An Adaptive Ultrasonic Backscattered Signal Processing Technique for Accurate Object Localization Based on the Instantaneous Energy Density Level

摘要

In clinical medicine, Ultrasonics is utilized to acquire information of the human body tissues to perform diagnosis since it is convenient and nondestructive. In this article, an adaptive ultrasonic backscattered echoes processing technique for the accurate object localization based on the instantaneous energy density level (IE) is for the first time presented. In this technique, a series of intrinsic mode functions (IMFs) of the ultrasonic backscattered signal processed are acquired by the Ensemble Empirical Mode Decomposition (EEMD) algorithm firstly. And then the Hilbert transform (HT) is applied on IMFs screened by the IMF selection process to gain the Hilbert spectrum. Finally, the time-frequency information in the Hilbert spectrum is utilized to extract the instantaneous energy density level to detect and localize the objective. In this study, the proposed technique based on the IE shows its high-performance comparing with the classic envelope detection (CED) method, and the relative error of localization is no more than 1.8% even in strong noise environment (SNR = -10 dB).