Application of wavelet theory for denoising in acoustic logging

Acoustic velocity is an important physical parameter in the geotechnical engineering domain. At present, acoustic log in the borehole is the main tool for providing acoustic velocity. However, noise, as a key factor, affects the accuracy and reliability of measure. It often leads to the wrong estimation and classification of rock and soil due to the uncertainty of velocity. Therefore, noise suppression is a most important task for accurate velocity estimation. The wavelet transform is a very useful method for modem signal processing for its multiresolution analysis. This paper applied wavelet transform to attenuate noise in acoustic velocity curve. Its basic theory is to decompose velocity data according to multiresolution theory, then separate noise from velocity curve, and mute die decomposition coefficients in small scales. So velocity data almost without any noise can be reconstructed through the inverse discrete wavelet transform. Real examples show that this method can inherit random noise effectively and present a reliable velocity of acoustic wave in rock. Finally a practical application of acoustic wave velocity after denoising to engineering geology is given to prove its effectiveness.