A comparison of algorithms for subsurface target detection and identification using time-domain electromagnetic induction data

Tantum, Stacy L.; Collins, Leslie M.

doi:10.1109/36.927453

Cited by 56 publications

(42 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As is evident from Fig. 1 and, indeed, as we have found to be true in general, such an assumption is not, in fact, warranted in practice [16].…”

Section: Introductionmentioning

confidence: 70%

“…The fact that our approach is to be able to localize and classify the object at the same time adds more to its value, while most classification approaches assume that geometrical information of the object is available at the classification stage: an assumption which is not valid for the real field data scenarios. For example, in [16], a very simple dipole model is used as the basis of processing. Classification results using the simulations for simple targets for the case of unknown coefficient factors at an SNR of about 40 dB shows the probability of correct classification of about 50%.…”

Section: Classification Resultsmentioning

confidence: 99%

“…It is known that if the model parameters (such as poles in our case) are sufficient statistics [34], then the optimal classification can be done in a lower dimensional feature space. Unfortunately, for the UXO, problem poles are not sufficient statistics [16] and, therefore, do not lead to optimal classifiers. Despite this theoretical shortcoming, these parameters are still widely used as the basis for solving the UXO classification problem due to convenience and the success of such methods in practice where other issues (such as the inaccuracies of the underlying dipole model) become important.…”

Section: B Optimal Classificationmentioning

confidence: 99%

“…These estimated parameters are then input to a classifier to determine the type of target under investigation. In many cases [11], [16], underlying both of these processing steps is an implicit assumption that the object-specific parameter values are independent of the orientation and location of the target. A broad array of modeling approaches is available that meets these criteria, from exact numerical models based on finite element and finite difference approaches to models that approximate the buried object as a magnetic dipole.…”

Section: Introductionmentioning

confidence: 99%

“…A broad array of modeling approaches is available that meets these criteria, from exact numerical models based on finite element and finite difference approaches to models that approximate the buried object as a magnetic dipole. We have chosen to implement the latter "dipole model," which is a decision motivated by the practical need for rapid data processing and by the ubiquity of its use in the UXO detection and classification literature [16]- [22]. We note that more accurate models of the physics of UXO-like targets are available, and in some sense, the dipole model could be considered the first-order application of the techniques developed in this paper, which are extensible to any parameterized model representing UXO.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Statistical Classification of Buried Unexploded Ordnance Using Nonparametric Prior Models

Aliamiri¹,

Stalnaker

Miller³

2007

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Abstract-We used kernel density estimation (KDE) methods to build a priori probability density functions (pdfs) for the vector of features that are used to classify unexploded ordnance items given electromagnetic-induction sensor data. This a priori information is then used to develop a new suite of estimation and classification algorithms. As opposed to the commonly used maximumlikelihood parameter estimation methods, here we employ a maximum a posteriori (MAP) estimation algorithm that makes use of KDE-generated pdfs. Similarly, we use KDE priors to develop a suite of classification schemes operating in both "feature" space as well as "signal/data" space. In terms of feature-based methods, we construct a support vector machine classifier and its extension to support M -ary classification. The KDE pdfs are also used to synthesize a MAP feature-based classifier. To address the numerical challenges associated with the optimal data-space Bayesian classifier, we have used several approximation techniques, including Laplacian approximation and generalized likelihood ratio tests employing the priors. Using both simulations and real field data, we observe a significant improvement in classification performance due to the use of the KDE-based prior models.

show abstract

“…As is evident from Fig. 1 and, indeed, as we have found to be true in general, such an assumption is not, in fact, warranted in practice [16].…”

Section: Introductionmentioning

confidence: 70%

Section: Classification Resultsmentioning

confidence: 99%

Section: B Optimal Classificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Statistical Classification of Buried Unexploded Ordnance Using Nonparametric Prior Models

Aliamiri¹,

Stalnaker

Miller³

2007

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

Transient electromagnetic inversion based on particle swarm optimization and differential evolution algorithm

Zhuang

et al. 2020

Near Surface Geophysics

View full text Add to dashboard Cite

For transient electromagnetic inversion, a gradient‐based algorithm is strongly dependent on the quality of the initial model, while any non‐gradient‐based algorithm often falls too easily into local optima. This paper proposes a joint differential‐evolution–particle‐swarm‐optimization inversion algorithm, which provides a better global optimization. A dual‐population evolution strategy and information exchange mechanism is presented. For verification, this is followed by adoption of a layered inversion model in the transient electromagnetic inversion with a central loop. The results show that the differential‐evolution–particle‐swarm‐optimization joint algorithm can reduce the probability of a premature phenomenon (i.e. falling into local optima) and improve the inversion accuracy, efficiency and stability, with a fast convergence occuring in the early stages. Furthermore, the proposed algorithm has a higher degree of fitting (prediction ability) for data inversion and is feasible for transient electromagnetic inversion.

show abstract

Modeling position error probability density functions for statistical inversions using a Goff–Jordan rough surface model

Tantum

Zhu

Torrione

et al. 2008

Stoch Environ Res Risk Assess

View full text Add to dashboard Cite

Buried unexploded ordnance (UXO) continues to be a difficult remediation problem from both a sensing and a discrimination point of view. Modern approaches to both the sensing and discrimination problems utilize high bandwidth electromagnetic induction (EMI) sensors to collect geo-referenced data which is then inverted, or fit, using a forward model in order to obtain features that can be directly interpreted using the physics associated with electromagnetic induction-based sensing. These features are then used in a variety of classification architectures. One aspect of this process that has captured recent interest is that uncertainty in the positions at which data was collected can degrade the inversion performance and thus the subsequent classification. Several mechanisms to address this issue have been explored that range from filtering and prediction of actual positions to exploiting Bayesian approaches for uncertainty mitigation. In the Bayesian approach, a statistical model of the position errors is used as a prior for integrating over the uncertainty in the inversion process. In this study, we demonstrate that errors in the statistical priors used in this process can negatively impact subsequent classification performance, thus highlighting the need for an accurate statistical model for the position errors. Next, we propose a mechanism by which to obtain such models. Specifically, we utilize a Goff-Jordan rough surface model and simulate the sensor data collection system motion over the simulated ground or ocean surfaces to calculate errors and generate statistical models. Our results suggest that this approach can be used to develop the statistical models necessary for mitigating uncertain position information.

show abstract

A comparison of algorithms for subsurface target detection and identification using time-domain electromagnetic induction data

Cited by 56 publications

References 7 publications

Statistical Classification of Buried Unexploded Ordnance Using Nonparametric Prior Models

Statistical Classification of Buried Unexploded Ordnance Using Nonparametric Prior Models

Transient electromagnetic inversion based on particle swarm optimization and differential evolution algorithm

Modeling position error probability density functions for statistical inversions using a Goff–Jordan rough surface model

Contact Info

Product

Resources

About