Ionograms denoising via curvelet transform

Ionograms are used to obtain important information of the ionosphere. Unfortunately, ionograms are always contaminated by several kinds of noises. In this paper, curvelet transform denoising algorithm is used to obtain high-quality ionograms. This algorithm is based on image processing and can preserve the layer traces better than other methods. In the process of curvelet transform denoising, we propose an adaptive threshold based on Bayes theory to improve the performance of this method. For practical applications to ionogram denoising, this curvelet transform method is combined with the traditional method to deal with a variety of ionogram noise such as radio interferences. This combined approach has been validated using data from Chinese Academy of Science-Digital Ionosonde (CAS-DIS), and can be used for ionogram automatic scaling.