A fast seismic data processing approach based on parallel wavelet transform

A new implementation of seismic data processing approach based on parallel wavelet transform is presented in this paper. Firstly, the feasibility and necessity of the parallel processing method for seismic data are elaborated from three primary aspects. Based on the analysis, the parallel wavelet transform is implemented by using MATLAB on a distributed cluster. Finally, both sequential and parallel implements are employed to test and compare the performance of our implementation. A series of experiments have demonstrated that parallel wavelet transform can significantly reduce the running time and improve the speedup greatly.