Tutorial 1 Random Noise Suppression Transform We will implement the noise2void methodology of blind spot networks for denoising. this involves performing a pre processing step which identifies the 'active' pixels and then replaces their original noisy value with that of a neighbouring pixel. On completion of this tutorial you will have learnt how to write your own blind spot denoising procedure that is trained in a self supervised manner, i.e., the training data is the same as the inference data with no labels required! we will implement the noise2void methodology of blind spot networks for denoising.
Tutorial 1 Random Noise Suppression Transform In this representation, coherent seismic events become sparsely concentrated, whereas random noise remains broadly distributed, enabling more effective signal–noise separation through coefficient thresholding. To address these limitations, we introduce a self supervised conditional diffusion (sscdiff) model for 3 d seismic random noise attenuation. in sscdiff, we introduce the conditional diffusion model to solve the problem of random noise attenuation rather than gaussian noise only. We propose an adaptive time reassigned synchrosqueezing transform (atsst) by introducing a time varying window function to improve the time frequency concentration, and integrate an improved optshrink algorithm for the suppression of seismic random noise. We propose a robust method to suppress both random and erratic noise. both synthetic and field data examples are used to demonstrate the effectiveness. linear radon transform, or slant stack transform, can be used to obtain an optimally sparse representation of linear seismic events.
Tutorial 1 Random Noise Suppression Transform We propose an adaptive time reassigned synchrosqueezing transform (atsst) by introducing a time varying window function to improve the time frequency concentration, and integrate an improved optshrink algorithm for the suppression of seismic random noise. We propose a robust method to suppress both random and erratic noise. both synthetic and field data examples are used to demonstrate the effectiveness. linear radon transform, or slant stack transform, can be used to obtain an optimally sparse representation of linear seismic events. To address this problem, we develop a structure oriented 3d denoising method based on 3d curvelet transform, which uses the dip information to analyze local features complexity. Transforming the data to the fk domain, raising the amplitude spectrum to an exponential power and performing the inverse transform is a method sometimes used to reduce the levels of random noise. the method is available in promax and landmarks post stack processing system. Whilst originally proposed for the suppression of i.i.d. noise in natural images and microscopy data, in this work we show that n2v can be adapted into an efficient suppressor of band passed random noise in seismic data, i.e. data that has some degree of correlation along the time axis. Predictive filtering (pf) in the frequency domain is one of the most widely used denoising algorithms in seismic data processing. pf is based on the assumption of linear or planar events in the time space domain.
Comments are closed.