George Kamberov: Research Projects
- Vision and Visualization: sensor data; the mobile network
overarching goal is to develop an interactive immersive environment
for monitoring, video surveillance, control, and simulation of a network of wireless
agents deployed in an urban area and to integrate the environment
with the WiNSeC Testbed at Stevens.
- Funding: WiNSeC, Picatinny, US Army, March 2002-August 2004
- People: Y.Yukota, A. Jain, S. Goda, K. Chintalapati,
J. Oliensis (2004), E. Angeloupolou, G. Kamberov(PI)
Click to Enlarge/Shrink
- Real-Time Geographic Visualization for the Mobile Networked Multiple Input Multiple Output (MNM)
Integrated Test Environment (MITE)
Click to Enlarge/Shrink
- Summary: Develop real-time methods for mobile network visualization, and implement and deploy a real-time geographic visualization GUI (RTGV) for the MNM
Integrated Test Environment at the Lakehurst naval airfield.
The RTGV supports on-the fly mobile network and environment data collection, tracking and display of the mobile agents' positions in
the field, drill-down functionalities to show the network links status and
connections properties, and real-time rendering and interaction with large
scenes involving geo-referenced 3D terrain models of the testing area.
- Funding: Lucent, DARPA, March 2004 - April 2005
- People: A. Jain, R. Bader, A. Corrighan , B. Luczynsky, G. Kamberov
- Video Surveillance and Visualization Suite component of the Secure Infrastructure Technology Laboratory at Stevens
- Summary: Build infrastructure and develop methods for video-based
surveillance of the SINTEL test area and the surrounding eight miles radius
sector of the Hudson river. The two principal objectives are: (i) to support
automatic correlation of the acoustic noise with sources, locations, and
behavior; (ii) develop automatic video-based methods to identify and track
aerial and surface vessels.
- Funding: ONR, August 2005 - June 2009
- People: R.Bader, B.Luczynsky, I. Meraly, M. Powers, R. Stolkin,
G. Kamberov (PI)
|Double click to download movie mpg (left) and wmv (right)
- Video-acoustic-seismic sensor fusion in an urban environment
- Summary: Investigate, develop, and test
design concepts for sensor fusion for arrays combining video, acoustic, and
seismic sensors in an urban environment and demonstrate their
applicability to tasks including threat and object identification and tracking,
object and incident classification and geo-location. Key objectives are to
develop new methods for learning and exploiting vehicle signatures based on shape/color, acoustic, and seismic tags, and
assisted adaptive beam forming in a cluttered environment.
- Funding: ARDEC US Army, September 2006- September 2007
- People: TBA, G. Kamberov (PI)
surveillance and forensics
- Summary: This is a large project whose main research topics fall in two main groups:
(i) Robust real-time surveillance and scene analysis based on new methods for object classification, identification of scene changes and
from low grade video, rapid terrain acquisition and GPS-free geolocation. (ii) Forensic analysis of video recordings
and creation and maintenance of reference data bases.
- Corporate Funding, Period: August 2006- April 2009.
- People: TBA, G. Kamberov (PI)
- Manifold learning and scene segmentation from unorganized 3D point
- Summary: This is a long term project - we are developing a novel automatic point-based
approach for assigning topology and for defining and computing
surface geometry from unorganized 3D point clouds obtained from 3D
scenes. The methods work on noisy data from scenes involving
multiple objects, partial views, surfaces with boundary points, and
tangent points. The topology is based on a implicit estimate
of a surface distance and is used to segment the 3D point cloud into
manifold components. To achieve robustness and speed the topology
and geometry are extracted without curve and surface reconstruction.
- Funding: ARDEC US Army, WiNSeC PicatinnyUS Army, NSF pending.
November 2000 -
- People: A. Jain, G. Kamberova, G. Kamberov (PI)
Double click to play a GIF animation.
- Modeling and particle-in-cell (PIC) simulation of ELM-produced particle
and energy pulses and their effect on the SOL in TOKAMAKs .
- Summary: This is an international collaboration with a number of Bulgarian
mathematicians and physicists. We are developing phenomenological models
and simulation code for the outward flow of relativistic particles propagating from the separatrix to the divertor target, and for the inward flow of impurities in
the SOL. These models will be validated using complex diagnostics from
different regimes of JET operations. They are needed for more realistic
simulations of fuel plasma in ITER - like runs with JET. The new PIC code
is used to perform extensive simulations of full
discharge of plasma. The code incorporates a novel approach to Monte
Carlo simulation of alpha particle collisions, processes with relativistic
electrons, and key collisions of impurities (e.g. ionization, recombination,
- Funding: EURATOM Key Action Controlled Thermo-Nuclear
Fusion, ITM PROJECT 3 campaigns C15-C17 JET work program,
the European Union, Bulgarian Ministry of Education and
Science. September 2001-
- People: Bulgaria: T. Nikolov, D.Vasileva, P. Marinov, V. Hristov, N. Manev, P. Godev, L. Popova (PI);
USA: G. Kamberov
- Solutions of the Dirac Equation on Higher Genus Surfaces
- Summary: We will study solutions of the Dirac equation for spinor
transforms. These solutions are used to construct immersed surfaces in R3
with prescribed properties by reducing the geometric constructions to analyzing
first order linear systems of partial differential equations. In particular, we
will investigate the construction of Bonnet surfaces and conformal deformations
of closed surfaces of genus one. The work on Bonnet surfaces is part of an ongoing
program on the Bonnet conjecture. The deformations of tori will be studied in
context of human gait analysis.
- Funding: International research development grants from Stevens and the
German Government, August 2006-February 2007.
- People: C. Moller (Germany) and G. Kamberov (PI)
- Flexible mobile platforms for continuous range and imagery
- Summary: This project will acquire a
state-of-the-art, custom sensor suite, the 3D Content Digitization Suite
(3DCDS), to support research and performance evaluation of computer vision
and robotics algorithms in challenging real-world, outdoors, dynamic scenes.
The project will result in novel methods for
2D-3D fusion for real-time geometry processing; large-scale scene
understanding and hierarchical semantic context inference.
- Funding: National Science Foundation, August 2009-July 2012.
- People: P. Mordohai, G. Kamberova, J. Oliensis, (Co-Pis) G. Kamberov (PI)
Last Update 09/26/09