Well Log Denoising by Wavelet Transform and Classification Algorithm

It is well known that signals obtained as responses from geological media contain kind of noises that can affect the lithological interpretation. Wavelet transform is more suitable than Fourier transform to denoise non-stationary signals, such as well logs. However, there are several parameters that must be considered when working with the wavelet transform, such as the wavelet basis function, the thresholding rule and the decomposition level. This study analyzes the well logs denoising process by discrete wavelet transform combined with classification algorithm, and investigates how different combination of parameters affects the output signals and their performance on the lithological classification.