The goal of this project is to develop a highly intelligent, next-generation video-based surveillance system. A service-oriented architecture as infrastructural basis of a scalable video surveillance system will be implemented.
Services are self-contained and can be used independently. Every service owns a public interface, which can be used by the subscribers within a network. Two of those services are a fully automatic camera calibration and a state-of-the-art semantic interpretation of the video content. The calibration process needs to be established without user interaction and should be able to calculate intrinsic and extrinsic parameters for cameras monitoring the traffic in a tunnel as well as for cameras monitoring a crowded place or other scenarios.
A middleware, which will be placed on top of the network layer, assures the communication between the ressources within the network.
This project will be performed in cooperation with:
Publications
M. Hödlmoser, H. Zollner, M. Kampel., “An Evaluation of Camera Calibration Methods Using Digital Low Cost Cameras” accepted at the 15th Computer Vision Winter Workshop (CVWW’10), Nove Hrady, Czech Republic, February 2010.
M. Hödlmoser, M. Kampel., “Towards Calibration Using Off-the-shelf Cameras” accepted at the 34th Annual Workshop of the Austrian Association for Pattern Recognition (AAPR), pp 97 – 104, Zwettl, Austria, May 2010.
M. Hödlmoser, M. Kampel., “Multiple Camera Self-Calibration and 3D Reconstruction Using Pedestrians”. 6th International Symposium on Visual Computing (ISVC), Las Vegas, Nevada, USA, Nov. 29 – Dec. 1, 2010.