- ホーム
- > 洋書
- > 英文書
- > Science / Mathematics
Full Description
In smart cities, video surveillance is essential for public safety, evolving beyond simple camera installations and centralized monitoring due to the overwhelming amount of footage that challenges human operators. To enhance anomaly detection, experts have developed sophisticated computer vision techniques that classify events as normal or abnormal.
Smart Public Safety Video Surveillance System explores an end-to-end urban video surveillance system, which aims to address asymmetric threats through three key strategies: firstly, it employs a corrective signal called "task-specific QoE" that considers contextual factors; secondly, it utilizes machine learningdriven predictive systems and a method known as "similarity-based meta-reinforcement learning" for effective anomaly detection; and thirdly, it advocates for "zero-touch" self-management systems based on autonomous computing. This holistic approach ensures rapid adaptation and situational awareness, effectively meeting the demands of modern businesses and enhancing overall safety in dynamic urban environments.
Contents
Preface ix
List of Acronyms xi
Introduction xvii
Chapter 1. Literature Review on an End-to-End Video Surveillance System for Public Safety 1
1.1. General description: human threats in urban areas and abnormal situation detection 2
1.2. Analytics for video surveillance 2
1.2.1. Crowd behavior analysis 4
1.2.2. Traffic analysis 9
1.2.3. Environment analysis 11
1.2.4. Individual behavior analysis 12
1.2.5. General human threat-centric urban situation analysis 14
1.3. System architecture for video surveillance 21
1.3.1. Network architecture 21
1.3.2. Computing infrastructure 23
1.4. Analytics and architecture: studies and reflections 24
1.4.1. Threats: from cyber-to-physical space or physical-to-cyber space? 25
1.4.2. Video quality impact on video surveillance: from monitoring to task-specific analytics 27
1.4.3. End-to-end measurement: from traditional QoE to task-specific QoE 28
1.5. Challenges 29
1.6. Conclusion 31
Chapter 2. A Development Platform for Integration and Testing 33
2.1. Introduction 33
2.2. Proposed framework - QoE-driven SA-centric DSS 34
2.2.1. High-level view of the system: reinforcement signal and QoE 34
2.2.2. Detailed system framework: SA-centric DSS 36
2.3. Use case -Airbus DSSLC's target market 48
2.3.1. Introduction 48
2.3.2.Challenges 48
2.3.3. Purposes 49
2.3.4. Application case: Airbus DS SLC's business opportunity 49
2.3.5. Target system: Airbus DS SLC's flagship product 53
2.4. Conclusion 54
Chapter 3. A Multi-Criteria Enriched Corrective Signal with Endogenous, Exogenous and Human Factors 55
3.1. Context 55
3.2. Problem statement 57
3.3. Proposals 58
3.3.1. QoP for endogenous factor assessment 61
3.3.2. Task-specific QoE for endogenous, exogenous and human factors 66
3.4. Conclusion 74
Chapter 4. A Situational Awareness-centric Predictive System for Anomaly Detection 77
4.1. Context 77
4.2. Baseline 78
4.3. Problem statement 79
4.4. Proposals 81
4.4.1. Feature extraction experimentation and reviewing 82
4.4.2. Capability-oriented classifier study 83
4.4.3. Result-oriented classifier study 101
4.5. Conclusion 117
Chapter 5. Towards an Autonomic Intelligent Video Surveillance System 119
5.1. Context 119
5.2. Problem statement 120
5.3. Proposals 121
5.3.1.Time-based control 122
5.3.2. Event-triggered control 124
5.4. Conclusion 142
Conclusions and Perspectives 143
References 149
Index 161
