Sequential Adaptive Multi-Modality Target Detec-tion and Classification using Physics-Based Models

1
Sequential Adaptive Multi-Modality Target
Detec-tion and Classification using Physics-Based
Models

• Professor Andrew E. Yagle (PI) (EECS)
• Mine detection, channel identification
• Professor Alfred O. Hero III (EECS)
• Sensor scheduling, nonparametric statistical
  models
• Professor Kamal Sarabandi (Director, Rad Lab)
• Vehicle and foliage physics-based modelling
• Assistant Professor Marcin Bownik (Mathematics)
• Basis functions and mathematical modelling

2
Sequential Adaptive Multi-Modality Target
Detec-tion and Classification using Physics-Based
Models

• Professor Andrew E. Yagle
• Jay Marble, Siddharth Shah
• Professor Alfred O. Hero III
• Chris Kreucher, Doron Blatt, Jose Costa, Neal
  Patwari, Raghuram Rangarajan, Krishnakanth
  Subramanian, Mike Fitzgibbons, Cyrille Hory
• Professor Kamal Sarabandi
• Mark Casciato, Il-Suek Koh, M. Dehmolaian

3
Sequential Adaptive Multi-Modality Target
Detec-tion and Classification using Physics-Based
Models

• PROJECT SUPERVISION
• Dr. Douglas Cochran (DARPA)
• Dr. Russell Harmon (ARO)
• INDUSTRY COLLABORATION
• Veridian (formerly ERIM) of Ann Arbor

4
Sequential Adaptive Multi-Modality Target
Detec-tion and Classification using Physics-Based
Models

• Mine detection Yagle, Marble
• Vehicle modeling Sarabandi, Casciato
• Foliage modeling Sarabandi, Koh, Dehmolaian
• Sensor scheduling Hero, Kreucher
• Nonparametric statistics Hero, Blatt
• Distributed detection Hero, Patwari
• Basis functions Yagle, Bownik

5
HERO Accomplishments

• Developed non-parametric statistical modelling
  using MRFs for targetclutter vs. clutter
• Developed target model reduction technique
• Developed distributed multisensor detection using
  hierarchical sensor aggregation
• Developed myopic sequential adaptive sensor
  management for tracking

6
Sarabandi Accomplishments

• Performed phenomenological studies of
• (a) physics-based clutter models
• (b) physics-based target models
• Developed SAR/INSAR image simulator
• Developed time-reversal method for foliage
  camouflaged target detection
• Developed iterative frequency-correlation-based
  forest radar channel identification

7
YAGLE Accomplishments

• Developed mine detection algorithm from SAR using
  range migration imaging (with Jay Marble)
• Developed 2-D and3-D blind deconvolution algs for
  radar channel identification (with Siddharth
  Shah)
• Developing basis-function-based inverse
  scattering approach (work in progress with Marcin
  Bownik)

8
Synergistic Activities Hero

• VERIDIAN INTL, Ann Arbor
• C. Kreucher sensor management scheduling
• K. Kastella sensor management
• J. Ackenhusen mine detection
• ARL NAS-SED review panel member
• N. Patwari (student) summer internship
• ERIM B. Thelen, N. Subotic collaborators

9
Synergistic Activities Sarabandi
VERIDIAN John Ackenhusen BAE Norm Byer FCS
COMMUNICATIONS Jim Freibersiser (DARPA
PM) Barry Perlman (CECOM) ARL Ed Burke (mm
wave), Brian Sadler, Bruce Wallace
10
Synergistic Activities Yagle

• VERIDIAN INTL, Ann Arbor, MI
• Jay Marble, student (ARO mine research)
• Brian Fischer, student (Low RCS material design)
• Chris Wackerman, former Ph.D. student

11
Research Project Objectives

• Develop overall algorithm for detection of
  Tanks under trees landmines
• Initial focus TUT (can hit the ground running)
• Features of algorithm sequential detection,
  sensor management selection, physics-based
  models
• Simplify stochastic physics-based models using
  functional-analysis-based approximation
• Evaluate the resulting procedure on realistic
  models (statistical simulations) and real data

12
Issues Overall Algorithm

• How to select sensing modalities?
• What is value-added for combining other
  modalities? Is it worth additional cost?
• How do we implement data-adaptive
  configu-rations, e.g., selection of
  sources/receivers, based on scattering of targets
  and propagation in medium?
• What are the figures of merit?
• How to select decision thresholds?

13
(No Transcript)
14
Summary

• Sequential detection and classification
• Sensor scheduling and management
• Physics-based models with dimensionality reduced
  using functional analysis
• Vehicle and canopy scattering models already at
  UM permit test evaluations