Abstract

A technique is suggested for denoising the ship gravity data using wavelet transform. The wavelet transform is a useful tool for analyzing the nonstationary signals. The denoising of a signal using wavelet transform is generally carried out using different types of thresholding. Three thresholding techniques of soft, hard and customized are tested on the synthetic gravity signals with white noise. The root mean square error and coefficient of determination between synthetic and denoised gravity signals indicate the superiority of customized thresholding. Further all the thresolding approaches are applied to the ship and satellite gravity data. The satellite gravity data was extracted along the ship tracks. Both (ship and satellite) gravity data is compared before and after denoising. Again the customized thresholding is found to provide better results.