Coding theoretic approach to segmentation and robust CFAR-detection for LADAR images

Unoma Ndili; Richard Baraniuk; Hyeokho Choi; Robert Nowak; Mário Figueiredo

doi:10.1117/12.445354

Coding theoretic approach to segmentation and robust CFAR-detection for LADAR images

Unoma Ndili, Richard Baraniuk, Hyeokho Choi, Robert Nowak, Mário Figueiredo

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Scopus citations

Abstract

In this paper, we present an unsupervised scheme aimed at segmentation of laser radar (LADAR) imagery for Automatic Target Detection. A coding theoretic approach implements Rissanen's concept of Minimum Description Length (MDL) for estimating piecewise homogeneous regions. MDL is used to penalize overly complex segmentations. The intensity data is modeled as a Gaussian random field whose mean and variance functions are piecewise constant across the image. This model is intended to capture variations in both mean value (intensity) and variance (texture). The segmentation algorithm is based on an adaptive rectangular recursive partitioning scheme. We implement a robust constant false alarm rate (CFAR) detector on the segmented intensity image for target detection and compare our results with the conventional cell averaging (CA) CFAR detector.

Original language	English (US)
Title of host publication	Proceedings of SPIE - The International Society for Optical Engineering
Editors	F.A. Sadjadi
Pages	86-94
Number of pages	9
Volume	4379
DOIs	https://doi.org/10.1117/12.445354
State	Published - 2001
Event	Automatic Target Recognition XI - Orlando, FL, United States Duration: Apr 17 2001 → Apr 20 2001

Other

Other	Automatic Target Recognition XI
Country/Territory	United States
City	Orlando, FL
Period	4/17/01 → 4/20/01

Keywords

Adaptive Recursive Partitioning
Automatic Target Detection
CFAR
Laser Radar
MDL
Segmentation

ASJC Scopus subject areas

Electrical and Electronic Engineering
Condensed Matter Physics

Access to Document

10.1117/12.445354

Cite this

@inproceedings{63a5332881a3475a853533b33db34de0,

title = "Coding theoretic approach to segmentation and robust CFAR-detection for LADAR images",

abstract = "In this paper, we present an unsupervised scheme aimed at segmentation of laser radar (LADAR) imagery for Automatic Target Detection. A coding theoretic approach implements Rissanen's concept of Minimum Description Length (MDL) for estimating piecewise homogeneous regions. MDL is used to penalize overly complex segmentations. The intensity data is modeled as a Gaussian random field whose mean and variance functions are piecewise constant across the image. This model is intended to capture variations in both mean value (intensity) and variance (texture). The segmentation algorithm is based on an adaptive rectangular recursive partitioning scheme. We implement a robust constant false alarm rate (CFAR) detector on the segmented intensity image for target detection and compare our results with the conventional cell averaging (CA) CFAR detector.",

keywords = "Adaptive Recursive Partitioning, Automatic Target Detection, CFAR, Laser Radar, MDL, Segmentation",

author = "Unoma Ndili and Richard Baraniuk and Hyeokho Choi and Robert Nowak and M{\'a}rio Figueiredo",

year = "2001",

doi = "10.1117/12.445354",

language = "English (US)",

volume = "4379",

pages = "86--94",

editor = "F.A. Sadjadi",

booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

note = "Automatic Target Recognition XI ; Conference date: 17-04-2001 Through 20-04-2001",

}

TY - GEN

T1 - Coding theoretic approach to segmentation and robust CFAR-detection for LADAR images

AU - Ndili, Unoma

AU - Baraniuk, Richard

AU - Choi, Hyeokho

AU - Nowak, Robert

AU - Figueiredo, Mário

PY - 2001

Y1 - 2001

N2 - In this paper, we present an unsupervised scheme aimed at segmentation of laser radar (LADAR) imagery for Automatic Target Detection. A coding theoretic approach implements Rissanen's concept of Minimum Description Length (MDL) for estimating piecewise homogeneous regions. MDL is used to penalize overly complex segmentations. The intensity data is modeled as a Gaussian random field whose mean and variance functions are piecewise constant across the image. This model is intended to capture variations in both mean value (intensity) and variance (texture). The segmentation algorithm is based on an adaptive rectangular recursive partitioning scheme. We implement a robust constant false alarm rate (CFAR) detector on the segmented intensity image for target detection and compare our results with the conventional cell averaging (CA) CFAR detector.

AB - In this paper, we present an unsupervised scheme aimed at segmentation of laser radar (LADAR) imagery for Automatic Target Detection. A coding theoretic approach implements Rissanen's concept of Minimum Description Length (MDL) for estimating piecewise homogeneous regions. MDL is used to penalize overly complex segmentations. The intensity data is modeled as a Gaussian random field whose mean and variance functions are piecewise constant across the image. This model is intended to capture variations in both mean value (intensity) and variance (texture). The segmentation algorithm is based on an adaptive rectangular recursive partitioning scheme. We implement a robust constant false alarm rate (CFAR) detector on the segmented intensity image for target detection and compare our results with the conventional cell averaging (CA) CFAR detector.

KW - Adaptive Recursive Partitioning

KW - Automatic Target Detection

KW - CFAR

KW - Laser Radar

KW - MDL

KW - Segmentation

UR - http://www.scopus.com/inward/record.url?scp=0035758523&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035758523&partnerID=8YFLogxK

U2 - 10.1117/12.445354

DO - 10.1117/12.445354

M3 - Conference contribution

AN - SCOPUS:0035758523

VL - 4379

SP - 86

EP - 94

BT - Proceedings of SPIE - The International Society for Optical Engineering

A2 - Sadjadi, F.A.

T2 - Automatic Target Recognition XI

Y2 - 17 April 2001 through 20 April 2001

ER -

Coding theoretic approach to segmentation and robust CFAR-detection for LADAR images

Abstract

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this