Downloads: 127 | Views: 138

Research Paper | Environmental Science Studies | India | Volume 2 Issue 2, February 2013

Denoising and Reconstruction of Very Low Frequency Signal with Wavelet Thresholding

Deepak Kumar Sondhiya | Shivali Verma ^[2] | A. K Gwal

Abstract: Over past few decades ambient environmental noises increased tremendously due to man made and natural factors. There are various reasons for these noises such as noisy heavy engine, pumps, spacecraft, lighting and earthquake. These noises often degrade the quality of very low frequency (VLF) signal observed by satellite. Due to this lot of information associated with these signals are lost. To retrieve signal with significant geologic information many types of denoising methods are used, but these traditional methods are based on linear pass band filter. These families of filters are useful according to phase properties but their efficiency is reduced when we used these filters for denoising of VLF signals. This is due to the fact that these signals are nongaussian, which contains gaussian background and narrow pulses due to lighting discharge, Power Line Harmonic Radiation (PLHR) and some time due to earthquake and volcanic eruptions. In this work we developed a new method for signal denoising which is based on wavelet threshold algorithms. We show that thresholding the wavelet coefficients of a VLF signal allows to restore the complete shape of the original signal. In this approach we substantially improve the performance of classical wavelet denoising algorithms, both in terms of SNR and of visual artifacts.

Keywords: Very Low Frequency Signal, denoising, Wavelet thresholding

Edition: Volume 2 Issue 2, February 2013,

Pages: 513 - 518

Similar Articles with Keyword 'Very'