Using convolutional neural networks to classify unexploded ordnance from multicomponent electromagnetic induction data

Electromagnetic induction (EMI) methods are commonly used to classify unexploded ordnance (UXO) in both terrestrial and marine settings. Modern time-domain systems used for classification are multicomponent which means they acquire many transmitter-receiver pairs at multiple time-channels. Traditionally, classification is performed using a physics-based inversion approach where polarizability curves are estimated from the EMI data. These curves are then compared with those in a library to look for a match based on some misfit measure. In this work, we developed a convolutional neural network (CNN) that classifies UXO directly from EMI data. Analogous to an image segmentation problem, our CNN outputs a classification map that preserves the spatial dimensions of the input. In this way, our CNN produces high-resolution results and can handle the multiple transmitter-receiver pairs and the acquisition of multicomponent systems. We train the CNN using synthetic data generated with a dipole forward model considering relevant UXO and clutter objects. A careful design of the clutter classes is needed to maximize clutter discrimination. 

We use a two-step workflow. First, we train a CNN to detect metallic objects in field data. From this, we extract patches of data that contain only background signal and use these to generate a new training data set by adding this background noise to our synthetic data. A second CNN is trained with these data to perform the classification. We test our approach using field data acquired with the UltraTEMA-4 system in the Sequim Bay marine test site. Using this workflow, classification results for the field data show that our approach detects all of the UXOs and classifies more than 90% as the correct type while also discriminating ~70% of the clutter. A key advantage of our CNN is that, once trained, it may be used to provide real-time classification results on the field.