next up previous [pdf]

Next: Acknowledgement Up: Zheng et al.: Interpolation Previous: Discussion

Conclusion

We introduced a fast approach to adaptive PF for missing data interpolation in the frequency domain. Instead of using the iterative optimization algorithm, we proposed a two-step interpolation strategy with noniterative SPF in the $ f$ -$ x$ and $ f$ -$ x$ -$ y$ domains. The proposed method employs a local and multidimensional similarity to constrain the autoregression equations for adaptive PFs in the frequency domain, which are based on the streaming computation framework. The SPF in the frequency domain provides a fast and reasonably accurate estimation of nonstationary seismic data. To guarantee the interpolation results, we also designed the filter structure and the processing path according to the characteristics of the interpolation problem. The synthetic and field examples show that the proposed SPF in the frequency domain can depict nonstationary signal variation and provide a reliable description of complex wavefield with low computational cost even when analyzing large-scale seismic data. The properties are suitable for missing data interpolation in practice. Finally, we discussed the problems of parameter selection, interpolation of regularly decimated data, and interpolation of low SNR data; the proposed methods can cope with such problems.


next up previous [pdf]

Next: Acknowledgement Up: Zheng et al.: Interpolation Previous: Discussion

2022-04-15