This course starts by looking at background information including smart camera architectures, distributed vision processing algorithms, collaborative feature extraction, data and decision fusion, and topics in seamless object detection and tracking. The intent is to familiarize the student with vision-based smart environments, so that the student can apply it to their problem of interest in computer vision research, including applications such as surveillance and tracking applications, multi-view vision for human-computer interaction, 3D scene analysis, and distributed multimedia and gaming.
Distributed vision networks is a multidisciplinary topic that interfaces the fields of computer vision, signal processing, pervasive computing, embedded programming, wireless sensor networks, and ambient intelligence. It creates opportunities for design paradigm shifts in these fields through emphasizing distributed vision processing and collaborative fusion of information over the network. In this premise, novel smart environment applications can be envisioned that are scalable, real-time, adaptive, interpretive, context-aware, and user-centric in nature.
Offered In: Winter 2009, University of Windsor
Text (not optional)
- Computer Vision: A Modern Approach, Forsyth, Ponce; Pearson Education, 2003. ISBN 0-13-085198-1
- Other readings as provided
- Multi-camera networks, Hamid Aghajan, Andrea Cavallaro; Elsevier, 2009. ISBN: 978-0-12-374633-7
Text (optional)
- Distributed Sensor Networks, S. Sitharama Iyengar, Richard R. Brooks; Chapman & Hall/Crc, 2004. ISBN: 978-1584883838
- Embedded Computer Vision (Advances in Pattern Recognition), Branislav Kisacanin, Shuvra S. Bhattacharyya, Sek Chai; Springer, 2008. ISBN: 978-1848003033
Project: Survey paper on Smart Camera Architectures
