|
Accession Number
|
ADA565336
|
|
Title
|
Robust Dynamic Vision Methods for Persistent Surveillance and Enhanced Vehicle Autonomy.
|
|
Publication Date
|
Feb 2012
|
|
Media Count
|
10p
|
|
Personal Author
|
M. Sznaier O. I. Camps
|
|
Abstract
|
This research addressed the USAF s unprecedented proactive persistent surveillance Long Term Challenge. Specifically, we aimed at a substantial enhancement of the ability to conduct autonomous, video based, persistent intelligent surveillance, reconnaissance and threat assessment in highly uncertain, adversarial scenarios such as urban environments. At its core was a novel approach, stressing dynamic models as key enablers for finding, tracking and anticipating/assessing behavior of multiple targets using as inputs data streams from spatially distributed sensors. It included both theory developments in an emerging new field dynamics based extraction of information sparsely encoded in high dimensional data and an investigation of implementation issues.
|
|
Keywords
|
Actionable information
Computer vision
Controlled dynamic vision
Data fusion
Dynamics
Hammerstein wiener systems
Hybrid systems
Hybrid systems identification and model (Invalidation)
Learning machines
Manifold embedding
Manifolds(Mathematics)
Monitoring
Multiple camera tracking systems
Optical detection
Optical tracking
Persistent surveillance
Sparsely encoded visual information
Surveillance
Vision based tracking and activity monitoring
|
|
|
Source Agency
|
Non Paid ADAS
|
|
NTIS Subject Category
|
62 - Computers, Control & Information Theory
63C - Infrared & Ultraviolet Detection
46C - Optics & Lasers
|
|
Corporate Author
|
Northeastern univ., Boston, MA. Dept. of Electrical and Computer Engineering.
|
|
Document Type
|
Technical report
|
|
Title Note
|
Final rept. 1 May 2009-30 Nov 2011.
|
|
NTIS Issue Number
|
1304
|
|
Contract Number
|
FA9550-09-1-0253
|
