Suchen und Finden
Preface
4
Contents
8
Contributors
10
Programming-by-Demonstration of Robot Motions
13
Introduction
13
Learning from Human Demonstration
15
Interpretation of Demonstrations in Hand-State Space
15
Skill Encoding Using Fuzzy Modeling
16
Generation and Execution of Robotic Trajectories Based on Human Demonstration
18
Mapping Between Human and Robot Hand States
19
Definition of Hand-States for Specific Robot Hands
20
Next-State-Planners for Trajectory Generation
22
Demonstrations of Pick-and-Place Tasks
24
Variance from Multiple Demonstrations
24
Experimental Platform
24
Experimental Evaluation
26
Experiment 1: Learning from Demonstration
26
Importance of the Demonstration
27
Experiment 2: Generalization in Workspace
29
Experiment 3: a Complete Pick-and-Place Task
32
Conclusions and Future Work
32
References
34
Grasp Recognition by Fuzzy Modeling and Hidden Markov Models
36
Introduction
36
An Experimental Platform for PBD
37
Simulation of Grasp Primitives
39
Geometrical Modeling
39
Modeling of Inverse Kinematics
40
Modeling of Grasp Primitives
42
Modeling by Time-Clustering
42
Training of Time Cluster Models Using New Data
43
Recognition of Grasps-Three Methods
44
Recognition of Grasps Using the Distance Between Fuzzy Clusters
44
Recognition Based on Qualitative Fuzzy Recognition Rules
45
Distance Norms
45
Extrema in the Distance Norms and Segmentation
46
Set of Fuzzy Rules
48
Similarity Degrees
49
Recognition Based on Time-Cluster Models and HMM
50
Experiments and Simulations
53
Time Clustering and Modeling
53
Grasp Segmentation and Recognition
55
Conclusions
57
References
58
Distributed Adaptive Coordinated Control of Multi-Manipulator Systems Using Neural Networks
59
Introduction
59
Preliminaries
61
Multi-Manipulator System Description
61
Radial Basis Function Neural Network
63
Controller Design
65
Performance Analysis
67
Simulation Example
71
Conclusion
75
References
78
A New Framework for View-Invariant Human Action Recognition
80
Introduction
80
Overview of the Proposed Approach
85
Exemplar Selection and Representation
87
Key Pose Extraction
87
2D Silhouette Image Generation
88
Contour Shape Feature
89
Action Modelling and Recognition
91
Exemplar-based Hidden Markov Model
91
Action Modelling
92
Action Recognition
92
Action Category Revalidation
93
Experiments
95
Conclusion
99
References
100
Using Fuzzy Gaussian Inference and Genetic Programming to Classify 3D Human Motions
103
Introduction
103
Human Skeletal Representation
104
The Learning Method
106
Model Description: the Fuzzy Membership Function
106
Model Generation: Fuzzy Gaussian Inference
107
Membership Evaluation
109
Mathematical Properties
109
Extracting Fuzzy Rules Using Genetic Programming
112
Experiment and Results
113
Apparatus
113
Participants
113
Procedure
115
Results
118
Discussion
121
Conclusion
122
References
122
Obstacle Detection Using Cross-Ratio and Disparity Velocity
125
Introduction
125
Background
125
Algorithm Overview
126
Generation of Mesh Maps
128
Mesh Generation
128
Estimation of the Ground Floor
130
Identification of Safe Regions within the Ground Plane
134
Incremental Addition of Feature Points
134
Safe Path Detection
137
Further Evaluation
141
Estimation of Ground Floor
141
Obstacle Detection
144
Summary
145
References
148
Learning and Vision-Based Obstacle Avoidance and Navigation
150
Introduction
150
Depth Perception
152
Absolute Depth and Binocular Vision
152
Absolute Depth
152
Relative Depth and Monocular Vision
155
Edge Direction and Perspective
155
Clarity of Detail and Texture Gradient
156
Size Cues
157
Motion Depth Cues
157
Occlusion
158
Why Learning and How to Learn for Monocular Visions
158
The Role of Experience
158
Learning Methods
158
MILN Learning
161
Special Problem: Illumination Changes in Outdoor Scenes
162
Finding Passable Regions for Obstacle Avoidance from Single Image Using MILN
162
Feature Vector
162
Edge
163
Clarity of Detail and Texture Gradient
163
Color Similarity with Lighting Invariance
163
Pixel Position and Region Connection
165
Training Data Generation and Experiment
166
Performance Evaluation
168
Control Law and Navigation
169
From Obstacle Boundaries to Motor Commands
170
Discussion
171
Learning Ability
171
Changing Lighting Conditions
171
Learning from Experience
172
References
172
A Fraction Distortion Model for Accurate Camera Calibration and Correction
175
Introduction
175
Previous Work
176
The Proposed Work
177
A New Distortion Model
178
A Novel Calibration Algorithm
179
Pin-Hole Camera Model
183
Optimisation of All Parameters
184
The Correction of the Distorted Image Points
185
Summary of the Novel Camera Calibration and Correction Algorithm
185
Experimental Results
186
Synthetic Data
186
Calibration and Correction
187
Collinearity Constraint
190
Different Levels of Noise
191
Real Images
192
Conclusion
193
References
195
A Leader-Follower Flocking System Based on Estimated Flocking Center
197
Introduction
197
Flocking System
199
Flocking Algorithms
201
Algorithm Stability
202
Experiments
204
Simulations
211
Conclusions
213
References
213
A Behavior Based Control System for Surveillance UAVs
215
Introduction
215
Platform and Atomic Actions
218
UAV Platform
218
System Structure
219
Atomic Actions
220
Software Architecture and Behavior Development
221
Software Architecture
221
Behavior Development
223
Ground Behavior
223
Takeoff Behavior
223
Hovering Behavior
223
GPS Landing Behavior
224
Vision Landing Behavior
224
Emergency Landing Behavior
224
GPS Tracking Behavior
224
Vision Tracking Behavior
224
Obstacle Avoidance Behavior
225
Trajectory Tracking Behavior
225
Vision Module Development
225
SURF Algorithm
225
Coordination Transformation
226
Kalman Filter
227
Experiment Results
228
Hovering Behavior
228
Vision Tracking Behavior
229
Trajectory Tracking Behavior
230
Trajectory Tracking Behavior with Obstacle Avoidance capability
231
GPS Landing Behavior
231
Vision Landing Behavior
231
Conclusion and Future Work
232
References
233
Hierarchical Composite Anti-Disturbance Control for Robotic Systems Using Robust Disturbance Observer
235
Introduction
235
Formulation of the Problem
237
Hierarchical Composite Anti-Disturbance Control (HCADC)
238
Applications to a Two-Link Robotic System
240
Conclusions
243
Proof of the Lemma 11.1
245
Proof of the Lemma 11.2
247
References
248
Autonomous Navigation for Mobile Robots with Human-Robot Interaction
250
Introduction
250
Human-Robot Interaction
252
Subject Following with Target Pursuing
254
Correspondence
254
Multi-Cue Integration
256
Robust Tracking
257
Pursuing
258
Mapping
259
Qualitative Localization
261
Scene Association
262
Scene Recognition
264
Planning and Navigation
266
Conclusion
269
References
271
Prediction-Based Perceptual System of a Partner Robot for Natural Communication
274
Introduction
274
Prediction-Based Perceptual System for A Partner Robot
276
A Partner Robot; Hubot
276
A Prediction-Based Perceptual System
277
Perceptual Modules
279
Differential Extraction
279
Human Detection
280
Object Detection
281
Hand Motion Recognition
282
Architecture of Prediction Based Perceptual System
282
Input Layer Based on Spiking Neurons
282
Clustering Layer Based on Unsupervised Learning
283
Prediction Layer and Perceptual Module Selection Layer
284
Update of Learning Rate for Perceptual Module Selection
285
Learning for Prediction and Perceptual Module Selection
286
Experimental Results
287
Clustering for Prediction
287
Real-Time Learning in Interaction
290
Additional Learning
291
Conclusions
293
References
294
Index
296
Alle Preise verstehen sich inklusive der gesetzlichen MwSt.