Object detection and classification in aerial hyperspectral imagery using a multivariate hit-or-miss transform

Research output: Contribution to journalConference Contribution

Abstract

High resolution aerial and satellite borne hyperspectral imagery provides a wealth of information about an imaged scene allowing for many earth observation applications to be investigated. Such applications include geological exploration, soil characterisation, land usage, change monitoring as well as military applications such as anomaly and target detection. While this sheer volume of data provides an invaluable resource, with it comes the curse of dimensionality and the necessity for smart processing techniques as analysing this large quantity of data can be a lengthy and problematic task. In order to aid this analysis dimensionality reduction techniques can be employed to simplify the task by reducing the volume of data and describing it (or most of it) in an alternate way. This work aims to apply this notion of dimensionality reduction based hyperspectral analysis to target detection using a multivariate Percentage Occupancy Hit or Miss Transform that detects objects based on their size shape and spectral properties. We also investigate the effects of noise and distortion and how incorporating these factors in the design of necessary structuring elements allows for a more accurate representation of the desired targets and therefore a more accurate detection. We also compare our method with various other common Target Detection and Anomaly Detection techniques.
LanguageEnglish
Article number1098619
Number of pages11
JournalProceedings of SPIE
Volume10986
Early online date14 May 2019
DOIs
Publication statusE-pub ahead of print - 14 May 2019
EventSPIE Defense and Commercial Sensing 2019 - Baltimore, United States
Duration: 14 Apr 201918 Apr 2019

Fingerprint

aerial photography
Object Classification
Hyperspectral Imagery
Target Detection
Object Detection
Hits
Target tracking
Anomaly Detection
Dimensionality Reduction
Transform
Antennas
Earth Observation
Military applications
Curse of Dimensionality
Spectral Properties
Alternate
Military
Soil
Percentage
anomalies

Keywords

  • hyperspectral image processing
  • mathematical morphology
  • hit-or-miss transform
  • template matching
  • object detection

Cite this

@article{de9d71c5dc9146f097ff2b1ca9e2e8ea,
title = "Object detection and classification in aerial hyperspectral imagery using a multivariate hit-or-miss transform",
abstract = "High resolution aerial and satellite borne hyperspectral imagery provides a wealth of information about an imaged scene allowing for many earth observation applications to be investigated. Such applications include geological exploration, soil characterisation, land usage, change monitoring as well as military applications such as anomaly and target detection. While this sheer volume of data provides an invaluable resource, with it comes the curse of dimensionality and the necessity for smart processing techniques as analysing this large quantity of data can be a lengthy and problematic task. In order to aid this analysis dimensionality reduction techniques can be employed to simplify the task by reducing the volume of data and describing it (or most of it) in an alternate way. This work aims to apply this notion of dimensionality reduction based hyperspectral analysis to target detection using a multivariate Percentage Occupancy Hit or Miss Transform that detects objects based on their size shape and spectral properties. We also investigate the effects of noise and distortion and how incorporating these factors in the design of necessary structuring elements allows for a more accurate representation of the desired targets and therefore a more accurate detection. We also compare our method with various other common Target Detection and Anomaly Detection techniques.",
keywords = "hyperspectral image processing, mathematical morphology, hit-or-miss transform, template matching, object detection",
author = "Fraser Macfarlane and Paul Murray and Stephen Marshall and Henry White",
year = "2019",
month = "5",
day = "14",
doi = "10.1117/12.2518103",
language = "English",
volume = "10986",
journal = "Proceedings of SPIE",
issn = "0277-786X",

}

Object detection and classification in aerial hyperspectral imagery using a multivariate hit-or-miss transform. / Macfarlane, Fraser; Murray, Paul; Marshall, Stephen; White, Henry.

In: Proceedings of SPIE, Vol. 10986, 1098619, 14.05.2019.

Research output: Contribution to journalConference Contribution

TY - JOUR

T1 - Object detection and classification in aerial hyperspectral imagery using a multivariate hit-or-miss transform

AU - Macfarlane, Fraser

AU - Murray, Paul

AU - Marshall, Stephen

AU - White, Henry

PY - 2019/5/14

Y1 - 2019/5/14

N2 - High resolution aerial and satellite borne hyperspectral imagery provides a wealth of information about an imaged scene allowing for many earth observation applications to be investigated. Such applications include geological exploration, soil characterisation, land usage, change monitoring as well as military applications such as anomaly and target detection. While this sheer volume of data provides an invaluable resource, with it comes the curse of dimensionality and the necessity for smart processing techniques as analysing this large quantity of data can be a lengthy and problematic task. In order to aid this analysis dimensionality reduction techniques can be employed to simplify the task by reducing the volume of data and describing it (or most of it) in an alternate way. This work aims to apply this notion of dimensionality reduction based hyperspectral analysis to target detection using a multivariate Percentage Occupancy Hit or Miss Transform that detects objects based on their size shape and spectral properties. We also investigate the effects of noise and distortion and how incorporating these factors in the design of necessary structuring elements allows for a more accurate representation of the desired targets and therefore a more accurate detection. We also compare our method with various other common Target Detection and Anomaly Detection techniques.

AB - High resolution aerial and satellite borne hyperspectral imagery provides a wealth of information about an imaged scene allowing for many earth observation applications to be investigated. Such applications include geological exploration, soil characterisation, land usage, change monitoring as well as military applications such as anomaly and target detection. While this sheer volume of data provides an invaluable resource, with it comes the curse of dimensionality and the necessity for smart processing techniques as analysing this large quantity of data can be a lengthy and problematic task. In order to aid this analysis dimensionality reduction techniques can be employed to simplify the task by reducing the volume of data and describing it (or most of it) in an alternate way. This work aims to apply this notion of dimensionality reduction based hyperspectral analysis to target detection using a multivariate Percentage Occupancy Hit or Miss Transform that detects objects based on their size shape and spectral properties. We also investigate the effects of noise and distortion and how incorporating these factors in the design of necessary structuring elements allows for a more accurate representation of the desired targets and therefore a more accurate detection. We also compare our method with various other common Target Detection and Anomaly Detection techniques.

KW - hyperspectral image processing

KW - mathematical morphology

KW - hit-or-miss transform

KW - template matching

KW - object detection

UR - https://www.spiedigitallibrary.org/conference-proceedings-of-spie/10986.toc

U2 - 10.1117/12.2518103

DO - 10.1117/12.2518103

M3 - Conference Contribution

VL - 10986

JO - Proceedings of SPIE

T2 - Proceedings of SPIE

JF - Proceedings of SPIE

SN - 0277-786X

M1 - 1098619

ER -