Wavelet application in classification of strata

Abstract

Geophysical survey needs logging data to speculate stratum situation. The classification of strata based on logging curve is the first thing for the inversion problem. Electric logging curve is a stochastic time series. Because of the complex strata structure, the time series consist of real signal and white Gaussian noise. In order to get the real logging curves, the noise should be removed. In this paper, wavelet is used to filter the electric logging curves and the filtered signal properly reserves the curve edge and effectively wipes off the mutations of noise. The classification of strata based on the wavelet pre-treatment of electric logging curve becomes more ideal than the original curve. © 2012 Springer-Verlag.