Analysis of acoustic back scattered signals of two different underwater materials using Empirical Mode Decomposition and support vector machine

Abstract

<span style="font-size:10.0pt;font-family: " times="" new="" roman","serif";mso-fareast-font-family:"times="" roman";mso-bidi-font-family:="" mangal;mso-ansi-language:en-gb;mso-fareast-language:en-us;mso-bidi-language:="" hi"="" lang="EN-GB">In this work an attempt has been made to analyse and discriminate acoustic backscattered signals from underwater objects of two different materials of PVC and Brass. A laboratory study of underwater acoustic scattering of spherical objects of PVC and Brass material is carried out using Empirical Mode Decomposition (EMD), HilbertTransform (HT) and Support Vector Machine (SVM). Incident signal used for the measurement is a Linear Frequency Modulated (LFM) signal of finite duration with the signal bandwidth of 40 kHz to 80 kHz. More than 80 back scattered acoustic signals from the objects are recorded and processed for discrimination. An EMD method is designed to decompose the scattered signal and HT was used to extract the features for discrimination. EMD decomposes the backscattered signal into intrinsic mode functions ((IMFs) and the significant features are extracted from the HT. The classification or discrimination is investigated using support vector machine (SVM) with four types of kernels such as linear, quadratic, RBF and polynomial. Performance of the SVM shows that the proposed method using EMD and Hilbert transform is useful for underwater object discrimination.</span>