TY  - CONF
TI  - Blind deconvolution technique for de-noising of Non-stationary seismic signals using DWT
SP  - 1230
ID  - scholars169
KW  - Blind deconvolution; Convolutional noises; Denoising; Effective tools; Gaussian; Gaussian noises; Hard thresholding; Heat gradients; High frequency components; Input signals; Multiple subsystems; On times; Seismic datums; Seismic signals; Seismic sources; Shrinkage effects; Soft thresholding; Stationarity; Time frequencies; Time variants; Time varying; Wavelet coefficients
KW  -  Convolution; Discrete wavelet transforms; Earthquakes; Seismic response; Seismic waves; Seismology; Trellis codes
KW  -  Wavelet transforms
N2  - Discrete wavelet transform is an effective tool to disintegrate the time variant seismic data in time-frequency manner. This work incorporates the wavelet transform in the blind deconvolution technique to deal with the inherent non-stationarity present in seismic data and to improve the SNR of seismic data. Time varying nature of seismic data is the result of a depth varying character of seismic source wavelet (where high frequency components of the source wavelet get absorbs due to increasing heat gradient with depth) convolved with the non Gaussian distributed earth reflectivity in presence of additive Gaussian, color Gaussian noise. Seismic signal can thus be considered as a result of multiple subsystems with different constraints based on time-frequency localization convolved with input signal. Techniques based on stationarity assumptions are not effective in modeling the time variance character of source with depth. In this work we apply the discrete wavelet transform (DWT) to decompose the seismic data into different time-frequency signals. Denoising based on soft thresholding is applied to get the shrinkage effect of wavelet coefficients. Combination of blind deconvolution technique mixed with the discrete wavelet transform gives the best result in terms of reducing the noise and improving the resolution of seismic data with time. Denoising based on soft thresholding gives optimal minimum means square value, low convolutional noise and also low maximum distortion value than hard thresholding. ©2007 IEEE.
N1  - cited By 0; Conference of 2007 International Conference on Intelligent and Advanced Systems, ICIAS 2007 ; Conference Date: 25 November 2007 Through 28 November 2007; Conference Code:74506
AV  - none
CY  - Kuala Lumpur
EP  - 1235
A1  - Younis, M.S.
A1  - Hani, A.F.M.
UR  - https://www.scopus.com/inward/record.uri?eid=2-s2.0-57949102563&doi=10.1109%2fICIAS.2007.4658580&partnerID=40&md5=6db36d05a4c690428a86e3664ab45cd3
SN  - 1424413559; 9781424413553
Y1  - 2007///
ER  -