Hidden Markov models and morphological neural networks for GPR-based land mine detection

Abstract

Previous results with Hidden Markov models showed that they could be used to perform reliable classification between mines and background/clutter under a variety ofconditions. Since then, new features have been defined and continuous models have been implemented. In this paper, new results are presented for applying them to calibration lane GPR data obtained during the Vehicle Mounted Mine Detection (VMMD) Advanced Technology Demonstrations. Morphological Neural Networks can be trained to perform feature extraction and detection simultaneously. Generalizing these networks to incorporate Choquet Integrals provides the added capability ofrobustness and improved feature learning. These features can provide complementary information compared to those generated by humans. Results of applying these networks to calibration lane GPR data from the VMMD Advanced Technology Demonstrations are provided. Combinations of the various methodologies with previously developed algorithms are also evaluated.

Adaptive hidden Markov models for extended landmine detection

“…An important method of this approach is the Maximum a posteriori Estimation (MAP) 18 . MAP is different from ML in the sense that the MAP estimate of the parameter θ MAP is given by (2) where P(θ) is the prior distribution of the parameter θ. If there is some prior information available Bayesian methods present optimal way to use it.…”