Intelligent Distribuited Video Systems
for Surveillance & Quality Inspection
Introduction
for use in two topic scenarios: Surveillance and Quality Inspection.
Surveillance
The most recent surveillance applications require the development of a distribuited vision system in which different video/audio sensors are able to share information and co-operate in an intelligent andautonomous way.
The main objective of the project is the development of a system which will be able to:
- perform an intelligent data fusion procedure
- detect anomalous events
- let different sensors co-operate in order to track the target movement on the scene
The main steps to reach this result (leading to a solution which has to be applied from home security to factory/storage security) are:
- Step 1 (common to both Project Tasks)
- Development and implementation of a middleware for managing multimedia data streamings. This middleware is fundamental in such systems in order to let different sensors to share information and to let the user set the distribuited system parameters through a remote monitoring application.
- Development of a multi-vision system calibration procedure (both for non-overlapping and overlapping fields scenarios)
- Development of data-fusion algorithms for integrating information retrieved from different vision systems (omni and standard cameras)Development of co-operative algorithms for distribuited intelligent systems.
- Step 2
- Development of a common platform for eterogeneous sensors and applications
- Development of distribuited alorithms for detection and tracking
- Develompent of algorithms for events/objects recognition
Objectives
Development of a distribuited video system for monitoring of large areas in which different sensors co-operate and share data streams in order to track people and events in the area of interest.
Inspection
The main purpose of this Task is the development of a distributed vision system for quality inspection. The system will be able to look at the target product from different point of view and perform the control quality routine with proper speed and accuracy. These features will lead to:
- reduced cost per unity quality control
- augmented control speed
- augmented control accuracy
- Step 1 (common to both Project Tasks)
- Development and implementation of a middleware for managing multimedia data streamings. This middleware is fundamental in such systems in order to let different sensors to share information and to let the user set the distribuited system parameters through a remote monitoring application.
- Development of a multi-vision system calibration procedure (both for non-overlapping and overlapping fields scenarios)
- Development of data-fusion algorithms for integrating information retrieved from different vision systems (omni and standard cameras)Development of co-operative algorithms for distribuited intelligent systems.
- Step 2
- Development of algorithms for automatic calibration of the system for a given inspection scenario;
- Development of algorithms for optimal camera placing depending on target geometry and surface texture;
- Development of a system for automatic estimation of an optimal lighting technique for the requested defect detection;
- Data Logging & statistics
Objectives
Development of a distribuited vision system for quality inspection for different product types and defects.
Events
- June 18-19 2009: iDVS @ SMAU BUSINESS 2009 - Fiera di Bologna, Italy
- May 2009: visiting CIELLE (Treviso, Italy)
- May 2009: iDVS Project partecipates to "Forum della Ricerca e Innovazione" (University of Padua) - Centro Culturale Altinate, Padova (Italy)
- May 6-7 2009: presenting the iDVS Project @ SMAU BUSINESS 2009, Fiera di Padova (Italy)
People
Emanuele Menegatti
Assistant Professor of Computer Science
Stefano Ghidoni
Fellow Researcher
Riccardo Marogna
Fellow Researcher