Statistical Homomorphic Deconvolution

Homomorphic deconvolution has been applied to seismic data processing widely. It is a nonlinear system and it is not necessary to assume that wavelet is minimum phase. Wavelets and reflection coefficients can be separated in homomorphic deconvolution; however, how to choose the optimum wavelet needs a lot of manual time and processing experiences. The optimum wavelet can be chosen automatically in Statistical Homomorphic Deconvolution by usingstatistical methods. The criterions of choosing the optimum wavelet are maximum variance norm and PARSIMONY variance norm methods. The optimum wavelet which is derived by maximum variance norm criterion is always same as that of PARSIMONY variance norm criterion through many seismic processing tests. The wavelet derived by Statistical Homomorphic Deconvolution can be transformed into minmum phase wavelet and zero phase wavelet. The transformed minmum phase wavelet can be satisfied by the deconvolutions which assume that the wavelet is minimum phase wavelet. The transformed zero phase wavelet is expected to be the output of seismic sections. Statistical Homomorphic Deconvolution has been applied to seismic data processing successfully not only saving manual time but also giving the best wavelet. Specially, Statistical Homomorphic Deconvolution is very useful for seismic processing that needs mixed phase wavelet Deconvolution.