vision, an alternative framework based on projective geometry has emerged during the last two decades, which allows to use linear algebra techniques for three-dimensional scene reconstructionand camera calibration purposes. With special emphasis on the problems of stereo image analysis and camera calibration, these fairly different - proaches are related to each other in the presented work, and their advantages and drawbacks are stated. In this context, various state-of-the-artcamera calibration and self-calibration methods as well as recent contributions towards automated camera calibration systems are described. An overview of classical and new feature-based, correlation-based, dense, and spatio-temporal methods for establishing point c- respondences between pairs of stereo images is given.

Christian Wöhler received his Diploma in Physics from Würzburg University in 1996, the Doctorate degree in Computer Science from Bonn University in 2000, and the Habilitation in Applied Computer Science from Bielefeld University in 2009. He is working as a research scientist in the Environment Perception department of Daimler Group Research and Advanced Engineering in Ulm. Since 2005 he is a visiting lecturer at the Technical Faculty of Bielefeld University. His scientific interests are in the domain of pattern classification, three-dimensional computer vision, and photogrammetry, with applications in the fields of driver assistance systems, industrial machine vision, and planetary science.'

Autoren/Hrsg.

Wöhler, Christian

Inhaltsverzeichnis

1;Preface;6
2;Acknowledgements;9
3;Contents;10
4;Part I Methods of 3D Computer Vision;14
4.1;Chapter 1 Geometric Approaches to Three-dimensional Scene Reconstruction;15
4.1.1;1.1 The Pinhole Camera Model;15
4.1.2;1.2 Bundle Adjustment Methods;19
4.1.3;1.3 Geometric Aspects of Stereo Image Analysis;21
4.1.3.1;1.3.1 Euclidean Formulation of Stereo Image Analysis;21
4.1.3.2;1.3.2 Stereo Image Analysis in Terms of Projective Geometry;24
4.1.4;1.4 Geometric Calibration of Single and Multiple Cameras;29
4.1.4.1;1.4.1 Methods for Intrinsic Camera Calibration;29
4.1.4.2;1.4.2 The Direct Linear Transform (DLT) Method;30
4.1.4.3;1.4.3 The Camera Calibration Method by Tsai (1987);33
4.1.4.4;1.4.4 The Camera Calibration Method by Zhang (1999a);37
4.1.4.5;1.4.5 The Camera Calibration Method by Bouguet (2007);39
4.1.4.6;1.4.6 Self-calibration of Camera Systems from Multiple Views of a Static Scene;40
4.1.4.7;1.4.7 Semi-automatic Calibration of Multiocular Camera Systems;53
4.1.4.8;1.4.8 Accurate Localisation of Chequerboard Corners;63
4.1.5;1.5 Stereo Image Analysis in Standard Geometry;74
4.1.5.1;1.5.1 Image Rectification According to Standard Geometry;74
4.1.5.2;1.5.2 The Determination of Corresponding Points;78
4.1.6;1.6 Three-dimensional Pose Estimation and Segmentation Methods;99
4.1.6.1;1.6.1 Pose Estimation of Rigid Objects;100
4.1.6.2;1.6.2 Pose Estimation of Non-rigid and Articulated Objects;107
4.1.6.3;1.6.3 Point Cloud Segmentation Approaches;125
4.2;Chapter 2 Photometric Approaches to Three-dimensional Scene Reconstruction;139
4.2.1;2.1 Shape from Shadow;139
4.2.1.1;2.1.1 Extraction of Shadows from Image Pairs;140
4.2.1.2;2.1.2 Shadow-based Surface Reconstructionfrom Dense Sets of Images;142
4.2.2;2.2 Shape from Shading;144
4.2.2.1;2.2.1 The Bidirectional Reflectance Distribution Function (BRDF);144
4.2.2.2;2.2.2 Determination of Surface Gradients;149
4.2.2.3;2.2.3 Reconstruction of Height from Gradients;154
4.2.2.4;2.2.4 Surface Reconstruction Based on Eikonal Equations;156
4.2.3;2.3 Photometric Stereo;158
4.2.3.1;2.3.1 Classical Photometric Stereo Approaches;159
4.2.3.2;2.3.2 Photometric Stereo Approaches Based on Ratio Images;160
4.2.4;2.4 Shape from Polarisation;163
4.2.4.1;2.4.1 Surface Orientation from Dielectric Polarisation Models;163
4.2.4.2;2.4.2 Determination of Polarimetric Properties of Rough Metallic Surfaces for Three-dimensional Reconstruction Purposes;166
4.3;Chapter 3 Real-aperture Approaches to Three-dimensional Scene Reconstruction;170
4.3.1;3.1 Depth from Focus;172
4.3.2;3.2 Depth from Defocus;173
4.3.2.1;3.2.1 Basic Principles;173
4.3.2.2;3.2.2 Determination of Small Depth Differences;178
4.3.2.3;3.2.3 Determination of Absolute Depth Across Broad Ranges;181
4.4;Chapter 4 Integrated Frameworks for Three-dimensional Scene Reconstruction;191
4.4.1;4.1 Monocular Three-dimensional Scene Reconstruction at Absolute Scale;192
4.4.1.1;4.1.1 Combining Motion, Structure, and Defocus;193
4.4.1.2;4.1.2 Online Version of the Algorithm;194
4.4.1.3;4.1.3 Experimental Evaluation Based on Tabletop Scenes;195
4.4.1.4;4.1.4 Discussion;205
4.4.2;4.2 Self-consistent Combination of Shadow and Shading Features;206
4.4.2.1;4.2.1 Selection of a Shape from Shading Solution Based on Shadow Analysis;207
4.4.2.2;4.2.2 Accounting for the Detailed Shadow Structure in the Shape from Shading Formalism;210
4.4.2.3;4.2.3 Initialisation of the Shape from Shading Algorithm Based on Shadow Analysis;212
4.4.2.4;4.2.4 Experimental Evaluation Based on Synthetic Data;214
4.4.2.5;4.2.5 Discussion;215
4.4.3;4.3 Shape from Photopolarimetric Reflectance and Depth;216
4.4.3.1;4.3.1 Shape from Photopolarimetric Reflectance;217
4.4.3.2;4.3.2 Estimation of the Surface Albedo;221
4.4.3.3;4.3.3 Integration of Depth Information;222
4.4.3.4;4.3.4 Experimental Evaluation Based on Synthetic Data;227
4.4.3.5;4.3.5 Discussion;232
4.4.4;4.4 Stereo Image Analysis of Non-Lambertian Surfaces;233
4.4.4.1;4.4.1 Iterative Scheme for Disparity Estimation;235
4.4.4.2;4.4.2 Qualitative Behaviour of the Specular Stereo Algorithm;239
4.4.5;4.5 Three-dimensional Pose Estimation Based on Combinations of Monocular Cues;240
4.4.5.1;4.5.1 Appearance-based Pose EstimationRelying on Multiple Monocular Cues;241
4.4.5.2;4.5.2 Contour-based Pose Estimation Using Depth from Defocus;246
5;Part II Application Scenarios;251
5.1;Chapter 5 Applications to Industrial Quality Inspection;252
5.1.1;5.1 Inspection of Rigid Parts;253
5.1.1.1;5.1.1 Object Detection by Pose Estimation;253
5.1.1.2;5.1.2 Pose Refinement;257
5.1.2;5.2 Inspection of Non-rigid Parts;262
5.1.3;5.3 Inspection of Metallic Surfaces;265
5.1.3.1;5.3.1 Inspection Based on Integration of Shadow and Shading Features;265
5.1.3.2;5.3.2 Inspection of Surfaces with Non-uniform Albedo;266
5.1.3.3;5.3.3 Inspection Based on SfPR and SfPRD;268
5.1.3.4;5.3.4 Inspection Based on Specular Stereo;275
5.1.3.5;5.3.5 Discussion;282
5.2;Chapter 6 Applications to Safe Human–Robot Interaction;285
5.2.1;6.1 Vision-based Human-Robot Interaction;285
5.2.1.1;6.1.1 The Role of Gestures in Human–Robot Interaction;286
5.2.1.2;6.1.2 Safe Human–Robot Interaction;287
5.2.1.3;6.1.3 Pose Estimation of Articulated Objects in the Context of Human–Robot Interaction;290
5.2.2;6.2 Object Detection and Tracking in Three-dimensional Point Clouds;299
5.2.3;6.3 Detection and Spatio-temporal Pose Estimation of Human Body Parts;301
5.2.4;6.4 Three-dimensional Tracking of Human Body Parts;304
5.3;Chapter 7 Applications to Lunar Remote Sensing;310
5.3.1;7.1 Three-dimensional Surface Reconstruction Methods for Planetary Remote Sensing;311
5.3.1.1;7.1.1 Topographic Mapping of Solar System Bodies;311
5.3.1.2;7.1.2 Reflectance Behaviour of Planetary Regolith Surfaces;314
5.3.2;7.2 Three-dimensional Reconstruction of Lunar Impact Craters;318
5.3.2.1;7.2.1 Shadow-based Measurement of Crater Depth;318
5.3.2.2;7.2.2 Three-dimensional Reconstruction of Lunar Impact Craters at High Resolution;321
5.3.3;7.3 Three-dimensional Reconstruction of Lunar Wrinkle Ridges and Faults;329
5.3.4;7.4 Three-dimensional Reconstruction of Lunar Domes;332
5.3.4.1;7.4.1 General Overview of Lunar Mare Domes;332
5.3.4.2;7.4.2 Observations of Lunar Mare Domes;335
5.3.4.3;7.4.3 Image-based Determination of Morphometric Data;338
5.3.4.4;7.4.4 Geophysical Insights Gained from Topographic Data;350
5.4;Chapter 8 Conclusion;358
6;References;365