Ground Penetrating Radar Signal Research Articles

Fusion of ground penetrating radar and laser scanning for infrastructure mapping

Abstract Mobile mapping vehicles, equipped with cameras, laser scanners (in this paper referred to as light detection and ranging, LiDAR), and positioning systems are limited to acquiring surface data. However, in this paper, a method to fuse both LiDAR and 3D ground penetrating radar (GPR) data into consistent georeferenced point clouds is presented, allowing imaging both the surface and subsurface. Objects such as pipes, cables, and wall structures are made visible as point clouds by thresholding the GPR signal’s Hilbert envelope. The results are verified with existing utility maps. Varying soil conditions, clutter, and noise complicate a fully automatized approach. Topographic correction of the GPR data, by using the LiDAR data, ensures a consistent ground height. Moreover, this work shows that the LiDAR point cloud, as a reference, increases the interpretability of GPR data and allows measuring distances between above ground and subsurface structures.

Improved Mask R-CNN with distance guided intersection over union for GPR signature detection and segmentation

Ground penetrating radar (GPR) has been used for non-destructive inspection of civil infrastructure systems such as bridges and pipelines. Manually extracting useful data from a large amount of non-intuitive GPR scans is tedious and error-prone. To address this challenge, a generalizable end-to-end framework is developed and implemented to simultaneously detect and segment object signatures in GPR scans. The proposed approach improves the Mask Region-based Convolutional Neural Network (R-CNN) by incorporating a novel distance guided intersection over union (DGIoU) as a new loss function for detection and segmentation. The DGIoU considers the center distance between two bounding boxes and overcomes the weakness of intersection over union (IoU) in training and evaluation. In addition, a new method is proposed to extract data points from the segmented mask patches containing both object signatures and background noises. The extracted data points can be further processed for object localization and characterization. Experiments were conducted using GPR scans collected from a concrete bridge deck. The hyperbolic signatures of rebars can be accurately detected and segmented using the proposed method. It was demonstrated that using DGIoU improves the regression effect of bounding box and mask. The improved Mask R-CNN achieved an average accuracy (AP) of 58.64% and 47.64% for the detection and segmentation task, respectively.

1D waveform inversion of GPR trace by particle swarm optimization

This study investigates the identification of buried materials from ground-penetrating radar (GPR) signals by applying Particle Swarm Optimization (PSO) method for the inversion of individual traces. Interpretation of GPR data can be carried out either by applying some basic processing steps to raw data or by inverting GPR data. However, for obtaining more specific information about the buried object, some more advanced algorithms are required. In this study, the PSO, which has been used to solve many engineering problems, is utilized to invert the GPR traces for detecting various types of buried plastic pipe fillings. Firstly, we confirmed the applicability of PSO to GPR traces using numerical modeling. Then, the data acquired over buried pipe filling materials (air, water ice) were evaluated experimentally by the proposed method. Determination of the filling material is one of the most popular problems of shallow pipe exploration by GPR. The inversion process resulted in good data fit between observed and calculated traces for air-filled, water-filled, and ice-filled pipe. The estimated dielectric properties including relative dielectric permittivity, conductivity, and relative magnetic permeability values were in the range of the corresponding parameter values in the literature at the end of inversion for each model. These results show that the PSO method applied to GPR data can distinguish plastic pipe contents. Moreover, the results suggest that PSO can be used for the quantitative interpretation of GPR data.

Ground-Penetrating Radar Study of Progradational Units in Holocene Coastal Plains: Carchuna Beach (SE Spain)

The internal geometries of progradational units on the coastal plain in southeastern Spain (Carchuna Beach, Granada) since the maximum Holocene marine transgression (ca. 7000 cal BP) have been analysed using ground-penetrating radar (GPR). Radargram analysis shows the presence of sedimentary structures consistent with the progradation of coastal plains, interpreted as laminations associated with the foreset and topset. Laterally, five middle scale progradational units (H2, H3, H4, H5 and H6) are identified. Each of these units can be grouped into two major units. The general trend towards progradation is interrupted at the boundary of the units. Minor scale units (prograding wedges), deposited at topographically lower areas, are detected in GPR profiles. The radargrams also show channel erosional surfaces and sigmoidal landward dipping reflections interpreted as washover channel deposits, with the erosion and transference of sediments between the beach and the back-ridge domain. Furthermore, the saturation of coastline sediments with saltwater rapidly attenuates the GPR signal. Thus, the GPR is also a good tool to detect saltwater intrusions in shallow aquifers.

Prospecting and Evaluation of Underground Massive Ice by Ground-Penetrating Radar

Data from geocryological studies of soil and rock massifs in permafrost zone are very important as a basis for predicting possible negative consequences associated with climate change. A promising technique for studying geocryological structures (various types of underground ice) is the ground-penetrating radar (GPR) method. This paper presents the applications of the GPR method to prospect and evaluate massive ice in a frozen rock mass. To study the features of GPR signals received during sounding of underground ice, a model of a single GPR trace for the structure “frozen rock-ice-frozen rock” was developed. As a result, regularities were established in the kinematic and dynamic characteristics of GPR signals at the upper and lower boundaries of massive ice, depending on its geometric parameters. The established features were confirmed by the results of computer and physical simulation of GPR measurements of a frozen rock mass model. The main result of the study was to obtain a set of criteria for identifying massive ice according to GPR measurements. The developed criteria will allow the use of GPR for a detailed study of the structure of permafrost rocks to prevent the development of dangerous cryogenic processes in undisturbed and urban areas of the Arctic.

Different Characteristics of Radar Signal Attenuation Depending on Concrete Condition of Bare Bridge Deck

This paper examined the attenuation method, which is used abroad, to derive a method to evaluate the condition of bare concrete bridge decks using ground penetrating radar (GPR) technology. For this purpose, 12 GPR surveys of 10 bridge decks in public service from the beginning of its service to 25 years of service were carried out to examine the attenuation characteristics of GPR signals under various concrete conditions. The survey revealed the signal below the top rebar of concrete bridge deck under the condition of using de-icing chlorides for snow removal was not clear. Therefore, using the receiving wave signal from top rebar was reasonable when attenuation of GPR signal was to be used for bridge deck evaluation. Examining the signal attenuation by the condition of concrete, dispersion of the attenuated signal was overall large in the initial performance period. However, since this is not due to deterioration, exclusion of the large dispersion from the evaluation of bridge deck condition was desirable. The attenuation size was linearly proportional to its two way travel time (signal transmission time), i.e., the depth of the top rebar, in a sound bridge deck. Also, the dispersion of the attenuation was small and symmetric to the linear regression line. If the bridge deck was maintained adequately, the linear regression correlation was also maintained similarly for the next several years. However, if deterioration occurred to be accompanied by increased attenuation and dispersion, correlation coefficient of the linear regression line declined. Nonetheless, the upper value of the attenuation representing sound concrete condition was still tended to be linearly proportional to two way travel time. Since the attenuation quantity could result in considerable debilitation just by the construction error, it is necessary to deal with the depth-error of top rebar for evaluating the concrete condition by using the GPR signal attenuation. Calibration was carried out by deriving a linear regression line for the signal two way travel time and the upper 90th percentile values of the attenuation obtained from the top bar position of the bridge and then removing it from the total attenuation.

On the Use of Lateral Wave for the Interlayer Debonding Detecting in an Asphalt Airport Pavement Using a Multistatic GPR System

In this article, we focus on the detection of the interlayer debonding of the asphalt airport pavement by the ground-penetrating radar (GPR) system. Since the interlayer debonding usually occurs in the shallow region of the asphalt airport pavement (several centimeters), it is difficult to interpret the anomalies or the defects from the GPR signals composed of many waves under the boundary conditions. Moreover, the wavelength of the ordinary GPR system is over several centimeters. Therefore, the spatial resolution of the system is not accurate enough to consider the millimeter thickness of the debonding layer. To overcome these problems, we propose a new method based on evaluating the lateral wave behavior of common midpoint (CMP) gathers collected by a multiple static GPR system. The multistatic GPR system is a stepped frequency continuous wave (SFCW) radar system, which consists of eight transmitting and eight receiving bowtie antennas. The system operates in the frequency range from 50 MHz to 1.5 GHz. After the validation of the simulation, the results of the interlayer debonding detection were evaluated by a field experiment obtained at Tokyo International Airport. The proposed method can detect the debonding layers which are less than 1 mm. Also, it is shown that our proposed method has a high consistency with the conventional acoustic finding method in the field measurement. It provides an innovative and effective method for the interlayer debonding detection of a partially damaged airport asphalt pavement, which is difficult to be observed by the ordinary GPR signals.

Underground Object Size Approximation using GPR Signal and Image Processing

This paper comprises a step wise method of approximating the size of an underground object using GPR (Ground Penetrating Radar). It involves more than just using predefined filters and techniques. Usage of Trivial method of mathematics to calculate the top surface dimensions of the buried objects is the main purpose of this paper. Problem that is faced that, only the presence of any object can be known using the GPR resource, but not exactly how to derive the size of the object using the same data. This method consists of a dual approach to the problem to make sure that the data that is being given out is accurate. The objectives of this paper are to use the GPR to calculate the top surface dimension of a buried object at a suitable depth according to the frequency. The steps that are incorporated include pre-processing of raw data, determination of ROI (Region of interest) from the pre-processed data, Application of appropriate filters for image processing and estimating surface area and depth of the concealed object. The main reason of this paper is to serve the purpose of detecting what is under the ground in a quick and simpler way using the algorithm proposed.

Comment on: “Fast 3D inversion of gravity data using solution space priorconditioned Lanczos bidiagonalization”, by Mohammad Rezaie, Ail Moradzadeh and Ali Nejati Kalateh

To reduce the shielding effect of strong jamming signals such as direct waves and multiple echoes on the effective signals in ground penetrating radar (GPR) images and effectively improve the GPR image resolution, a method to remove strong interference signals in GPR that combines the wavelet transform and F-K filtering is proposed. First, we choose a biorthogonal wavelet basis that matches the waveform of GPR signal, and an original GPR image is decomposed into subimages of different frequency bands based on the optimal wavelet basis. Second, according to the difference in the distribution of the effective signals and the strong interference of different subsignal images in the wavenumber domain, different bandpass filters are designed to accurately separate the effective signals and strong interference, and the subimages are reconstructed after the interference being removed. The processing of the forward simulated and measured GPR images shows that the combined method of wavelet transform and F-K filtering can produce different F-K filters according to the characteristics of different subsignal images and accurately process the image, suppressing the strong interference and retaining the effective signals to the greatest extent. Compared to the results of the predictive deconvolution method and the conventional F-K method, the cavity reflection in the image processed by the proposed method is more obvious, the coaxial signal is continuous and clear, and the image resolution is improved, which proves the feasibility and effectiveness of this method in separating strong interference waves.

A Practical Approach to Condition Assessment of Asphalt-Covered Concrete Bridge Decks on Korean Expressways by Dielectric Constant Measurements Using Air-Coupled GPR.

The main objectives of this study are to investigate the variations of the dielectric constant of concrete on Korean expressways by using a 1 GHz air-coupled Ground Penetrating Radar (GPR) system and to develop a practical approach to the condition assessment of concrete bridge decks with asphalt overlay on Korean expressways by dielectric constant measurements. A total of 684 GPR investigations of 601 actual concrete bridge decks, which are in service between 2 and 43 years, were carried out during the period between 1999 and 2013. Statistical analysis revealed that the dielectric constant of asphalt-covered concrete bridge decks reduced with service age and this trend continued until service age of over 40 years. As a result, this study provides a practical dielectric constant curve that could be used for condition evaluation of top concrete in asphalt-covered bridge decks with consideration of concrete age. Based on regression analyses of the GPR field survey data and experiences through the field survey, a double cut-off dielectric constant criterion was proposed for condition assessment of asphalt-covered concrete bridge decks on Korean expressways. In addition, a GPR field survey was performed at an actual bridge on the Yeongdong expressway in Korea to test the proposed GPR signal interpretation method. The field survey results provide fundamental data to better understand the variation of the dielectric constant of concrete in actual bridges with asphalt overlay and to develop a practical approach to condition assessment of asphalt-covered concrete bridge decks on Korean expressways by dielectric constant measurements using air-coupled GPR.

Use of Ground Penetrating Radar, Hydrogeochemical Testing, and Aquifer Characterization to Establish Shallow Groundwater Supply to the Rehabilitated Ni-les’tun Unit Floodplain: Bandon Marsh, Coquille Estuary, Oregon, USA

Fluvial-tidal wetlands in the Ni-les’tun Unit (~200 hectares) of the Bandon Marsh, Coquille Estuary, Oregon, were analyzed for shallow aquifer conditions that could influence surface water-qualities in reconstructed marsh, pond, and discharge/tidal channels. The wetlands were surveyed for pre-historic channel features, depth to groundwater surface (GWS), and subsurface salinity intrusion by ground penetrating radar (GPR) in 50 profiles, totaling 11.1 km in track line distance. Only small flood-discharge/tidal channel features (<10 m width and 1–2 m depth) were recorded in the interior floodplain areas. GWS reflections were observed at 0.5–2.0 depth, where the GPR signal was not obscured by localized salinity intrusion (~0.5 km landward distance) from the adjacent Coquille Estuary channel. Top-sealed piezometers (1.5–2.0 m depth) were installed at 10 sites, where in-situ groundwaters were monitored for temperature (8.5–16.5° C), conductivity (<100–18,800 μS cm-1), and pH (2.5–7.8) on a seasonal basis. Dissolved oxygen was semi-quantitatively measured (ChemSticks) at some sites, and all sites were monitored (fall, winter, summer) for GWS level. Low dissolved oxygen (DO <1 ppm) at four sites was of particular concern for potential discharge into small channels that were to be constructed for juvenile salmonid nursery habitat. The horizontal and vertical asymmetries of conductivity (salinity), used as a conservative groundwater source tracer, and measured GWS elevation trends (gradients) led to a four-part flow model for shallow groundwater supply in the Ni-les’tun floodplain. Freshwater supplied, in part, by hillslope discharge contributes to low pH and low DO water quality in the shallow aquifer. Saline water, supplied by subsurface salinity intrusion and evaporative capillary rise, could introduce salinity toxicity to isolated (stagnant) surface ponds. Following construction of a dense channel network (2009–2011) by the Bandon Marsh National Wildlife Refuge, selected Ni-les’tun channel waters (13 sites) were monitored (2011-2012) for resulting water-quality. The tidally-connected channels generally showed improved water-quality relative to groundwater in some nearby piezometer sites. However, low-quality groundwater supply compromised some channel reaches (DO ~2.0–4.7 ppm) that depended on groundwater recharge from hillslope discharge during either summer or winter conditions.

Characterization of the Internal Structure of Landmines Using Ground-Penetrating Radar

One of the principal limitations of employing a ground-penetrating radar (GPR) for landmine detection is the presence of clutter, i.e., reflections from the surrounding environment, which might interfere with the landmine echoes. Clutter presents similar scattering characteristics as typical targets and may significantly raise the detection threshold of the system. The capability to characterize the internal structure of a buried target might provide key unique information to develop advanced landmine-clutter discrimination algorithms, considering that the presence of internal scattering components can be univocally associated with man-made targets. In this letter, the possibility of identifying and characterizing these contributions from the GPR signature of a landmine is numerically assessed and experimentally validated. The simulated response from a landmine-like target shows that the presence of an internal structure generates additional reflection peaks, as a consequence of the layered structure of the object, and the field trials corroborate that it is possible to identify these scattering components and delineate their spatial distribution.

GPR pattern recognition of shallow subsurface air voids

Countless subsurface voids in urban areas of cities threaten people’s lives and property. A workflow for automatically identifying subsurface voids from ground penetrating radar (GPR) data was developed in this study. The workflow consists of 3 stages: locating voids automatically from C-scans, then verifying voids from corresponding B-scans, and finally making judgements based upon the previous 2 sets of results. This study adopted 2 (Lai et al., 2016) approaches: approach 1 quantified the GPR response of air voids using forward modelling, while approach 2 used workflow prototyping and validation with inverse modelling. Forward simulations indicated that different ratios of void size to GPR signal footprint could result in a variety of patterns in B-scans: they can be hyperbolas, cross patterns, bowl shaped patterns and reverberations. With a database of void patterns of both C-scans and B-scans established, in approach 2 the workflow uses a pyramid pattern recognition method – with pixel value or gradient being used for feature identification – to search automatically for air-filled void responses in GPR data. The workflow was tested using 2 laboratory and field experiments and the results were promising. The constraint values proposed by the 2 experiments were validated with another site experiment. Given the huge workload involved in city-scale subsurface health inspections, a standardized workflow can help improve efficiency and effectiveness of subsurface void identification.

Object Detection of Ground-Penetrating Radar Signals Using Empirical Mode Decomposition and Dynamic Time Warping

An object detection method of ground-penetrating radar (GPR) signals using empirical mode decomposition (EMD) and dynamic time warping (DTW) is proposed in this study. Two groups of timber specimens were examined. The first group comprised of Douglas fir (Pseudotsuga menziesii) timber sections prepared in the laboratory with inserts of known internal characteristics. The second group comprised of timber girders salvaged from the timber bridges on historic Route 66 over 80 years. A GSSI Subsurface Interface Radar (SIR) System 4000 with a 2 GHz palm antenna was used to scan these two groups of specimens. GPR sensed differences in dielectric constants (DC) along the scan path caused by the presence of water, metal, or air within the wood. This study focuses on the feature identification and defect classification. The results show that the processing methods were efficient for the illustration of GPR information.

In-situ recognition of moisture damage in bridge deck asphalt pavement with time-frequency features of GPR signal

A complete solution, including an effective non-destructive evaluation (NDE) method and an automatic recognition model, was provided for the rapid diagnosis of moisture damage in the asphalt pavement by using ground-penetrating radar (GPR) signals. A ground-coupled 2.3 GHz antenna was used to conduct a GPR survey on an asphalt pavement of a bridge deck, where the moisture damage areas were detected and visually recognized in processed GPR B-scan images and further validated in subsequent pavement coring. Field GPR traces of the asphalt layer were read and classified to build a dataset which included 8215 moisture damage and 8215 normal pavement traces. A 28-element time-frequency feature vector was extracted and further reduced to an 11-element sensitive feature vector via the linear discriminant analysis (LDA) method. Principal component analysis (PCA) was adopted to decompose the feature vector into the PCs (principal components), which was used to train a BP-ANN model. The result indicates the high accuracy of the ANN model with sensitive feature vectors, i.e., 95.3% for normal and 92.4% for moisture classification. Finally, the ANN model was used to evaluate the GPR survey data, and its result is consistent with the GPR B-scan feature. These findings suggest that the ground-coupled GPR system with 2.3 GHz antenna and the recognition model will enable an innovative quality evaluation system for asphalt pavement.

Mapping the Physical and Dielectric Properties of Layered Soil Using Short-Time Matrix Pencil Method-Based Ground-Penetrating Radar

This paper presents the mapping of the physical and dielectric properties of layered soil by using ground-penetrating radar (GPR). Poles extracted from the GPR signals by using the short-time matrix pencil method, along with preprocessing, including filtering and antenna calibration, were employed in order to map the physical and dielectric properties of the layered soil. With the proposed system, the conjugate gradient method was also introduced to solve the time-domain inverse problem faced in the antenna calibration. Experimentations were conducted on four different days in Nakhon Ratchasima province in Thailand, which is a potential area for the occurrence of hardpan. These were done based on the hypothesis that the soil properties, such as water content and dielectric constant, should change when the experimentation day changes in spite of the experimentations in the same area. In the experimentations, soil samples were collected using a core method, and their physical and dielectric properties were measured by using a standard laboratory method and a commercial dielectric probe kit. The measured soil properties and extracted poles for each experiment and each soil layer are shown and analyzed. The results indicate that the real part of the dielectric constants, strongly related to water content, can be mapped using extracted natural frequency. In order to map the dielectric properties, i.e. the water content and dielectric constants of the layered soil, the experimental results were fitted to $2^{\mathrm {nd}}$ -order polynomial curve. The largest regression value of the fitted curve was 0.9994. The bulk density, which is a physical property of the soil, distinguishes the soil type here differentiated by different soil layers. According to the experimental results, the hardpan occurred at the second soil layer because its bulk density was higher than 1.8 g/cm 3 . The bulk density was mapped by extracted poles, including damping factors and natural frequencies.

Estimation of sand water content using GPR combined time-frequency analysis in the Ordos Basin, China

Abstract Groundwater is the key factor of determining the growth of vegetation. Identifying the characteristics of groundwater is an important basis to formulate a management plan for water resources and develop the technology of controlling desertification in arid areas scientifically. It is also important to the environmental protection in China. Ground penetrating radar (GPR) signals produce a special response to the changes in water content during propagation, thus it is essential to study the propagation of electromagnetic pulse in aeolian sand with different water content. The GPR tests of aeolian sand samples with different water content were conducted by a GPR system, dielectric constant meter, and conductivity meter. The temporal and frequency spectral characteristics of electromagnetic signals measured from aeolian sand samples were analyzed. The results show that the dielectric constant and conductivity of aeolian sand increase nonlinearly as the water content increases, and the attenuation coefficient of electromagnetic pulses increases parabolically. Meanwhile, the temporal waveform fluctuates significantly at the boundary of different media, and its two-way travel time increases nonlinearly as the water content increases, the pulse velocity decreases nonlinearly. Furthermore, the peak frequency of the spectrum for a signal propagating through aeolian sand decreases as the water content increases. The peak frequency is concentrated in the 1000 to 1400 MHz range, but the shape and bandwidth of the spectrum are less affected by water content. The above-mentioned correlations can provide a water content estimation of aeolian sand with direct value to the local authorities which are involved in the decision-making process for effective water management in arid and semi-arid area.

Modeling of EM Wave Coherent Scattering From a Rough Multilayered Medium With the Scalar Kirchhoff Approximation for GPR Applications

This article presents a new asymptotic modeling of electromagnetic (EM) wave coherent scattering from a rough multilayered medium, based on the scalar Kirchhoff-tangent plane approximation. The proposed EM model is developed to simulate a realistic ground-penetrating radar (GPR) signal that considers the interface roughness of the multilayer. It allows us to investigate the influence of the interface roughness on the amplitude of the GPR echoes coming from the multilayered medium. Sounded multilayered medium generally has a low contrast between the successive layers, so that the multiple reflections inside each layer may be neglected; this assumption will be evaluated. The very low computational burden of this EM method is an important advantage as compared with a rigorous numerical method. First, numerical results in the frequency domain are presented to validate the proposed model, by comparison with a reference method based on the Method of Moments (MoM). Then, numerical results in the time domain are presented to analyze the behavior and performance of this new method, and the impact of both the interface roughness and the medium conductivity on the results.

Lab Non Destructive Test to Analyze the Effect of Corrosion on Ground Penetrating Radar Scans

Corrosion is a significant damage in many reinforced concrete structures, mainly in coastal areas. The oxidation of embedded iron or steel elements degrades rebar, producing a porous layer not adhered to the metallic surface. This process could completely destroy rebar. In addition, the concrete around the metallic targets is also damaged, and a dense grid of fissures appears around the oxidized elements. The evaluation of corrosion is difficult in early stages, because damage is usually hidden. Non-destructive testing measurements, based on non-destructive testing (NDT) electric and magnetic surveys, could detect damage as consequence of corrosion. The work presented in this paper is based in several laboratory tests, which are centered in defining the effect of different corrosion stage on ground penetrating radar (GPR) signals. The analysis focuses on the evaluation of the reflected wave amplitude and its behavior. The results indicated that an accurate analysis of amplitude decay and intensity could most likely reveal an approach to the state of degradation of the embedded metallic targets because GPR images exhibit characteristics that depend on the effects of the oxidized rebar and the damaged concrete. These characteristics could be detected and measured in some cases. One important feature is referred to as the reflected wave amplitude. In the case of corroded targets, this amplitude is lower than in the case of reflection on non-oxidized surfaces. Additionally, in some cases, a blurred image appears related to high corrosion. The results of the tests highlight the higher amplitude decay of the cases of specimens with corroded elements.

Application of Polarimetric Features and Support Vector Machines for Classification of Improvised Explosive Devices

In this letter, we present a binary classification problem of improvised explosive devices, where true and false target features are taken from polarimetric ground-penetrating radar (GPR) signals. The radar measurements are carried out in laboratory and field scenarios by using dual-polarized Vivaldi antennas and ultrawideband multiple-input-multiple-output GPR (MIMO-GPR) systems in the frequency range from 0.8 to 5 GHz. Recursive algorithms, recursive least square, and linear predictive coding are used for clutter removal. To extract targets features of polarimetric measurements, eight data-processing methods are assembled, combining clutter removal algorithms, time-frequency transformations, and singular value decomposition. Moreover, for every method, 13 target feature vectors are constructed. Each vector is used to train and test a support vector machine algorithm for the classification problem. Classification results are validated by using leave-two-out cross-validation. Accuracy of 87.02% and a false positive rate of 10.53% in the best classifier were obtained.