國立高雄大學圖資館 |

Language: English

Back

Ultra-Wideband EMI Sensing: Non-Meta...

Dartmouth College.

Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance.
Author:	Sigman, John Brevard.
Published:	Ann Arbor : ProQuest Dissertations & Theses, 2017
Description:	155 p.
Notes:	Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.
Notes:	Adviser: Fridon Shubitidze.
Contained By:	Dissertation Abstracts International79-04B(E).
Subject:	Electrical engineering.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10636032
ISBN:	9780355390216

Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance.
Sigman, John Brevard.

Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance. - Ann Arbor : ProQuest Dissertations & Theses, 2017 - 155 p.

Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.

Thesis (Ph.D.)--Dartmouth College, 2017.

Buried explosive hazards present a pressing problem worldwide. Millions of acres and thousands of sites are contaminated in the United States alone [1, 2]. There are three categories of explosive hazards: metallic, intermediate-electrical conducting (IEC), and non-conducting targets. Metallic target detection and classification by electromagnetic (EM) signature has been the subject of research for many years. Key to the success of this research is modern multi-static Electromagnetic Induction (EMI) sensors, which are able to measure the wideband EMI response from metallic buried targets. However, no hardware solutions exist which can characterize IEC and non-conducting targets. While high-conducting metallic targets exhibit a quadrature peak response for frequencies in a traditional EMI regime under 100 kHz, the response of intermediate-conducting objects manifests at higher frequencies, between 100 kHz and 15 MHz.

ISBN: 9780355390216Subjects--Topical Terms:

454503
Electrical engineering.

Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance.
LDR:03509nmm a2200337 4500 001 523951
005 20180517120325.5
008 180709s2017 ||||||||||||||||| ||eng d
020 $a 9780355390216
035 $a (MiAaPQ)AAI10636032
035 $a (MiAaPQ)dartmouth:11268
035 $a AAI10636032
040 $a MiAaPQ $c MiAaPQ
100 1 $a Sigman, John Brevard. $3 795452
245 1 0 $a Ultra-Wideband EMI Sensing: Non-Metallic Target Detection and Automatic Classification of Unexploded Ordnance.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2017
300 $a 155 p.
500 $a Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.
500 $a Adviser: Fridon Shubitidze.
502 $a Thesis (Ph.D.)--Dartmouth College, 2017.
520 $a Buried explosive hazards present a pressing problem worldwide. Millions of acres and thousands of sites are contaminated in the United States alone [1, 2]. There are three categories of explosive hazards: metallic, intermediate-electrical conducting (IEC), and non-conducting targets. Metallic target detection and classification by electromagnetic (EM) signature has been the subject of research for many years. Key to the success of this research is modern multi-static Electromagnetic Induction (EMI) sensors, which are able to measure the wideband EMI response from metallic buried targets. However, no hardware solutions exist which can characterize IEC and non-conducting targets. While high-conducting metallic targets exhibit a quadrature peak response for frequencies in a traditional EMI regime under 100 kHz, the response of intermediate-conducting objects manifests at higher frequencies, between 100 kHz and 15 MHz.
520 $a In addition to high-quality electromagnetic sensor data and robust electromagnetic models, a classification procedure is required to discriminate Targets of Interest (TOI) from clutter. Currently, costly human experts are used for this task. This expense and effort can be spared by using statistical signal processing and machine learning.
520 $a This thesis has two main parts. In the first part, we explore using the high frequency EMI (HFEMI) band (100 kHz-15 MHz) for detection of carbon fiber UXO, voids, and of materials with characteristics that may be associated with improvised explosive devices (IED). We constructed an HFEMI sensing instrument, and apply the techniques of metal detection to sensing in a band of frequencies which are the transition between the induction and radar bands. In this transition domain, physical considerations and technological issues arise that cannot be solved via the approaches used in either of the bracketing lower and higher frequency ranges.
520 $a In the second half of this thesis, we present a procedure for automatic classification of UXO. For maximum generality, our algorithm is robust and can handle sparse training examples of multi-class data. This procedure uses an unsupervised starter, semi-supervised techniques to gather training data, and concludes with supervised learning until all TOI are found. Additionally, an inference method for estimating the number of remaining true positives from a partial Receiver Operating Characteristic (ROC) curve is presented and applied to live-site dig histories.
590 $a School code: 0059.
650 4 $a Electrical engineering. $3 454503
650 4 $a Electromagnetics. $3 708560
690 $a 0544
690 $a 0607
710 2 $a Dartmouth College. $b Engineering. $3 795453
773 0 $t Dissertation Abstracts International $g 79-04B(E).
790 $a 0059
791 $a Ph.D.
792 $a 2017
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10636032