| Description |
1 online resource (xiv, 270 pages) : illustrations |
| Series |
Cognitive systems monographs ; v. 4 |
|
Cognitive systems monographs ; v. 4.
|
| Contents |
Cover -- Contents -- 1 Introduction -- 1.1 Motivation and Scope -- 1.2 Contribution -- 1.3 Outline -- 2 State of the Art in Object Recognition and Pose Estimation -- 2.1 Appearance-Based Methods -- 2.1.1 Global Approaches -- 2.1.2 Local Approaches -- 2.2 Model-Based Methods -- 2.2.1 Edge-Based Object Tracking -- 2.2.2 Edge-Based Object Recognition -- 2.2.3 Pose Estimation Based on Matched Feature Points -- 2.2.4 Object Recognition Based on 3D Point Clouds -- 2.2.5 Hybrid Approaches -- 2.3 Comparison -- 3 State of the Art in Human Motion Capture -- 3.1 VICON -- 3.2 Systems Using a Search Method -- 3.3 Systems Using a Minimization Method -- 3.3.1 Minimization Method for Articulated Objects -- 3.3.2 Systems Using a 3D-3D Minimization Method -- 3.3.3 Systems Using a 2D-3D Minimization Method -- 3.4 Systems Based on Particle Filtering -- 3.5 Pose Estimation Based on Silhouettes -- 3.6 Comparison -- 4 Fundamentals of Image Processing -- 4.1 Camera Model -- 4.1.1 Coordinate Systems -- 4.1.2 Intrinsic Camera Parameters of the Linear Mapping -- 4.1.3 Extrinsic Camera Parameters -- 4.1.4 Distortion Parameters -- 4.1.5 Overview -- 4.2 Segmentation -- 4.2.1 Thresholding -- 4.2.2 Background Subtraction -- 4.2.3 Color Segmentation -- 4.3 Correlation Methods -- 4.3.1 General Definition -- 4.3.2 Non-normalized Correlation Functions -- 4.3.3 Normalized Correlation Functions -- 4.4 Homography -- 4.4.1 General Definition -- 4.4.2 Least Squares Computation of Homography Parameters -- 4.5 Principal Component Analysis -- 4.5.1 Mathematical Definition -- 4.5.2 Eigenspace -- 4.5.3 Application -- 4.6 Particle Filtering -- 4.7 RANSAC -- 5 Guiding Principles -- 5.1 Ways of Using Calibrated Stereo Camera Systems -- 5.2 Eye Movements -- 5.3 Rectification of Stereo Image Pairs -- 5.4 Undistortion of Images -- 5.5 Grasp Execution -- 5.6 Imitiation Learning -- 6 Stereo-Based Object Recognition and Pose Estimation System -- 6.1 Recognition and Pose Estimation Based on the Shape -- 6.1.1 Problem Definition -- 6.1.2 Basic Approach -- 6.1.3 Orientation Correction -- 6.1.4 Position Correction -- 6.1.5 Increasing Robustness -- 6.1.6 Summary of the Algorithm -- 6.2 Recognition and Pose Estimation Based on Texture -- 6.2.1 Accuracy Considerations -- 6.2.2 Feature Calculation -- 6.2.3 Recognition and 2D Localization -- 6.2.4 6-DoF Pose Estimation -- 6.2.5 Occlusions -- 6.2.6 Increasing Robustness -- 6.2.7 Runtime Considerations -- 6.2.8 Summary of the Algorithm -- 7 Stereo-Based Markerless Human Motion Capture System -- 7.1 Problem Definition -- 7.2 Human Upper Body Model -- 7.2.1 Kinematic Model -- 7.2.2 Geometric Model -- 7.3 General Particle Filtering Framework for Human Motion Capture -- 7.3.1 Edge Cue -- 7.3.2 Region Cue -- 7.3.3 Fusion of Multiple Cues -- 7.4 Cues in the Proposed System -- 7.4.1 Edge Cue -- 7.4.2 Distance Cue -- 7.4.3 Cue Comparison -- 7.4.4 Using a Calibrated Stereo System -- 7.5 Image Processing Pipeline -- 7.6 Hand/Head Tracking -- 7.7 Hierarchical Search -- 7.8 Fusing the Edge Cue and the Distance Cue -- 7.9 Adaptive Noise -- 7.10 Adaptive Shoulder Position -- 7.11 Incorporating Inverse Kinematics -- 7.12 Summary of the Algorithm -- 8 Software and Interfaces -- T$24593 |
| Summary |
The development of humanoid robots is one of the most challenging research fields within robotics. One of the crucial capabilities of such a humanoid is the ability to visually perceive its environment. The present monograph deals with visual perception for the intended applications manipulation and imitation, supporting higher-level cognition. In particular, stereo-based methods and systems for object recognition and 6 DoF pose estimation as well as for markerless human motion capture are presented. After an extensive presentation of the state of the art in these areas, three real-time systems that have been developed by the author are presented in great detail: object recognition and pose estimation for textured and for single-colored objects, and a markerless human motion capture system. As only sensor a stereo camera system is used. All experiments have been performed using the humanoid robot ARMAR-III. The systems presented in this monograph are successfully applied for various research activities in the context of humanoid robotics at the University of Karlsruhe, including manipulation, imitation, visual serving, motion planning, and higher-level planning |
| Bibliography |
Includes bibliographical references (pages 255-265) and index |
| Notes |
English |
|
Print version record |
| In |
Springer eBooks |
| Subject |
Robot vision.
|
|
Robots -- Control systems.
|
|
Engineering.
|
|
Artificial intelligence.
|
|
engineering.
|
|
artificial intelligence.
|
|
Ingénierie.
|
|
Robot vision
|
|
Robots -- Control systems
|
|
Humanoider Roboter
|
|
Lokalisierung Robotik
|
|
Maschinelles Lernen
|
|
Motion Capturing
|
|
Objekterkennung
|
|
Pose
|
| Form |
Electronic book
|
| LC no. |
2009934346 |
| ISBN |
9783642042294 |
|
3642042295 |
|
