Sub-array processing techniques applied to the detection of buried targets

A sub-array processing (SAP) technique for the localization of conductive buried objects is proposed. A Direction of Arrival (DoA) approach is implemented by using two different methodologies: a bearing algorithm based on the conventional MUltiple SIgnal Classification (MUSIC) implementation is compared with a Kernel-based technique, such as Support Vector Machines (SVM). An array of sensors is considered and partitioned in sub-arrays. A dominant DoA is found for each sub-array. By triangulating the DoAs, a crossing pattern is derived and it is statistically processed in order to estimate the object position.