Application of Entropy Based Adaptive Thrsholding and Phase Retrieval Technique for Target Detection and Identification in Ground Penetrating Radar

The ground-penetrating radar (GPR) is used to develop images for the buried objects under the ground surface. An indigenously developed GPR system in the frequency range 12 GHz is used here to obtain the raw data after time gating by placing the metal plate at the mouth of the antenna. In this paper, we discuss the identification and classification of the various objects using the phase retrieval technique and the supervised machine learning. To remove the clutter from the raw data finding the adaptive threshold value is an important task, the major challenge in the target detection is to reduce the false alarm rate. The Simulated Annealing (SA) algorithm is modified using the entropy of the raw data to find the energy level in the threshold value. The proposed algorithm based on entropy and SA gives encouraging results to detect different objects buried underground.