Biologically-Inspired Smart Camera Design for Distributed Surveillance Applications
Speaker: Ehsan Norouznezhad, ITEE
When: 2008-10-27 13:00:00
Venue: 78-420
Host: Abbas Bigdeli
Abstract:Intelligent video surveillance is currently a hot topic in computer vision research.
Due to the high computational complexity of vision tasks and the real-time nature of these systems,
current software-based intelligent video surveillance systems are unable to perform sophisticated operations.
Smart cameras are a key component for future intelligent surveillance systems. They use embedded processing to
offload computationally intensive vision tasks from the host processing computers and increasingly reduce the data
flows over the network. This project proposes new class of Smart Cameras which is inspired by Human Vision System (HVS).
The three main properties of human vision system are modeled to enhance and accelerate vision tasks.
The first characteristic is to have multiple foveal visions alongside peripheral vision.
Inspired by dorsal stream of human visual pathway, the second feature is to process motion inside the camera and provide
the coordinates of moving objects in the scene. The last feature which is inspired by ventral stream of human visual
pathway is to extract the key features of the scene through successive processing layers. Furthermore based on the
proposed smart camera, a mathematical model for distributed surveillance systems is presented.
Biography:
Type: Ph.D confirmation
Contact:Abbas Bigdeli, seminar host (a.bigdeli@uq.edu.au)
or Heng Tao SHEN (ITEE seminar co-ordinator)
(shenht@itee.uq.edu.au)