ArticlePDF Available

Identification of Metallic Spheroids by Classification of Their Electromagnetic Induction Responses

Authors:
R. H. Chesney
R. H. Chesney
R. H. Chesney
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Yogadhish Das
Yogadhish Das
Yogadhish Das
  • This person is not on ResearchGate, or hasn't claimed this research yet.
John E. Mcfee at McFysics Consulting
John E. Mcfee
  • McFysics Consulting
Mabo Robert Ito at University of British Columbia
Mabo Robert Ito
Download full-text PDFDownload full-text PDF
Read full-text
Download citation

Abstract

An investigation into the feasibility of applying pattern recognition concepts to the classification of metallic objects by their electromagnetic induction responses was performed. The effect on the response of a limited set of steel spheroids due to various factors such as object shape, size, and orientation was examined and a pattern recognition scheme based on these results was proposed. Implementation of the scheme involved the development of a novel extension to the nearest mean vector type of classifier in which the concept of the class mean as a point in feature space was generalized to be a curve. The resultant pattern recognition scheme was tested on a representative test set which included 815 responses, corresponding to 104 variations in object and orientation. A success rate of greater than 96 percent was achieved. It is noted that the classifier extension developed provides a viable approach to classification of responses that very continuously with respect to a single parameter.
ieee-pami-6-1984-Identification-of-Metallic-Spheroids by-Classificati
on.pdf
Content uploaded by John E. Mcfee
Author content
All content in this area was uploaded by John E. Mcfee on Feb 23, 2015
Content may be subject to copyright.
ieee-pami-6-1984-Identification-of-Metallic-Spheroids by-Classificati
on.pdf
Content uploaded by John E. Mcfee
Author content
All content in this area was uploaded by John E. Mcfee on Oct 28, 2014
Content may be subject to copyright.
... Depending on the number of variables involved, such a locus could be a curve, a surface or a higher order manifold. A novel approach t o solving such pattern classi cation problems was developed Chesney et al., 1984, McFee and . This technique has been already discussed under magnetostatics. ...
... This technique has been already discussed under magnetostatics. The earlier Chesney et al., 1984 v ariant o f t h e technique was applied to measured EMI responses of four steel spheroids, similar in dimensions to artillery shells. The measurements were taken in a nonmetallic laboratory, and data were collected at di erent orientations and depths of the objects. ...
View
... The CSEM method also can detect and, to some extent, classify man-made targets such as pipes, underground structures and unexploded ordnance (UXO) (Chesney et al., 1984; Qian and Boerner, 1995; Huang and Won, 2003). The method is sensitive to subsurface spatial variations in electrical conductivity. ...
... In near-surface geophysical applications, controlled-source electromagnetic (CSEM) methods are widely used because they are capable of detecting and mapping the distribution of fluids, large voids, faults or lithological contacts (Nobes, 1996; Tezkan, 1999; Pellerin, 2002; Everett and Meju, 2003). The CSEM method also can detect and, to some extent, classify man-made targets such as pipes, underground structures and unexploded ordnance (UXO) (Chesney et al., 1984; Qian and Boerner, 1995; Huang and Won, 2003). The method is sensitive to subsurface spatial variations in electrical conductivity. ...
Experimental time-domain controlled source electromagnetic induction for highly conductive targets detection and discrimination
Article
Full-text available
The response of geological materials at the scale of meters and the response of buried targets of different shapes and sizes using controlled-source electromagnetic induction (CSEM) is investigated. This dissertation focuses on three topics; i) frac- tal properties on electric conductivity data from near-surface geology and processing techniques for enhancing man-made target responses, ii) non-linear inversion of spa- tiotemporal data using continuation method, and iii) classification of CSEM transient and spatiotemporal data. In the first topic, apparent conductivity profiles and maps were studied to de- termine self-affine properties of the geological noise and the effects of man-made con- ductive metal targets. 2-D Fourier transform and omnidirectional variograms showed that variations in apparent conductivity exhibit self-affinity, corresponding to frac- tional Brownian motion. Self-affinity no longer holds when targets are buried in the near-surface, making feasible the use of spectral methods to determine their pres- ence. The difference between the geology and target responses can be exploited using wavelet decomposition. A series of experiments showed that wavelet filtering is able to separate target responses from the geological background. In the second topic, a continuation-based inversion method approach is adopted, based on path-tracking in model space, to solve the non-linear least squares prob- lem for unexploded ordnance (UXO) data. The model corresponds to a stretched- exponential decay of eddy currents induced in a magnetic spheroid. The fast inversion of actual field multi-receiver CSEM responses of inert, buried ordnance is also shown. Software based on the continuation method could be installed within a multi-receiver CSEM sensor and used for near-real-time UXO decision. In the third topic, unsupervised self-organizing maps (SOM) were adapted for data clustering and classification. The use of self-organizing maps (SOM) for central- loop CSEM transients shows potential capability to perform classification, discrimi- nating background and non-dangerous items (clutter) data from, for instance, unex- ploded ordnance. Implementation of a merge SOM algorithm showed that clustering and classification of spatiotemporal CSEM data is possible. The ability to extract tar- get signals from a background-contaminated pattern is desired to avoid dealing with forward models containing subsurface response or to implement processing algorithm to remove, to some degree, the effects of background response and the target-host interactions.
View
... The role of pattern recognition methods such as machine learning and target feature extraction, as an alternative to inversion, is gaining rapid acceptance in areas such as unexploded ordnance (UXO) and landmine detection. An important early paper in classification of buried spheroids by their CSEM response is by Chesney et al. (1984). It would be interesting to explore whether such concepts can be applied to hydrogeophysical settings in which the subsurface target is not necessarily an isolated, well-defined man-made object but instead could be a subtle, finely-distributed and irregular variation in the subsurface electrical conductivity distribution. ...
Transient Em Induction
Chapter
Full-text available
  • Jan 2007
Landfill sites commonly use the space available in disused quarries or special purpose-built structures but not all past landfill operations were adequately controlled or documented such that the site boundaries, and the type and volume of fill are unknown in some old covered landfill sites. Even in controlled sites, the final form and depth extent of the landfill may not conform to those indicated in the original plan submitted to the regulatory authorities during the application for a site license. Thus, a significant amount of work is required in order to accurately define the relevant parameters of a covered landfill site. Our hydrogeophysical interest in landfill sites lies in assessing the pollution threat they pose since they may contain hazardous substances. In conventional geophysical investigation of landfill sites, the usual goals are to determine the geometrical characteristics (size and shape) of the repository and the physiochemical properties of the infill. Of the several non-invasive geophysical methods used in landfill studies, the electrical and electromagnetic (EM) methods are the most popular owing to their inherent ability to detect changes related to variations in fluid content, chemical composition and temperature in the subsurface, and the minimum capital and labor outlay required to use them in small-scale surveys (Whiteley and Jewell, 1992; Meju 2000). Since the presence of saline fluids in the ground enhances its ability to conduct electrical current, it is possible to locate a leachate plume by measuring the resistivity distribution in the subsurface. The main ground resistivity measurement techniques employed in landfill studies are the direct current (dc) resistivity and/or induced polarization (IP) methods (e.g.
View
... The role of pattern recognition methods such as machine learning and target feature extraction, as an alternative to inversion, is gaining rapid acceptance in areas such as unexploded ordnance (UXO) and landmine detection. An important early paper in classification of buried spheroids by their CSEM response is by Chesney et al. (1984). It would be interesting to explore whether such concepts can be applied to hydrogeophysical settings in which the subsurface target is not necessarily an isolated, well-defined man-made object but instead could be a subtle, finely-distributed and irregular variation in the subsurface electrical conductivity distribution. ...
View
... The design set for an object consisted of the feature vectors for all orientations (15' increments) at a given depth. A continuous parameter pattern classifier [19] , similar to the one used in magnetostatics , was able to classify the objects with a probability of misclassification of about 1% if the design and test sets were obtained for the same object depth. If the two sets were taken from depths differing by 0.1 m, the probability increased to about 11%. ...
View
... In near-surface geophysical applications, controlled-source electromagnetic (CSEM) methods are widely used because they are capable of detecting and mapping the distribution of fluids, large voids, faults, or lithological contacts (Nobes, 1996; Tezkan, 1999; Pellerin, 2002; Everett and Meju, 2003). The CSEM method also can detect and, to some extent, classify man-made targets such as pipes, underground structures, and unexploded ordnance (UXO) (Chesney et al., 1984; Qian and Boerner, 1995; Huang and Won, 2003). The method is sensitive to subsurface spatial variations in electrical conductivity . ...
Target signal enhancement in near-surface controlled-source electromagnetic data
Article
Full-text available
  • May 2005
Controlled-source electromagnetic-conductivity pro-files and maps were obtained on the Brazos Valley, Texas, floodplain to study the fractal statistics of ge-ologic noise and the effects of man-made conduc-tive metal targets. Fourier transform, discrete wavelet transforms, and variogram analyses were used. Targets tend to flatten the power-law power spectrum at small wavenumbers and shift power to higher wavenumbers. Detection and localization of targets can be achieved us-ing wavelet spectrogram techniques. Additionally, vari-ograms from pure background conductivity maps show a power-law trend for all lags, whereas, in the presence of targets, a short power-law trend is followed by a sill corresponding to a loss in spatial correlation. A sim-ple preprocessing step that combines responses from two perpendicular transmitter-receiver orientations en-hances the localization of targets and rejects back-ground signals in profiles and 2D apparent-conductivity maps. Finally, a field example shows how the use of wavelet filtering is able to separate target responses from the geologic background.
View
... The role of pattern recognition methods such as machine learning and target feature extraction, as an alternative to inversion, is gaining rapid acceptance in areas such as unexploded ordnance (UXO) and landmine detection. An important early paper in classification of buried spheroids by their CSEM response is by Chesney et al. (1984). It would be interesting to explore whether such concepts can be applied to hydrogeophysical settings in which the subsurface target is not necessarily an isolated, well-defined man-made object but instead could be a subtle, finely-distributed and irregular variation in the subsurface electrical conductivity distribution. ...
Near-Surface Controlled-Source Electromagnetic Induction
Chapter
Full-text available
  • Jan 2005
The controlled-source electromagnetic (CSEM) induction method is emerging as a leading geophysical technique in hydrogeological studies. However, the technique is quite often misunderstood compared to other common techniques of applied geophysics: namely, seismic reflection and refraction, magnetics, gravity, and ground-penetrating radar (GPR). In this chapter we review the fundamental physical principles behind the CSEM prospecting technique, with emphasis on near-surface applications, and present some recent advances in this field that have been made by the authors. CSEM methods are defined here to be those in which the experimenter has knowledge of and control over the electromagnetic field transmitted into the ground and hence excludes magnetotellurics, related natural-source methods, and the various uncontrolled-source methods involving, for example, radio transmissions.
View
Wideband measurement of the magnetic susceptibility of soils and the magnetic polarizability of metallic objects
Conference Paper
  • Jul 2012
Broadband electromagnetic induction (EMI) sensors have been shown to be able to reduce false alarm rates and increase the probability of detecting landmines. To aid in the development of these sensors and associated detection algorithms, a testing facility and inversion technique have been developed to measure the magnetic response of soils and the magnetic polarizability of metal targets. When theoretical predictions are possible, these measurements show good agreement with theory.
View
Statistical classification of buried objects from spatially sampled time or frequency domain electromagnetic induction data
Article
  • Aug 2004
Methods for classifying objects based on spatially sampled electromagnetic induction data taken in the time or frequency domain are developed and analyzed. To deal with nuisance parameters associated with the position of the object relative to the sensor as well as the object orientation, a computationally tractable physical model explicit in these unknowns is developed. The model is also parameterized by a collection of decay constants (or equivalently Laplace-plane poles) whose values in theory are independent of object position and orientation. These poles are used as features for classification. The overall algorithm consists of two stages. First, we estimate the values of the unknown parameters and then we do classification. Classification is done by comparing either the raw data or some low-dimensional collection of features extracted from the data to entries in a library. The library can be constructed using either simulated or calibration data. A maximum likelihood method is developed and analyzed for the problem of joint pole, location, and orientation parameter determination. Here we examine and compare two classification schemes. The first classification method is based on data residuals generated from estimated signal parameters. This scheme performs well in low SNR cases. The second is based on estimated pole values themselves, which performs well in high SNR cases. We validate our methods on both simulated and field data taken from frequency and time domain sensors.
View
Canadian landmine detection research program
Article
  • Sep 2003
  • Proceedings of SPIE
Defence R&D Canada (DRDC), an agency within the Department of National Defence, has been conducting research and development (R&D) on the detection of landmines for countermine operations and of unexploded ordnance (UXO) for range clearance since 1975. The Canadian Centre for Mine Action Technologies (CCMAT), located at DRDC Suffield, was formed in 1998 to carry out R&D related to humanitarian demining. The lead group responsible for formulating and executing both countermine and humanitarian R&D programs in detection is the Threat Detection Group at DRDC Suffield. This paper describes R&D for both programs under the major headings of remote minefield detection, close-in scanning detection, confirmation detection and teleoperated systems. Among DRDC's achievements in landmine and UXO detection R&D are pioneering work in electromagnetic and magnetic identification and classification; the first military-fielded multisensor, teleoperated vehicle-mounted landmine detection system; pioneering use of confirmation detectors for multisensor landmine detection systems; the first fielded thermal neutron activation landmine confirmation sensor; the first detection of landmines using a real-time hyperspectral imager; electrical impedance imaging detection of landmines and UXO and a unique neutron backscatter landmine imager.
View
LEARNING IN STATISTICAL PATTERN RECOGNITION.
Article
  • Jan 1980
After a brief review of the basic principles of machine learning, the recent progress is examined. The interrelationships among learning algorithms are considered. Examples are drawn from seismic data study. The basic problem of learning sample size versus dimensionality and other special topics are also reviewed.
View
View
View
View
Statistical and Structural 12 Approaches to Texture
Article
  • Jan 1978
  • PHOTOGRAMM ENG REM S
A review is presented of the image processing literature on the various approaches and models investigators have used for textures. These include statistical approaches of autocorrelation function, optical transforms, digital transforms, textural edgeness, structural element, gray tone co-occurrence, run lengths, and auto-regressive models. A discussion and generalization is presented of some structural approaches to texture based on more complex primitives than gray tone. Some structural-statistical generalizations which apply the statistical techniques to the structural primitives are given.
View
View
An Introduction to Long Memory Time Series Models and Fractional Differencing
Article
  • Jun 2008
  • J TIME SER ANAL
The idea of fractional differencing is introduced in terms of the infinite filter that corresponds to the expansion of (1-B)d. When the filter is applied to white noise, a class of time series is generated with distinctive properties, particularly in the very low frequencies and provides potentially useful long-memory forecasting properties. Such models are shown to possibly arise from aggregation of independent components. Generation and estimation of these models are considered and applications on generated and real data presented.
View
Spatial Interaction and the Statistical Analysis of Lattice Systems
Article
  • Jan 1974
The formulation of conditional probability models for finite systems of spatially interacting random variables is examined. A simple alternative proof of the Hammersley–Clifford theorem is presented and the theorem is then used to construct specific spatial schemes on and off the lattice. Particular emphasis is placed upon practical applications of the models in plant ecology when the variates are binary or Gaussian. Some aspects of infinite lattice Gaussian processes are discussed. Methods of statistical analysis for lattice schemes are proposed, including a very flexible coding technique. The methods are illustrated by two numerical examples. It is maintained throughout that the conditional probability approach to the specification and analysis of spatial interaction is more attractive than the alternative joint probability approach.
View
Experimental time domain electromagnetic induction system
Article
  • Jul 1984
An automated system which measures, coherently averages, and stores time domain electromagnetic induction responses of conductive objects and allows on‐line analysis of the data is described. The sensor portion of the system consists of a set of coils, transmit and receive circuitry situated in a remote nonmetallic laboratory. The circuitry is controlled by a specialized high‐speed data‐acquisition system which communicates, through a microcomputer, with a distant mainframe computer. Advanced inductive load switching techniques are employed and errors due to temperature variations in the coils are minimized by suitable sensor design. The user can change date‐collection parameters, such as pulse length, dwell time, and transmitter waveform shape, under program control to tailor the system to a particular application. The design also allows for easy replacement of the sensor elements; i.e., coils, receive and transmit circuitry as different applications require. This flexibility allows the system to be readily adapted for research in a number of areas that rely on measurement of electromagnetic induction responses. A typical response may be obtained in less than 1 min. The minimum time resolution of the system is 4 μs. Noise levels of 3 to 5 mV standard deviation are obtainable for 500 averages. Data are presented to demonstrate the capabilities of the system.
View
Additive estimators for probabilities of correct classification
Article
  • Dec 1978
  • PATTERN RECOGN
Several methods for estimating a sample-based discriminant's probability of correct classification are compared with respect to bias, variance, robustness, and computation cost. “Smooth” modification of the counting estimator, or sample success proportion, is recommended to reduce bias and variance while retaining robustness. Also the “bootstrap” method of Efron(8) can approximately correct an additive estimator's bias using an ancillary computer simulation. In contrast, bias reduction achieved by the popular “leave-one-out” modification of counting method is vitiated by corresponding increase in variance.
View