Method for converting seismic reflected waves into machine learning sample library

Seismic data can be used for explaining geologic structures and researching reservoir lithology, for a large set of strata, seismic events are top and bottom interfaces of the strata, when the strata present thin interbed deposition, seismic reflection is no longer the characteristic of standard wavelets, but presents a complex wave structure, and the seismic reflection is not the characteristic of standard wavelets. The complex wave event feature not only can identify the height of the structure, but also has a thin interbed combination feature. According to the seismic exploration principle, the seismic data resolution is lower than the seismic reflection wave quarter wavelength limit, that is to say, the seismic reflection limit can be used for distinguishing lambda/4, lambda is the seismic emission wavelength, the calculation method is that lambda = V/F, and F is the seismic data video rate. Seismic amplitudes of seismic reflection are restructured, and according to the requirement that each sample has th