Diffusion filtering for seismic data enhancement

Abstract

The application of diffusion filtering for seismic data enhancement will be discussed in this abstract. After a short review of diffusion filtering, some tests were conducted for comparing this method with conventional linear-filtering methods such as band-pass filtering and non-linear-filtering methods such as median filtering. The results show that diffusion filtering is more advantageous for keeping the seismic image edges while attenuating noise. It can be smartly used for de-noising non-coherent noises caused by unexpected heterogeneities in pre-stack data and enhancing the structure properties in post-stack data. It is a meaningful tool for seismic interpretation.