M.Sc Tezi Görüntüleme

       Deconvolution is an important processing step of seismic data processing. Firstly, different form of Wiener filters and the least square method which known the starting point of Wiener filters in literature are shown solving of the linear time invariant system in the matrix form. Furthermore, deconvolution operator which suggests with convolution processing steps return and approaches which used Fourier methods are given.

      In this problem conventional approaches have instability problem due to the zeros of the convolution kernel. This study suggests a new method of multidimensional exact deconvolution mathematically. Advantage and disadvantage of solving 1-dimensional closed-form of this method between the other conventional methods on the prepared synthetic seismograms have determined with matching.

      In here, minimum and zero phase wavelet are used in the seismograms which are built with synthetic. Then behaviors of the processing, which are shown in case adding the noisy to synthetic seismogram are examined. At the same time, amplitude spectrums have been computed with used the Fourier transform for comparing in the frequency domain in this study.Therefore, the most effective method in deterministic deconvolution methods can be given the exact deconvolution method.

      Key Words : Deconvolution, inverse filters, multidimensional filters, deconvolution operator, least square filters, Wiener filter.