RECOGNITION OF RICKER WAVELETS BY SYNTACTIC ANALYSIS

Abstract

Syntactic pattern recognition techniques are applied to the analysis of 1-D seismic traces to classify picker wavelets. Seismic picker wavelets have structural information, and each wavelet can be represented by a string of symbols. To recognize the strings, we use a finite-state automaton to identify each string. The automaton can accept strings having substitution, insertion, and deletion errors of the symbols. There are two attributes, terminal symbol and weight, in each transition of the automaton. A minimum-cost, error-correcting, finite-state automaton is proposed to parse the input string.