Foresti / Ellis Ambient Intelligence

A Novel Paradigm
1. Auflage 2006
ISBN: 978-0-387-22991-1
Verlag: Springer US
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 240 Seiten

Ambient Intelligence (AmI) is an integrating technology for supporting a pervasive and transparent infrastructure for implementing smart environments. Such technology is used to enable environments for detecting events and behaviors of people and for responding in a contextually relevant fashion. AmI proposes a multi-disciplinary approach for enhancing human machine interaction. Ambient Intelligence: A Novel Paradigm is a compilation of edited chapters describing current state-of-the-art and new research techniques including those related to intelligent visual monitoring, face and speech recognition, innovative education methods, as well as smart and cognitive environments. The authors start with a description of the iDorm as an example of a smart environment conforming to the AmI paradigm, and introduces computer vision as an important component of the system. Other computer vision examples describe visual monitoring for the elderly, classic and novel surveillance techniques using clusters of cameras installed in indoor and outdoor application domains, and the monitoring of public spaces. Face and speech recognition systems are also covered as well as enhanced LEGO blocks for novel educational purposes. The book closes with a provocative chapter on how a cybernetic system can be designed as the backbone of a human machine interaction.

Autoren/Hrsg.

Foresti, Gian Luca

Ellis, Tim

Inhaltsverzeichnis

1;Contents;7
2;Preface;9
3;Foreword;11
4;1 AMBIENT INTELLIGENCE;14
4.1;1. Introduction;14
4.2;2. The Essex approach;15
4.2.1;2.1 The iDorm - A Testbed for Ubiquitous Computing and Ambient Intelligence;15
4.2.2;2.2 The iDorm Embedded Computational Artifacts;17
4.3;3. Integrating Computer Vision;21
4.3.1;3.1 User Detection;22
4.3.2;3.2 Estimating reliability of detection;24
4.3.3;3.3 Vision in the iDorm;26
4.4;4. Conclusions;26
4.5;References;26
5;2 TOWARDS AMBIENT INTELLIGENCE FOR THE DOMESTIC CARE OF THE ELDERLY;28
5.1;1. Introduction;28
5.2;2. An Integrated Supervision System;29
5.2.1;2.1 E-service Based Integration Schemata;32
5.3;3. People and Robot Localization and Tracking System;34
5.3.1;3.1 System architecture and implementation;36
5.4;4. The Plan Execution Monitoring System;39
5.4.1;4.1 Representing Contingencies;42
5.4.2;4.2 The Execution Monitor;43
5.5;5. Integrating Sensing and Execution Monitoring: a Running Example;46
5.6;6. Conclusions and Future Work;49
5.7;References;51
6;3 SCALING AMBIENT INTELLIGENCE;52
6.1;1. Ambient Intelligence: the contribution of different disciplines;52
6.2;2. I-BLOCKS technology;55
6.3;3. Design process;57
6.4;4. Scaling Ambient Intelligence at level of compositional devices: predefined activities;58
6.4.1;4.1 Arithmetic training;59
6.4.2;4.2 Storytelling Play Scenario;60
6.4.3;4.3 Linguistic scenario;63
6.5;5. Scaling Ambient Intelligence at level of compositional devices: free activities;65
6.6;6. Scaling Ambient Intelligence at the level of configurable environments: future scenarios;67
6.6.1;6.1 The Augmented Playground;67
6.6.2;6.2 Self-reconfigurable Robots;70
6.7;7. Discussion and conclusions;71
6.8;References;73
7;4 VIDEO AND RADIO ATTRIBUTES EXTRACTION FOR HETEROGENEOUS LOCATION ESTIMATION;76
7.1;1. Introduction;76
7.2;2. Related work;77
7.3;3. Main tasks of Ambient Intelligence systems;78
7.4;4. Architecture design;79
7.4.1;4.1 Inspiration;79
7.4.2;4.2 Mapping the Model into an AmI Architecture;81
7.4.3;4.3 Artificial Sensing;82
7.4.4;4.4 Proposed structure;82
7.5;5. Context aware systems;84
7.5.1;5.1 Location feature;85
7.5.2;5.2 The formalism;86
7.5.3;5.3 Alignment and Extraction of Video and Radio Object Reports;88
7.6;6. Results;93
7.6.1;6.1 The environment;93
7.6.2;6.2 Results for video object extraction;93
7.6.3;6.3 Results for radio object extraction;93
7.6.4;6.4 Alignment results;95
7.7;7. Conclusions;95
7.8;8. Acknowledgments;96
7.9;References;96
8;5 DISTRIBUTED ACTIVE MULTICAMERA NETWORKS;102
8.1;1. Introduction;102
8.2;2. Sensing modalities;102
8.3;3. Vision for Ambient Intelligence;103
8.4;4. Architecture;104
8.5;5. Tracking and object detection;105
8.5.1;5.1 Object detection;105
8.5.2;5.2 Tracking;106
8.5.3;5.3 Appearance models;107
8.5.4;5.4 Track data;108
8.6;6. Normalization;108
8.7;7. Multi-camera coordination;110
8.8;8. Multi-scale image acquisition;111
8.8.1;8.1 Active Head Tracking and Face Cataloging;112
8.8.2;8.2 Uncalibrated, multiscale data acquisition;114
8.8.3;8.3 Extensions;115
8.9;9. Indexing Surveillance Data;115
8.9.1;9.1 Visualization;116
8.10;10. Privacy;116
8.11;11. Conclusions;117
8.12;References;117
9;6 A DISTRIBUTED MULTIPLE CAMERA SURVEILLANCE SYSTEM;120
9.1;1. Introduction;120
9.2;2. System architecture;121
9.3;3. Motion detection and single-view tracking;121
9.3.1;3.1 Motion Detection;122
9.3.2;3.2 Scene Models;124
9.3.3;3.3 Target Tracking;125
9.3.4;3.4 Partial Observation;126
9.3.5;3.5 Target Reasoning;129
9.4;4. Multi view tracking;133
9.4.1;4.1 Homography Estimation;133
9.4.2;4.2 Least Median of Squares;134
9.4.3;4.3 Feature Matching Between Overlapping Views;135
9.4.4;4.4 3D Measurements;136
9.4.5;4.5 Tracking in 3D;137
9.4.6;4.6 Non-Overlapping Views;139
9.5;5. System architecture;142
9.5.1;5.1 Surveillance Database;143
9.6;6. Summary;145
9.7;7. Appendix;147
9.7.1;7.1 Kalman Filter;147
9.7.2;7.2 Homography Estimation;148
9.7.3;7.3 3D Measurement Estimation;149
9.8;References;150
10;7 LEARNING AND INTEGRATING INFORMATION FROM MULTIPLE CAMERA VIEWS;152
10.1;1. Introduction;152
10.1.1;1.1 Semantic Scene Model;154
10.2;2. Learning point-based regions;156
10.3;3. Learning trajectory-based regions;159
10.3.1;3.1 Route model;159
10.3.2;3.2 Learning algorithm;161
10.3.3;3.3 Segmentation to paths and junctions;162
10.4;4. Activity analysis;163
10.5;5. Integration of information from multiple views;164
10.5.1;5.1 Multiple Camera Activity Network (MCAN);166
10.6;6. Database;168
10.6.1;6.1 Metadata Generation;171
10.7;7. Summary;175
10.8;References;175
11;8 FAST ONLINE SPEAKER ADAPTATION FOR SMART ROOM APPLICATIONS;178
11.1;1. Introduction;178
11.2;2. Description of the proposed on-line adaptation technique;179
11.3;3. Implementation details of proposed approach;183
11.3.1;3.1 Calculation of in an FST framework;183
11.4;4. Experimental details and results;185
11.5;5. Conclusions;187
11.6;References;187
12;9 STEREO-BASED 3D FACE RECOGNITION SYSTEM FOR AMI;190
12.1;1. Introduction;190
12.2;2. Face Recognition: Review;192
12.2.1;2.1 Face Recognition from Still Images;192
12.2.2;2.2 Face Recognition from Image Retrievals;193
12.2.3;2.3 3D Face Recognition;194
12.2.4;2.4 NIVA System Overview;195
12.3;3. NIVA 3D Vision System;195
12.3.1;3.1 NIVA 3D Stereo-based Face Database;196
12.4;4. Face Recognition in NIVA;196
12.4.1;4.1 Fisher/Linear Discriminant Analysis;197
12.4.2;4.2 Face Classification in NIVA;198
12.4.3;4.3 Pattern Vectors;198
12.5;5. NIVA Dynamic Indexing to Database and Recognition;199
12.6;6. NIVA Implementation of Indexing and Recognition;199
12.6.1;6.1 Feature Space;200
12.6.2;6.4 Step 2: Face Recognition;202
12.7;7. Testing and Results;202
12.7.1;7.1 Indexing and Recognition Performance;203
12.7.2;7.2 Conclusion and Future Work;205
12.8;References;209
13;10 SECURITY AND BUILDING INTELLIGENCE;212
13.1;1. Introduction;212
13.2;2. System Description;213
13.3;3. People tracking and counting;215
13.3.1;3.1 People tracking;215
13.3.2;3.2 People counting;216
13.4;4. Event detection and association;217
13.5;5. Experimental results;217
13.6;6. AmI for training environments;218
13.7;7. Conclusions;222
13.8;References;223
14;11 SUSTAINABLE CYBERNETICS SYSTEMS;226
14.1;1. Encoding Interplay and Co-Evolution;229
14.1.1;1.1 Encoding Interplay between Natural and Cybernetic Systems;229
14.1.2;1.2 Encoding Co-Evolution of Natural and Cybernetic Systems;236
14.2;2. Sustaining Ambient Intelligence;245
14.2.1;2.1 Propagating Structure and Function;245
14.2.2;2.2 Indicators of Sustainability;249
14.2.3;2.3 Collective Intelligent Agents;250
14.3;3. Conclusion;250
14.4;References;251
15;Index;252

Leseproben

Chapter 6

A DISTRIBUTED MULTIPLE CAMERA SURVEILLANCE SYSTEM (p.107-108)

T.Ellis, J.Black, M.Xu and D.Makris
Digital Imaging Research Centre (DIRC), Kingston University, UK
{t.ellis,j.black,m.xu,d.makris}@kingston.ac.uk

1. Introduction

An important capability of an ambient intelligent environment is the capacity to detect, locate and identify objects of interest. In many cases interesting can move, and in order to provide meaningful interaction, capturing and tracking the motion creates a perceptively-enabled interface, capable of understanding and reacting to a wide range of actions and activities. CCTV systems fulfill an increasingly important role in the modern world, providing live video access to remote environments. Whilst the role of CCTV has been primarily focused on rather specific surveillance and monitoring tasks (i.e. security and traffic monitoring), the potential uses cover a much wider range.

The proliferation of video security surveillance systems over the past 5-10 years, in both public and commercial environments, is extensively used to remotely monitor activity in sensitive locations and publicly accessible spaces. In town and city centres, surveillance has been acknowledged to result in significant reductions in crime. However, in order to provide comprehensive and large area coverage of anything but the simplest environments, a large number of cameras must be employed.

In complex and cluttered environments with even moderate numbers of moving objects (e.g. 10-20) the problem of tracking individual objects is significantly complicated by occlusions in the scene, where an object may be partially occluded or totally disappear from camera view for both short or extended periods of time. Static occlusion results from objects moving behind (with respect to the camera) fixed elements in the scene (e.g. walls, bushes), whilst dynamic occlusion occurs as a result of moving objects in the scene occluding each other, where targets may merge or separate (e.g. a group of people walking together).

Information can be combined from multiple viewpoints to improve reliability, particularly taking advantage of the additional information where it minimises occlusion within the field-of-view (FOV). We treat the non-visible regions between camera views as simply another type of occlusion, and employ spatio-temporal reasoning to match targets moving between cameras that are spatially adjacent. The "boundaries" of the system represent locations from which previously unseen targets can enter the network.

To aid robust tracking across the camera network requires the system to maintain a record of each target entering the system and throughout its duration. When a target disappears from any camera FOV, motion prediction, colour identification, and learnt route patterns are used to re-establish tracking when the target reappears. Each target is maintained as a persistent object in the active database and spatial and temporal reasoning are used to detect these activities and ensure that entries are not retained for indefinite periods.

This chapter describes a multi-camera surveillance network that can detect and track objects (principally pedestrians and vehicles) moving through an outdoor environment. The remainder of this chapter is divided into four sections. The first describes the architecture of our multi-camera surveillance system. The second considers the image analysis methods for detecting and tracking objects within a single camera view. The next section deals with the integration of information from multiple cameras. The final section describes the structure of the database.