A new GPR image de-nosing method based on BEMD

摘要

This paper presents a new de-noising method for GPR image based on BEMD and wavelet. This method complies with the adaptability from BEMD. The method decomposes the image into a series of IMF components, then applies wavelet threshold de-noising on the selected high frequency IMF components for de-noising. In the reconstruction course, the de-noising IMF and low frequency IMF are combined. The experiment results shows the effectiveness of the method on GPR image.