Deep convolutional neural networks for classifying GPR B-scans

Symmetric and asymmetric buried explosive hazards (BEHs) present real, persistent, deadly threats on the modern battlefield. Current approaches to mitigate these threats rely on highly trained operatives to reliably detect BEHs with reasonable false alarm rates using handheld Ground Penetrating Radar (GPR) and metal detectors. As computers become smaller, faster and more efficient, there exists greater potential for automated threat detection based on state-of-the-art machine learning approaches, reducing the burden on the field operatives. Recent advancements in machine learning, specifically deep learning artificial neural networks, have led to significantly improved performance in pattern recognition tasks, such as object classification in digital images. Deep convolutional neural networks (CNNs) are used in this work to extract meaningful signatures from 2-dimensional (2-D) GPR B-scans and classify threats. The CNNs skip the traditional “feature engineering” step often associated with machine learning, and instead learn the feature representations directly from the 2-D data. A multi-antennae, handheld GPR with centimeter-accurate positioning data was used to collect shallow subsurface data over prepared lanes containing a wide range of BEHs. Several heuristics were used to prevent over-training, including cross validation, network weight regularization, and “dropout.” Our results show that CNNs can extract meaningful features and accurately classify complex signatures contained in GPR B-scans, complementing existing GPR feature extraction and classification techniques.

[1]  Leslie M. Collins,et al.  Texture Features for Antitank Landmine Detection Using Ground Penetrating Radar , 2007, IEEE Transactions on Geoscience and Remote Sensing.

[2]  Joseph N. Wilson,et al.  An evaluation of several fusion algorithms for anti-tank landmine detection and discrimination , 2012, Inf. Fusion.

[3]  Joseph N. Wilson,et al.  Compactometry, the density distribution, and their use in discriminating landmines and clutter , 2005, SPIE Defense + Commercial Sensing.

[4]  Joseph N. Wilson,et al.  Region processing of ground-penetrating radar and electromagnetic induction for handheld landmine detection , 2004, SPIE Defense + Commercial Sensing.

[5]  Kenneth D. Morton,et al.  Sensor fusion approaches for EMI and GPR-based subsurface threat identification , 2011, Defense + Commercial Sensing.

[6]  Lance E. Besaw,et al.  Deep learning algorithms for detecting explosive hazards in ground penetrating radar data , 2014, Defense + Security Symposium.

[7]  Shengen Yan,et al.  Deep Image: Scaling up Image Recognition , 2015, ArXiv.

[8]  Paul D. Gader,et al.  Detection and Discrimination of Land Mines in Ground-Penetrating Radar Based on Edge Histogram Descriptors and a Possibilistic $K$-Nearest Neighbor Classifier , 2009, IEEE Transactions on Fuzzy Systems.

[9]  Pascal Vincent,et al.  Unsupervised Feature Learning and Deep Learning: A Review and New Perspectives , 2012, ArXiv.

[10]  M. Ivimey Annual report , 1958, IRE Transactions on Engineering Writing and Speech.

[11]  Yoshua Bengio,et al.  Convolutional networks for images, speech, and time series , 1998 .

[12]  Joseph N. Wilson,et al.  Landmine detection using frequency domain features from GPR measurements and their fusion with time domain features , 2005, SPIE Defense + Commercial Sensing.

[13]  Nitish Srivastava,et al.  Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[14]  Joseph N. Wilson,et al.  Development of region processing algorithm for HSTAMIDS: status and field test results , 2007, SPIE Defense + Commercial Sensing.

[15]  Humanitarian Demining,et al.  Mines Other than Anti-Personnel Mines , 2012 .

[16]  Paul D. Gader,et al.  A linear prediction land mine detection algorithm for hand held ground penetrating radar , 2002, IEEE Trans. Geosci. Remote. Sens..

[17]  Razvan Pascanu,et al.  Theano: new features and speed improvements , 2012, ArXiv.

[18]  Kenneth D. Morton,et al.  Adaptive Gaussian mixture models for pre-screening in GPR data , 2011, Defense + Commercial Sensing.

[19]  Joseph N. Wilson,et al.  DynaMax+ ground-tracking algorithm , 2011, Defense + Commercial Sensing.