A Non-destructive Approach for Noise Reduction in Time Domain
Amin Zehtabian, Hamid Hassanpour · 2009
Abstract: In this paper a new time domain noise reduction approach is presented. In existing noise reduction techniques, the structure of the enhanced signal might be slightly changed compared to the original signal. In the proposed non-destructive approach, the noisy signal is initially represented in a Hankel Matrix. Then the Singular Value Decomposition (SVD) operator is applied on the matrix to divide the data into signal subspace and noise subspace. Reducing the effect of noise from the singular vectors and using them in reproducing the matrix, leads to significant enhancement of information embedded in the matrix. This matrix is finally used to obtain the time-series signal. There are several important parameters, such as size of the matrix and the degree of the filter, affecting the performance of the proposed approach. These parameters are optimally set using the genetic algorithm. The results of applying the proposed method on different synthetic noisy signals indicate its better efficiency in noise reduction compared to the other time series methods. The obtained results also indicate that using the proposed approach keeps the main structure of the original signal unchanged.