ALFT2D

This process uses the anti-leakage Fourier transform to reconstruct dead traces. It computes Fourier components from irregular sampling, which are then used to interpolate dead traces. This is performed on frequency slices, as with F-X algorithms. The user specifies how many iterations to use, each one adding another component. Unfortunately, this is a slow Fourier transform, unlike the FFT used in many algorithms.

The plots below compare ALFT2D with POCS2D and CADZOW, which can all be used to restore dead traces. The input was the AVO12 open data, which was rather good quality, so noise was added. Channels 11,22,44 and 45 were killed to be reconstructed.

The images will open in a new tab/window:

ALFT2D
POCS2D
CADZOW



Reference:
Xu, Zhang, Pham, Lambaré (2005) Antileakage Fourier transform for seismic data regularization, Geophysics vol 70, issue 4

Note that you need a sourceforge revision of at least 313.