Landmine detection and classification with complex-valued hybrid neural network using scattering parameters dataset

Neural networks have been applied to landmine detection from data generated by different kinds of sensors. Real-valued neural networks have been used for detecting landmines from scattering parameters measured by ground penetrating radar (GPR) after disregarding phase information. This paper presents results using complex-valued neural networks, capable of phase-sensitive detection followed by classification. A two-layer hybrid neural network structure incorporating both supervised and unsupervised learning is proposed to detect and then classify the types of landmines. Tests are also reported on a benchmark data.     

Robust landmine detection from thermal image time series using Hough transform and rotationally invariant features

ABSTRACT

Automated detection of buried anti-personnel landmines using remote sensing techniques is very important for clearing minefields without putting lives in danger. Although thermal infrared imaging is promising, it is far from applicable to the real world in its current state-of-the-art. The most serious problem is that experiments are generally held using sandboxes or levelled and cleared soil, but real fields are, at least partially, covered with plants. In this study, we present an algorithm for landmine detection that is robust enough to detect beyond the clutter caused by partial plant cover. The first part is a hypothesis generator based on circular Hough Transform applied to images that are filtered to enhance circular structures. The second part tests the candidate landmine coordinates using rotationally invariant features, including modified Histogram of Oriented Gaussians (HOG), over multiple images taken at different times after Wiener filtering to maximize signal-to-clutter ratio. The performances of various features and classifiers are compared. The overall performance of the algorithm is demonstrated on a dataset of real-world landmine images contaminated by simulated plants. Satisfactory results are obtained up to 40% equivalent plant coverage where more than 65% of the pixels are fully or partially covered by plants.     

Covariance-guided One-Class Support Vector Machine

In one-class classification, the low variance directions in the training data carry crucial information to build a good model of the target class. Boundary-based methods like One-Class Support Vector Machine (OSVM) preferentially separates the data from outliers along the large variance directions. On the other hand, retaining only the low variance directions can result in sacrificing some initial properties of the original data and is not desirable, specially in case of limited training samples. This paper introduces a Covariance-guided One-Class Support Vector Machine (COSVM) classification method which emphasizes the low variance projectional directions of the training data without compromising any important characteristics. COSVM improves upon the OSVM method by controlling the direction of the separating hyperplane through incorporation of the estimated covariance matrix from the training data. Our proposed method is a convex optimization problem resulting in one global optimum solution which can be solved efficiently with the help of existing numerical methods. The method also keeps the principal structure of the OSVM method intact, and can be implemented easily with the existing OSVM libraries. Comparative experimental results with contemporary one-class classifiers on numerous artificial and benchmark datasets demonstrate that our method results in significantly better classification performance.     

Using one-class and two-class SVMs for multiclass image annotation

We propose using one-class, two-class, and multiclass SVMs to annotate images for supporting keyword retrieval of images. Providing automatic annotation requires an accurate mapping of images' low-level perceptual features (e.g., color and texture) to some high-level semantic labels (e.g., landscape, architecture, and animals). Much work has been performed in this area; however, there is a lack of ability to assess the quality of annotation. In this paper, we propose a confidence-based dynamic ensemble (CDE), which employs a three-level classification scheme. At the base-level, CDE uses one-class support vector machines (SVMs) to characterize a confidence factor for ascertaining the correctness of an annotation (or a class prediction) made by a binary SVM classifier. The confidence factor is then propagated to the multiclass classifiers at subsequent levels. CDE uses the confidence factor to make dynamic adjustments to its member classifiers so as to improve class-prediction accuracy, to accommodate new semantics, and to assist in the discovery of useful low-level features. Our empirical studies on a large real-world data set demonstrate CDE to be very effective.     

One-class remote sensing classification: one-class vs. binary classifiers

Many applications of remote sensing only require the classification of a single land type. This is known as the one-class classification problem and it can be performed using either binary classifiers, by treating all other classes as the negative class, or one-class classifiers which only consider the class of interest. The key difference between these two approaches is in their training data and the amount of effort needed to produce it. Binary classifiers require an exhaustively labelled training data set while one-class classifiers are trained using samples of just the class of interest. Given ample and complete training data, binary classifiers generally outperform one-class classifiers. However, what is not clear is which approach is more accurate when given the same amount of labelled training data. That is, for a fixed labelling effort, is it better to use a binary or one-class classifier. This is the question we consider in this article. We compare several binary classifiers, including backpropagation neural networks, support vector machines, and maximum likelihood classifiers, with two one-class classifiers, one-class SVM, and presence and background learning (PBL), on the problem of one-class classification in high-resolution remote sensing imagery. We show that, given a fixed labelling budget, PBL consistently outperforms the other methods. This advantage stems from the fact that PBL is a positive-unlabelled method in which large amounts of readily available unlabelled data is incorporated into the training phase, allowing the classifier to model the negative class more effectively.     

Detecting and locating landmine fields from vehicle- and air-borne measured IR Images

Air- and vehicle-borne sensor-based technique is a potentially attractive approach for fast detecting landmines and locating landmine fields towards humanitarian demining. For images measured from airborne and vehicle-borne cameras, landmines may be indicated by direct or indirect signs, e.g., spatial difference from their surroundings due to digging or, due to thermal and material signatures. The background in images usually consists of various types of noise and clutter, e.g., thermal noise, sand, gravel road and vegetation, thus making the detection even more difficult. This paper is focused on the following aspects: (1) Finding a robust detector that is suitable for detecting/locating landmine candidates and man-made landmarks by using infrared images measured from vehicle- or air-borne sensors; (2) Interpreting the detector using the 2D isotropic bandpass filter, matched filter, detection theory and thermodynamic-based landmine models; (3) Extending the detector to a multiscale version where landmine detectability is enhanced by automatically selecting a proper scale and localization is improved by inter-scale position tracing. We propose a special type of isotropic feature detector that exploits the characteristic difference between landmines and their surroundings in the spatial-frequency domain under the multiscale framework. Experiments were performed on several infrared images measured from vehicle-borne sensors as well as airborne sensors on a helicopter over the test bed scenarios. The performance of the detector was also evaluated in terms of detectability, localization, and automatic scale selection of the detector. These results and evaluations have shown the effectiveness of the method and its potential in landmine field detection.     

The impact of false alarm mitigation on surface landmine detection in MWIR imagery

Surface landmine and minefield detection from airborne imagery is a difficult problem. As part of the minefield detection process, anomaly detection is performed to identify potential landmines in individual airborne images. Post-processing is performed on the initial landmines identified to reduce the number of false alarms, referred to as false alarm mitigation. In this research, a circular harmonics transform image processing approach (the CHT method) and a constant false alarm rate technique (the RX approach) are investigated for surface landmine detection and false alarm mitigation in medium wave infrared (MWIR) image data. The false alarm mitigation approach integrates the CHT and RX methods to identify candidate landmine locations with one technique at a given false alarm rate and applies the other technique to confirm landmine locations and eliminate potential false alarms. Individual detector and false alarm mitigation experimental results are presented for 31 daytime and 43 nighttime MWIR images containing 76 and 142 landmines, respectively. At a 0.9 desired probability of landmine detection, experimental results show that false alarm mitigation reduces the false alarm rate by as much as 84.3% and 13.7% for daytime and nighttime images, respectively, maintaining the probability of detection at 0.85 and 0.90, respectively.     

A Back-propagation Neural Network Landmine Detector Using the Delta-technique and S-statistic

Landmines are a major problem facing the world today; there are millions of these deadly weapons still buried in various countries around the world. Humanitarian organizations dedicate an immeasurable amount of time, effort, and money to find and remove as many of these mines as possible. Unfortunately, landmines can be made out of common materials which make the correct detection of them very difficult. This paper analyzes the effectiveness of combining certain statistical techniques with a neural network to improve detection. The detection method must not only detect the majority of landmines in the ground, it must also filter out as many of the false alarms as possible. This is the true challenge to developing landmine detection algorithms. Our approach combines a Back-Propagation Neural Network (BPNN) with statistical techniques and compares the performance of mine detection against the performance of the energy detector and the δ-technique. Our results show that the combination of the δ-technique and the S-statistics with a neural network improves the performance.     

Multiclass classifiers based on dimension reduction with generalized LDA

Linear discriminant analysis (LDA) has been widely used for dimension reduction of data sets with multiple classes. The LDA has been recently extended to various generalized LDA methods that are applicable regardless of the relative sizes between the data dimension and the number of data items. In this paper, we propose several multiclass classifiers based on generalized LDA (GLDA) algorithms, taking advantage of the dimension reducing transformation matrix without requiring additional training or parameter optimization. A marginal linear discriminant classifier (MLDC), a Bayesian linear discriminant classifier (BLDC), and a one-dimensional BLDC are introduced for multiclass classification. Our experimental results illustrate that these classifiers produce higher ten-fold cross validation accuracy than kNN and centroid-based classifiers in the reduced dimensional space obtained from GLDA.     

超球体多类支持向量机理论

目前的多类分类器大多是经二分类器组合而成的,存在训练速度较慢的问题,在分类类别多的时候,会遇到很大困难,超球体多类支持向量机将超球体单类支持向量机扩展到多类问题,由于每类样本只参与一个超球体支持向量机的训练.因此,这是一种直接多类分类器,训练效率明显提高.为了有效训练超球体多类支持向量机,利用SMO算法思想,提出了超球体支持向量机的快速训练算法.同时对超球体多类支持向量机的推广能力进行了理论上的估计.数值实验表明,在分类类别较多的情况,这种分类器的训练速度有很大提高,非常适合解决类别数较多的分类问题.超球体多类支持向量机为研究快速直接多类分类器提供了新的思路.     

A review of satellite and airborne sensors for remote sensing based detection of minefields and landmines

This paper focuses on the use of space and airborne sensors that can be applied to detect landmines and minefields. First the landmine and minefield problem is addressed and examples of the use of remote sensing images are presented that could provide valuable information for the mine action process and assist in conventional minefield and landmine detection methods. This is followed by an overview on relevant (declassified) aspects related to strategic overhead detection techniques developed by the military/intelligence community as well as those of civilian space and airborne remote sensing programmes. The airborne sensing techniques describe the state of the art of sensors such as optical (film, multi- and hyperspectral sensors), thermal infrared as well as microwave sensors and their suitability--limitations for remote sensing based minefield and landmine detection purposes.     

An evaluation of several fusion algorithms for anti-tank landmine detection and discrimination

Many algorithms have been proposed for detecting anti-tank landmines and discriminating between mines and clutter objects using data generated by a ground penetrating radar (GPR) sensor. Our extensive testing of some of these algorithms has indicated that their performances are strongly dependent upon a variety of factors that are correlated with geographical and environmental conditions. It is typically the case that one algorithm may perform well in one setting and not so well in another. Thus, fusion methods that take advantage of the stronger algorithms for a given setting without suffering from the effects of weaker algorithms in the same setting are needed to improve the robustness of the detection system. In this paper, we discuss, test, and compare seven different fusion methods: Bayesian, distance-based, Dempster-Shafer, Borda count, decision template, Choquet integral, and context-dependent fusion. We present the results of a cross validation experiment that uses a diverse data set together with results of eight detection and discrimination algorithms. These algorithms are the top ranked algorithms after extensive testing. The data set was acquired from multiple collections from four outdoor sites at different locations using the NIITEK GPR system. This collection covers over 41,807 m² of ground and includes 1593 anti-tank mine encounters.     

Off-the-peg and bespoke classifiers for fraud detection

Detecting fraudulent plastic card transactions is an important and challenging problem. The challenges arise from a number of factors including the sheer volume of transactions financial institutions have to process, the asynchronous and heterogeneous nature of transactions, and the adaptive behaviour of fraudsters. In this fraud detection problem the performance of a supervised two-class classification approach is compared with performance of an unsupervised one-class classification approach. Attention is focussed primarily on one-class classification approaches. Useful representations of transaction records, and ways of combining different one-class classifiers are described. Assessment of performance for such problems is complicated by the need for timely decision making. Performance assessment measures are discussed, and the performance of a number of one- and two-class classification methods is assessed using two large, real world personal banking data sets.     

基于单类分类器的半监督学习

提出一种结合单类学习器和集成学习优点的Ensemble one-class半监督学习算法.该算法首先为少量有标识数据中的两类数据分别建立两个单类分类器.然后用建立好的两个单类分类器共同对无标识样本进行识别,利用已识别的无标识样本对已建立的两个分类面进行调整、优化.最终被识别出来的无标识数据和有标识数据集合在一起训练一个基分类器,多个基分类器集成在一起对测试样本的测试结果进行投票.在5个UCI数据集上进行实验表明,该算法与tri-training算法相比平均识别精度提高4.5%,与仅采用纯有标识数据的单类分类器相比,平均识别精度提高8.9%.从实验结果可以看出,该算法在解决半监督问题上是有效的.     

Outlier detection based on a dynamic ensemble model: Applied to process monitoring

This paper focuses on outlier detection and its application to process monitoring. The main contribution is that we propose a dynamic ensemble detection model, of which one-class classifiers are used as base learners. Developing a dynamic ensemble model for one-class classification is challenging due to the absence of labeled training samples. To this end, we propose a procedure that can generate pseudo outliers, prior to which we transform outputs of all base classifiers to the form of probability. Then we use a probabilistic model to evaluate competence of all base classifiers. Friedman test along with Nemenyi test are used together to construct a switching mechanism. This is used for determining whether one classifier should be nominated to make the decision or a fusion method should be applied instead. Extensive experiments are carried out on 20 data sets and an industrial application to verify the effectiveness of the proposed method.     

基于支持向量的单类分类方法综述

研究基于支持向量机理论和单类分类思想的2种支持向量域数据描述模型,即单分类支持向量机和支持向量描述模型,分析2类模型之间的区别和联系以及参数的优化设置,总结支持向量域单分类方法存在的缺点以及目前对这2类支持向量描述模型研究的改进方向。     

Feature and decision level sensor fusion of electromagnetic induction and ground penetrating radar sensors for landmine detection with hand-held units

Strategies for fusion of electromagnetic induction (metal detector (MD)) and ground penetrating radar (GPR) sensors for landmine detection are investigated. Feature and decision level algorithms are devised and compared. Features are extracted from the MD signals by correlating with weighted density distribution functions. A multi-frequency band linear prediction method generates features for the GPR. Feature level fusion combines MD and GPR features in a single neural network. Decision level fusion is performed by using the MD features as inputs to one neural network and the GPR features as inputs to the geometric mean and combining the output values. Experimental results are reported on a very large real data set containing 2315 mine encounters of different size, shape, content and metal composition that are measured under different soil conditions at three distinct geographical locations.     

Haar-like features with optimally weighted rectangles for rapid object detection

This article proposes an extension of Haar-like features for their use in rapid object detection systems. These features differ from the traditional ones in that their rectangles are assigned optimal weights so as to maximize their ability to discriminate objects from clutter (non-objects). These features maintain the simplicity of evaluation of the traditional formulation while being more discriminative. The proposed features were trained to detect two types of objects: human frontal faces and human heart regions. Our experimental results suggest that the object detectors based on the proposed features are more accurate and faster than the object detectors built with traditional Haar-like features.     

Adapted One-versus-All Decision Trees for Data Stream Classification

One versus all (OVA) decision trees learn k individual binary classifiers, each one to distinguish the instances of a single class from the instances of all other classes. Thus OVA is different from existing data stream classification schemes whose majority use multiclass classifiers, each one to discriminate among all the classes. This paper advocates some outstanding advantages of OVA for data stream classification. First, there is low error correlation and hence high diversity among OVA's component classifiers, which leads to high classification accuracy. Second, OVA is adept at accommodating new class labels that often appear in data streams. However, there also remain many challenges to deploy traditional OVA for classifying data streams. First, as every instance is fed to all component classifiers, OVA is known as an inefficient model. Second, OVA's classification accuracy is adversely affected by the imbalanced class distribution in data streams. This paper addresses those key challenges and consequently proposes a new OVA scheme that is adapted for data stream classification. Theoretical analysis and empirical evidence reveal that the adapted OVA can offer faster training, faster updating and higher classification accuracy than many existing popular data stream classification algorithms.