A computational efficient approach to identify very small mine-shaped plastic objects, e.g. M56 Anti-Personnel (AP) mines buried in the ground, is presented. The size of the objects equals the smallest AP-mines in use today, i.e., the most difficult mines to detect with respect to humanitarian mine clearance. Our approach consists of three stages, the phase stepped-frequency radar method, generation of a quarternary image and template crosscorrelation. The phase stepped-frequency radar method belongs to the class of stepped-frequency radar methods. In a two-dimensional mesh-grid above the ground a radar probe is moved automatically to measure in each grid point a set of reflection coefficients from which phase and amplitude information are extracted. Based on a simple processing of the phase information, quarternary image and template cross-correlation a successful detection of metal- and non-metal mine-shaped objects is possible. Measurements have been performed on loamy soil containing different mine-shaped objects
展开▼
