Adaptive target recognition

Bir Bhanu, Yingqiang Lin, Grinnell Jones, Jing Peng

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Target recognition is a multilevel process requiring a sequence of algorithms at low, intermediate and high levels. Generally, such systems are open loop with no feedback between levels and assuring their performance at the given probability of correct identification (PCI) and probability of false alarm (Pf) is a key challenge in computer vision and pattern recognition research. In this paper, a robust closed-loop system for recognition of SAR images based on reinforcement learning is presented. The parameters in model-based SAR target recognition are learned. The method meets performance specifications by using PCI and Pf as feedback for the learning system. It has been experimentally validated by learning the parameters of the recognition system for SAR imagery, successfully recognizing articulated targets, targets of different configuration and targets at different depression angles.

Original languageEnglish
Pages (from-to)289-299
Number of pages11
JournalMachine Vision and Applications
Volume11
Issue number6
DOIs
StatePublished - 1 Jan 2000

Fingerprint

Feedback
Open systems
Reinforcement learning
Closed loop systems
Computer vision
Pattern recognition
Learning systems
Specifications

Keywords

  • Parameter learning
  • Reinforcement learning
  • Target recognition

Cite this

Bhanu, Bir ; Lin, Yingqiang ; Jones, Grinnell ; Peng, Jing. / Adaptive target recognition. In: Machine Vision and Applications. 2000 ; Vol. 11, No. 6. pp. 289-299.
@article{29111c9a49014953bc285607c865d259,
title = "Adaptive target recognition",
abstract = "Target recognition is a multilevel process requiring a sequence of algorithms at low, intermediate and high levels. Generally, such systems are open loop with no feedback between levels and assuring their performance at the given probability of correct identification (PCI) and probability of false alarm (Pf) is a key challenge in computer vision and pattern recognition research. In this paper, a robust closed-loop system for recognition of SAR images based on reinforcement learning is presented. The parameters in model-based SAR target recognition are learned. The method meets performance specifications by using PCI and Pf as feedback for the learning system. It has been experimentally validated by learning the parameters of the recognition system for SAR imagery, successfully recognizing articulated targets, targets of different configuration and targets at different depression angles.",
keywords = "Parameter learning, Reinforcement learning, Target recognition",
author = "Bir Bhanu and Yingqiang Lin and Grinnell Jones and Jing Peng",
year = "2000",
month = "1",
day = "1",
doi = "10.1007/s001380050113",
language = "English",
volume = "11",
pages = "289--299",
journal = "Machine Vision and Applications",
issn = "0932-8092",
publisher = "Springer Verlag",
number = "6",

}

Adaptive target recognition. / Bhanu, Bir; Lin, Yingqiang; Jones, Grinnell; Peng, Jing.

In: Machine Vision and Applications, Vol. 11, No. 6, 01.01.2000, p. 289-299.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Adaptive target recognition

AU - Bhanu, Bir

AU - Lin, Yingqiang

AU - Jones, Grinnell

AU - Peng, Jing

PY - 2000/1/1

Y1 - 2000/1/1

N2 - Target recognition is a multilevel process requiring a sequence of algorithms at low, intermediate and high levels. Generally, such systems are open loop with no feedback between levels and assuring their performance at the given probability of correct identification (PCI) and probability of false alarm (Pf) is a key challenge in computer vision and pattern recognition research. In this paper, a robust closed-loop system for recognition of SAR images based on reinforcement learning is presented. The parameters in model-based SAR target recognition are learned. The method meets performance specifications by using PCI and Pf as feedback for the learning system. It has been experimentally validated by learning the parameters of the recognition system for SAR imagery, successfully recognizing articulated targets, targets of different configuration and targets at different depression angles.

AB - Target recognition is a multilevel process requiring a sequence of algorithms at low, intermediate and high levels. Generally, such systems are open loop with no feedback between levels and assuring their performance at the given probability of correct identification (PCI) and probability of false alarm (Pf) is a key challenge in computer vision and pattern recognition research. In this paper, a robust closed-loop system for recognition of SAR images based on reinforcement learning is presented. The parameters in model-based SAR target recognition are learned. The method meets performance specifications by using PCI and Pf as feedback for the learning system. It has been experimentally validated by learning the parameters of the recognition system for SAR imagery, successfully recognizing articulated targets, targets of different configuration and targets at different depression angles.

KW - Parameter learning

KW - Reinforcement learning

KW - Target recognition

UR - http://www.scopus.com/inward/record.url?scp=0006310978&partnerID=8YFLogxK

U2 - 10.1007/s001380050113

DO - 10.1007/s001380050113

M3 - Article

AN - SCOPUS:0006310978

VL - 11

SP - 289

EP - 299

JO - Machine Vision and Applications

JF - Machine Vision and Applications

SN - 0932-8092

IS - 6

ER -