Experimental evaluation of FLIR ATR approaches - A comparative study

Baoxin Li, R. Chellappa, Q. Zheng, S. Der, N. Nasrabadi, L. Chan, L. Wang

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

This paper presents an empirical evaluation of a number of recently developed Automatic Target Recognition algorithms for Forward-Looking Infrared (FLIR) imagery using a large database of real FLIR images. The algorithms evaluated are based on convolutional neural networks (CNN), principal component analysis (PCA), linear discriminant analysis (LDA), learning vector quantization (LVQ), modular neural networks (MNN), and two model-based algorithms, using Hausdorff metric-based matching and geometric hashing. The evaluation results show that among the neural approaches, the LVQ- and MNN-based algorithms perform the best; the classical LDA and the PCA methods and our implementation of the geometric hashing method ended up in the bottom three, with the CNN- and Hausdorff metric-based methods in the middle. Analyses show that the less-than-desirable performance of the approaches is mainly due to the lack of a good training set.

Original languageEnglish (US)
Pages (from-to)5-24
Number of pages20
JournalComputer Vision and Image Understanding
Volume84
Issue number1
DOIs
StatePublished - Oct 2001
Externally publishedYes

Fingerprint

Infrared radiation
Neural networks
Vector quantization
Discriminant analysis
Principal component analysis
Automatic target recognition

Keywords

  • Automatic target recognition
  • Performance Evaluation

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Experimental evaluation of FLIR ATR approaches - A comparative study. / Li, Baoxin; Chellappa, R.; Zheng, Q.; Der, S.; Nasrabadi, N.; Chan, L.; Wang, L.

In: Computer Vision and Image Understanding, Vol. 84, No. 1, 10.2001, p. 5-24.

Research output: Contribution to journalArticle

Li, B, Chellappa, R, Zheng, Q, Der, S, Nasrabadi, N, Chan, L & Wang, L 2001, 'Experimental evaluation of FLIR ATR approaches - A comparative study', Computer Vision and Image Understanding, vol. 84, no. 1, pp. 5-24. https://doi.org/10.1006/cviu.2001.0938
Li, Baoxin ; Chellappa, R. ; Zheng, Q. ; Der, S. ; Nasrabadi, N. ; Chan, L. ; Wang, L. / Experimental evaluation of FLIR ATR approaches - A comparative study. In: Computer Vision and Image Understanding. 2001 ; Vol. 84, No. 1. pp. 5-24.
@article{f61405562d314f7e8e5b2c7b56ac28ec,
title = "Experimental evaluation of FLIR ATR approaches - A comparative study",
abstract = "This paper presents an empirical evaluation of a number of recently developed Automatic Target Recognition algorithms for Forward-Looking Infrared (FLIR) imagery using a large database of real FLIR images. The algorithms evaluated are based on convolutional neural networks (CNN), principal component analysis (PCA), linear discriminant analysis (LDA), learning vector quantization (LVQ), modular neural networks (MNN), and two model-based algorithms, using Hausdorff metric-based matching and geometric hashing. The evaluation results show that among the neural approaches, the LVQ- and MNN-based algorithms perform the best; the classical LDA and the PCA methods and our implementation of the geometric hashing method ended up in the bottom three, with the CNN- and Hausdorff metric-based methods in the middle. Analyses show that the less-than-desirable performance of the approaches is mainly due to the lack of a good training set.",
keywords = "Automatic target recognition, Performance Evaluation",
author = "Baoxin Li and R. Chellappa and Q. Zheng and S. Der and N. Nasrabadi and L. Chan and L. Wang",
year = "2001",
month = "10",
doi = "10.1006/cviu.2001.0938",
language = "English (US)",
volume = "84",
pages = "5--24",
journal = "Computer Vision and Image Understanding",
issn = "1077-3142",
publisher = "Academic Press Inc.",
number = "1",

}

TY - JOUR

T1 - Experimental evaluation of FLIR ATR approaches - A comparative study

AU - Li, Baoxin

AU - Chellappa, R.

AU - Zheng, Q.

AU - Der, S.

AU - Nasrabadi, N.

AU - Chan, L.

AU - Wang, L.

PY - 2001/10

Y1 - 2001/10

N2 - This paper presents an empirical evaluation of a number of recently developed Automatic Target Recognition algorithms for Forward-Looking Infrared (FLIR) imagery using a large database of real FLIR images. The algorithms evaluated are based on convolutional neural networks (CNN), principal component analysis (PCA), linear discriminant analysis (LDA), learning vector quantization (LVQ), modular neural networks (MNN), and two model-based algorithms, using Hausdorff metric-based matching and geometric hashing. The evaluation results show that among the neural approaches, the LVQ- and MNN-based algorithms perform the best; the classical LDA and the PCA methods and our implementation of the geometric hashing method ended up in the bottom three, with the CNN- and Hausdorff metric-based methods in the middle. Analyses show that the less-than-desirable performance of the approaches is mainly due to the lack of a good training set.

AB - This paper presents an empirical evaluation of a number of recently developed Automatic Target Recognition algorithms for Forward-Looking Infrared (FLIR) imagery using a large database of real FLIR images. The algorithms evaluated are based on convolutional neural networks (CNN), principal component analysis (PCA), linear discriminant analysis (LDA), learning vector quantization (LVQ), modular neural networks (MNN), and two model-based algorithms, using Hausdorff metric-based matching and geometric hashing. The evaluation results show that among the neural approaches, the LVQ- and MNN-based algorithms perform the best; the classical LDA and the PCA methods and our implementation of the geometric hashing method ended up in the bottom three, with the CNN- and Hausdorff metric-based methods in the middle. Analyses show that the less-than-desirable performance of the approaches is mainly due to the lack of a good training set.

KW - Automatic target recognition

KW - Performance Evaluation

UR - http://www.scopus.com/inward/record.url?scp=0035481110&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035481110&partnerID=8YFLogxK

U2 - 10.1006/cviu.2001.0938

DO - 10.1006/cviu.2001.0938

M3 - Article

VL - 84

SP - 5

EP - 24

JO - Computer Vision and Image Understanding

JF - Computer Vision and Image Understanding

SN - 1077-3142

IS - 1

ER -