Active Sensor Planning for Multiview Vision Tasks

by Chen, Shengyong; Li, Youfu; Zhang, Jianwei; Wang, Wanliang

• ISBN13: 9783540770718
• ISBN10: 3540770712
• Format: Hardcover
• Copyright: 2008-04-03
• Publisher: Springer Verlag

Summary

Vision sensors have limited fields of views and can only "see" a portion of a scene from a single viewpoint. To make the entire object visible, the sensor has to be moved from one place to another around the object to observe all features of interest, which brings a multiview vision task that has to be solved by means of active perception. The sensor planning presented in this book describes some effective strategies to generate a sequence of viewing poses and sensor settings for optimally completing a perception task. Several methods are proposed to solve the problems in both model-based and nonmodel-based vision tasks. For model-based applications, the method involves determination of the best next view and sensor settings, to incrementally acquire the object information and to find geometrical cues to predict the unknown portion of an object or environment.

Table of Contents

Preface	p. V
Introduction	p. 1
Motivations	p. 1
The Tasks	p. 1
From a Biological View	p. 3
The Problems and Goals	p. 4
Significance and Applications	p. 6
Objectives and Solutions	p. 7
Book Structure	p. 8
Active Vision Sensors	p. 11
3D Visual Sensing by Machine Vision	p. 11
Passive Visual Sensing	p. 11
Active Visual Sensing	p. 14
3D Sensing by Stereo Vision Sensors	p. 19
Setup with Two Cameras	p. 19
Projection Geometry	p. 20
3D Measurement Principle	p. 21
3D Sensing by Stripe Light Vision Sensors	p. 23
Setup with a Switchable Line Projector	p. 23
Coding Method	p. 24
Measurement Principle	p. 25
3D Sensor Reconfiguration and Recalibration	p. 27
The Motivation for Sensor Reconfiguration and Recalibration	p. 28
Setup of a Reconfigurable System	p. 29
Geometrical Constraint	p. 33
Rectification of Stripe Locations	p. 34
Solution Using the Geometrical Cue	p. 35
Summary	p. 38
Active Sensor Planning - the State-of-the-Art	p. 39
The Problem	p. 39
Overview of the Recent Development	p. 40
Fundamentals of Sensor Modeling and Planning	p. 43
Planning for Dimensional Inspection	p. 48
Planning for Recognition and Search	p. 51
Planning for Exploration, Navigation, and Tracking	p. 54
Planning for Assembly and Disassembly	p. 59
Planning with Illumination	p. 60
Other Planning Tasks	p. 63
Interactive Sensor Planning	p. 63
Placement for Virtual Reality	p. 64
Robot Localization	p. 64
Attention and Gaze	p. 65
Summary	p. 66
Sensing Constraints and Evaluation	p. 67
Representation of Vision Sensors	p. 67
Placement Constraints	p. 68
Visibility	p. 68
Viewing Angle	p. 69
Field of View	p. 69
Resolution	p. 70
In Focus and Viewing Distance	p. 71
Overlap	p. 72
Occlusion	p. 72
Image Contrast	p. 73
Robot Environment Constraints	p. 73
Common Approaches to Viewpoint Evaluation	p. 75
Criterion of Lowest Operation Cost	p. 77
Summary	p. 80
Model-Based Sensor Planning	p. 81
Overview of the Method	p. 81
Sensor Placement Graph	p. 82
HGA Representation	p. 82
Min-Max Objective and Fitness Evaluation	p. 83
Evolutionary Computing	p. 84
The Shortest Path	p. 85
The Viewpoint Distance	p. 85
Determination of a Shortest Path	p. 86
Practical Considerations	p. 87
Geometry Scripts	p. 87
Inspection Features	p. 87
Sensor Structure	p. 88
Constraint Satisfaction	p. 89
Viewpoint Initialization	p. 90
Implementation	p. 92
The Viewpoint Planner	p. 92
Examples of Planning Results	p. 92
Viewpoint Observation	p. 97
Experiments with a Real System	p. 98
Summary	p. 100
Planning for Freeform Surface Measurement	p. 101
The Problem	p. 101
B-Spline Model Representation	p. 104
B-Spline Representation	p. 104
Model Selection	p. 105
Uncertainty Analysis	p. 108
Sensing Strategy for Optimizing Measurement	p. 110
Determining the Number of Measurement Data	p. 110
Optimizing the Locations of Measurement Data	p. 110
Experiments	p. 112
Summary	p. 118
Sensor Planning for Object Modeling	p. 119
Planning Approaches to Model Construction	p. 119
Model Construction from Multiple Views	p. 119
Previous Planning Approaches for Modeling	p. 122
The Procedure for Model Construction	p. 124
Self-Termination Criteria	p. 127
The Principle	p. 127
Termination Judgment	p. 128
Experiments	p. 131
Summary	p. 144
Information Entropy Based Planning	p. 147
Overview	p. 147
Model Representation	p. 148
Curve Approximation	p. 149
Improved BIC Criterion	p. 150
Expected Error	p. 154
Information Entropy of a B-Spline Model	p. 155
Information Gain	p. 156
View Planning	p. 157
Experiments	p. 159
Setup	p. 159
Model Selection	p. 160
Determining the NBV	p. 163
Another Example	p. 172
Summary	p. 175
Model Prediction and Sensor Planning	p. 177
Surface Trend and Target Prediction	p. 177
Surface Trend	p. 177
Determination of the Exploration Direction	p. 179
Surface Prediction	p. 182
Determination of the Next Viewpoint	p. 183
Simulation	p. 186
Practical Considerations	p. 186
Numerical Simulation	p. 187
Practical Implementation	p. 191
Discussion and Conclusion	p. 203
Discussion	p. 203
Conclusion	p. 205
Integrating Planning with Active Illumination	p. 207
Introduction	p. 207
From Human Vision to Machine Vision	p. 209
Evaluation of Illumination Conditions	p. 210
SNR	p. 210
Dynamic Range	p. 210
Linearity	p. 211
Contrast	p. 211
Feature Enhancement	p. 211
Controllable Things	p. 212
Brightness	p. 212
Color Temperature and Color Rendering Index	p. 212
Glare	p. 213
Uniform Intensity	p. 213
Glare Avoidance	p. 214
Disability Glare	p. 214
Discomfort Glare	p. 215
Intensity Estimation	p. 216
Sensor Sensitivity	p. 216
Estimation of Image Irradiance	p. 217
Intensity Control	p. 221
The Setpoint	p. 221
System Design	p. 223
Simulation	p. 224
Implementation	p. 227
Design for Active Illumination	p. 227
Experimental Robots	p. 228
Summary	p. 231
Bibliography	p. 233
A	p. 233
B	p. 234
C	p. 235
D, E	p. 237
F, G	p. 238
H	p. 240
I, J	p. 242
K	p. 243
L	p. 244
M	p. 248
N, O, P, Q	p. 248
R	p. 250
S	p. 251
T	p. 254
U, V, W	p. 255
X, Y, Z	p. 257
Index	p. 261
Table of Contents provided by Ingram. All Rights Reserved.