Dealiased interpolation by low-frequency constrained iterative local slope estimation

When interpolating the regularly missing traces by using a seislet-based POCS algorithm, there are three key factors that can affect the final reconstruction results: (1) The slope estimation. (2) The threshold value in the seislet transform. (3) The number of iterations. For the traditional approach, we calculate the local slope every a small number of iterations using the low-frequency components of seismic data. The large number of iterations is also used to ensure an acceptable slope estimation. The threshold value is designed assuming that a constant percentage of transform coefficients can optimally represent the true data. In the next section, we propose a very effective and efficient way for improving the traditional interpolation by using a different slope estimation approach, which is efficient and accurate enough to allow us to obtain a successful interpolation performance through a fairly small number of iterations.