Signal Processing Seminar


Title:  Real Time Data Integration and Pattern Recognition Across Disparate Sensing Systems: 
Toward Hazard Aware and Tele-Immersive Spaces

Speaker:  Peter Bajcsy
National Center for Supercomputing Applications (NCSA)
University of Illinois at Urbana-Champaign

Date:  Wednesday, April 9, 2008
Time:  4:00 PM to 5:00 PM
Where:  2269 Beckman Institute
 
________________________________________

Abstract: 

We have developed prototype solutions to many problems that require real-time data integration and 
pattern recognition across disparate sensing systems over both time and geography.  The sensing 
systems include wireless micro-electro-mechanical systems (MEMS) sensor networks, such as MICA 
sensors by Crossbow Inc., Radio Frequency Identification (RFID) tags and cameras that capture a 
variety of spectra, such as visible, thermal infrared, and hyperspectral information. Some of the sensing 
is adaptive in time and space by using a remotely controlled robot for the sensor deployment. Our research 
has focused on solving problems related to real-time data integration and pattern recognition, such as 
pattern recognition for robotic sensor deployment using various human-computer interfaces, synchronization 
of sensors and cameras, localization of sensors and objects by fusing acoustic time-of-flight localization and 
stereo vision approaches, mutual calibration of measurements from sensors and cameras, proactive camera 
control based on sensor readings, and real-time robust 3D reconstruction of dynamic scenes.

We applied the results of our research and development to two applications. First, it is a prototype of hazard 
aware spaces (HAS) to alert people in events like natural disasters, failures of human hazard attention, and 
intentionally harmful human behaviors. This HAS system integrates both emerging and standard sensing 
technologies to deliver sensor and camera data streams to a central location. Our data analysis algorithms then 
detect hazards, alert humans, and possibly analyze hazard sources. Once alerted, decision makers know when, 
where and what events occur. They can then use the gesture/voice/keyboard controlled robot to confirm the 
presence of hazards using sensor and video feedback, as well as to attempt to contain the hazards with the robot. 
Second, it is a prototype of tele-immersive spaces (TEEVE) to enable remote collaborations, training and art 
performances. The TEEVE system integrates multiple visible and thermal infrared video streams in order to 
detect objects of interest robustly, reconstruct them in 3D, integrate multiple 3D streams from geographically 
distributed locations and then display integrated 3D information in a virtual environment in real time.

Biography: 
 
Peter Bajcsy has earned his Ph.D. degree from the Electrical and Computer Engineering Department, 
University of Illinois at Urbana-Champaign, IL, 1997, and M.S. degree from the Electrical Engineering 
Department, University of Pennsylvania, Philadelphia, PA, 1994. He is currently with the National Center 
for Supercomputing Applications at the University of Illinois at Urbana-Champaign, Illinois, working as a 
research scientist on problems related to (1) theoretical modeling and experimental understanding of multi-
instrument measurement systems generating multi-dimensional multi-variate data, (2) automation of common 
image pre-processing and analysis tasks, and (3) development of cyber-environments. In the past, he had 
worked on real-time machine vision problems for semiconductor industry and synthetic aperture radar (SAR) 
technology for government contracting industry. He has developed several software systems for automatic 
feature extraction, feature selection, segmentation, classification, tracking and statistical modeling from 
electro-optical, SAR, laser and hyperspectral data sets. Dr. Bajcsy’s scientific interests include image and 
signal processing, statistical data analysis, data mining, pattern recognition, novel sensor technology, and 
computer and machine vision.