Computer Vision in the Age of IoT
1 like373 views
Computer vision and pattern recognition algorithms are important for IoT applications like smart homes and healthcare that involve large camera networks. Academic expertise is needed for accuracy and efficiency, while industrial concerns focus on system integration, configuration and management. The presentation describes a large-scale video surveillance system using heterogeneous information fusion and visualization across a university campus. It also discusses implementing system self-awareness through fault, environment and context awareness, and presents methods for real-time camera anomaly detection.
Computer Vision in the Age of IoT
- 1. COMPUTER VISION IN THE AGE OF IOT FROM BIG DATA TO SYSTEM SELF-AWARENESS YUAN-KAI WANG FU-JEN CATHOLIC UNIVERSITY 2014/8/26 Plenary speech in International Conference on Pattern Recognition, Stockholm, Sweden.1
- 2. 2 Healthcare Smart HomeEntertainment Security Transportation Industrial Others Internet of Things Retails
- 3. 3
- 4. IoC Challenges 4 ?? IoC is closely related to camera networks, which becomes a large-scale system. ?? System-level thinking is necessary ?? Academic expertise on algorithms, such as computer vision and pattern recognition, are important for accuracy, efficiency and convenience. ?? Industrial concerns on system integration are critical for configuration, maintenance and management. ?? Autonomic computing with self-awareness ?? Paul Horn, 2001. ?? Camera collaboration of anomaly events
- 5. Our Experience Wide-Area Video Surveillance in Campus 5
- 6. Video Surveillance Cloud 6 Intrusion detection PTZ pedestrian tracking Camera anomaly detection Video Synopsis Geographical visualizationMulti-resolution fusion and visualization Parking lots estimation Illegal parking detection Video based face extraction Storage Area Network PC Mobile device Multi-core Hypervisor GPGPU Yuan-Kai Wang et al. "A Large Scale Video Surveillance System with Heterogeneous Information Fusion and Visualization for Wide Area Monitoring," IEEE Int. Conf. IIHMSP, 2012.
- 8. System Self-Awareness 8 Fault awareness Environment awareness Context awareness
- 9. Camera Anomaly Detection: Self-awareness of Camera 9 Yuan-Kai Wang et al. "Real-Time Camera Anomaly Detection for Real-World Video Surveillance," ICMLC, 2011. Normal Anomaly! Background updating graph cut Online Kalman filter finite state machine Features of image quality and replacement
- 10. Traffic Camera Anomaly Detection Yuan-Kai Wang et al. "Traffic Camera Anomaly Detection," ICPR 2014. (8/28, Balcony Level5, PM 4:00-5:40, paper no. : ThCT4p.26) 10
- 11. 11
- 12. Related Publications "? "A Large Scale Video Surveillance System with Heterogeneous Information Fusion and Visualization for Wide Area Monitoring," IEEE International Conference on Intelligent Information Hiding and Multimedia Signal Processing, 2012. "? "Real-Time Camera Anomaly Detection for Real-World Video Surveillance," International Conference on Machine Learning and Cybernetics, 2011. "? "Traffic Camera Anomaly Detection," ICPR 2014. (8/28, Balcony Level5, PM 4:00-5:40, paper no. : ThCT4p.26) 12