SMARTVIS

 

SMARTVIS is a cooperation of several SMEs and a research organisation for the purpose of developing and marketing novel visual surveillance applications which are based on smart camera hardware.

Visual surveillance has important applications in various domains such as transport telematics, security, retail, advertisement and urban planning.

 

Automated video analysis usually involves

1) detection of interesting objects (e.g. cars, pedestrians),

2) tracking of objects over time in order to obtain their motion paths and

3) analysis of object tracks to reason about their behaviour and interaction.

Contradicting to claims of marketing departments, existing video surveillance techniques still have limited accuracy and usability when it comes to complex real world scenarios. Furthermore, a large number of cameras in a network prohibits the processing of video data in real-time (despite for the simplest tasks) by centralised systems based on PCs due to either bandwidth and/or computing limitations.
Smart cameras combine video sensing, processing and communication in a single embedded device. They have gained remarkable processing power and maturity since their invention in the mid 90‘s. Today their hardware is powerful enough so that they could be used for demanding computer vision tasks. One of the drawbacks of currently available smart camera solutions is the gap between software development tools and libraries for the PC and the software development facilities on the smart camera. Hence, compared to PC-based solutions, today’s software on smart cameras is years behind current research when it comes to the demanding tasks of object detection, tracking and object track analysis for automatic visual surveillance.
 

SMARTVIS objectives:

1) Development of high end algorithms for object detection, tracking and object track analysis exceeding the state-of-the-art level,

2) creation of a software framework for object detection and tracking, tracking and motion analysis both on smart cameras and PCs,

3) design of protocols and implementation of software for the integration and configuration of smart cameras over a network,

4) creation of new vision sensors based on the smart camera,

5) and setting up marketing and distribution for the new smart camera solutions.