LOCAL MESHLESS ALGORITHM FOR TOTAL VARIATION-BASED MODEL FOR SIGNAL RESTORATION HAVING ADDITIVE NOISE

Abstract

A noisy signal converts the signal to another signal containing fluctuation called signal fluctuation. This fluctuation in signal is a common factor in signal processing. In this article, we introduce a Local Meshless Collocation Method (LMCM) for the numerical solution of the Euler- Lagrange Partial Differential Equation (EL-PDE) associated with the Total Variation (TV)-based model for removing additive noise from the given data signals. This method uses the mul-ti-quadric Radial Basis Function (MQ-RBF) as the basis function. These features will eliminate the fluctuations from the given noisy signals well due to the meshless applications used in the suggested method. The experimental result demonstrates that the proposed LMCM will perform well in terms of Signal to Noise Ratio (SNR) compared to another traditional mesh-based scheme for different basis functions.