People Image Analysis Consortium
 

Home -> Projects ->Video Surveillance and Monitoring (VSAM) Member Login  

Jump to: Components | Demos | Publications | Releases

 Description

This project developed a prototype end-to-end surveillance system and a suite of automated video surveillance technologies.

The prototype surveillance system consisting of a network of active video cameras. These cameras cooperate to provide continuous coverage of people and vehicles moving throughout a cluttered urban environment. The automated video surveillance algorithms automatically "parse" people and vehicles from raw video, determine their geolocations, and insert them into a dynamic scene visualization.

 Components

This project demostrated the following components of advanced surveillance techniques. Details are described in the published papers below.
 1. Real-time moving object detection and tracking from stationary and moving camera platforms.
 2. Recognition of generic object classes (e.g. human, sedan, truck) and specific object types (e.g. campus police car, FedEx van).
 3. Object pose estimation with respect to a geospatial site model.
 4. Active camera control and multi-camera cooperative tracking.
 5. Human gait analysis.
 6. Recognition of simple multi-agent activities.
 7. Real-time data dissemination, data logging and dynamic scene visualization.

 Demos

VSAM Overview
VSAM Overall Demo
This is an demo video about the overall VSAM project.
Download: mpg [68.2MB]
Single Camera Surveillance
Motion Detection and Tracking
This is a video about how the system detects, classifies, and tracks vechivles and persons. During the active tracking, camera is moved based on tracking results to keep the target within field of view.
Download: avi [8.6MB]
Motion Detection from Static Camera
This is a video about how the system detects, classifies moving vehicle and persons from a static camera.
Download: avi [7.5MB]
Motion Detection from Static Camera 2
This is another video about how the system detects, classifies moving objects from a static camera.
Download: avi [21.1MB]
Motion Detection from Moving Camera
This is a video about how the system detects, classifies moving targets from a moving camera.
Download: avi [4.0MB]
Multiple Camera Surveillance
Master-Slave Surveillance on Vehicles
This is a video about how the system moitors vechicles in master-slave manner.
Download: avi [7.6MB]
Master-Slave Surveillance on Vehicles 2
This is another video about how the system moitors vechicles in master-slave manner.
Download: avi [7.2MB]
Master-Slave Surveillance on People
This is a video about how the systme moitors people in master-slave manner.
Download: avi [10.6MB]
Site Modeling and Visualization
Geolocation Visualization
This is a video about how the systme build a geographic model and visualize tracking trajectory.
Download: avi [6.8MB]
Site Visulization
This is a video about how the systme build a site model and visualize overall surveillance scene.
Download: avi [4.1MB]
Geolocation Modeling
This is a video about how the systme builds a site model and visualize tracking trajectory.
Download: avi [17.8MB]
Geolocation Modeling 2
This is a video about how the systme builds a site model and visualize tracking trajectory.
Download: avi [12.5MB]

 Selected Publications [ View All 14 Publications ]

We grant permission to use and publish all data and results from this site, provided the following paper is referenced:

 Releases [ Members Only ]


PIA Consortium is a part of Robotics Institute, School of Computer Science, Carnegie Mellon University