Mine detection using a maximum likelihood estimator and in-field GPR data

This contribution addresses the signal preprocessing and a detection algorithm applied to measured ground penetrating radar (GPR) data. The measurement setup consists of a GPR system and inert mines in five different environments. The statistics of the output variable of a maximum likelihood detection algorithm which also reduces the signal to clutter ratio (SCR) is analyzed. As a result the threshold level for a constant false alarm rate (CFAR) mode can be computed.