Ultrasonic Multitransducer System for Classification and 3-D Location of Reflectors Based on PCA

Abstract

This paper presents a system to classify and locate basic ultrasonic reflectors (plane, corner, and edge) in 3-D environments. The classification system is based on the principal-component-analysis (PCA) technique, using the time of flight (TOF) as the classification parameter. The system proposes a sensorial structure to simultaneously obtain up to 16 TOFs at every emission/scanning process. A particular and different macrosequence has been assigned to every transducer in the sensorial system to encode their emissions. These macrosequences, obtained from complementary sets of sequences, allow simultaneous emission and reception to be carried out with all the transducers for the same scanned environment. The set of obtained TOFs make it possible to identify the type of ultrasonic reflector, as well as its position in 3-D environments. The results achieved by the classification system are presented for both simulated and real data.