2-D continuous wavelet-based algorithm for extracting wave information from nautical radar images

Abstract

In this paper, a 2-D continuous wavelet transform-based algorithm for the extraction of wave information from X-band nautical radar images is presented. By reviewing the theory of the wavelet transform for image processing, a sophisticated scheme for the wavelet design is developed to obtain image spectra. First, the Morlet wavelet is selected and simplified as the mother wavelet. Then, a configuration and sampling strategy, which enables the non-dimensional continuous wavelet to discretely process the digital radar images in dimensional space, is proposed. Wavelet parameters are derived theoretically and calibrated numerically. Similar techniques employed in the Fourier analysis are incorporated into the proposed algorithm. These include image normalization and a modulation transfer function for the retrieval of wave spectra and wave parameters from the image spectra derived by wavelet. The algorithm is first evaluated via numerical simulation and then applied to horizontally-polarized radar field data collected on the East Coast of Canada. Good agreement of the results with the in-situ record from a TRIAXYS wave buoy indicates that the proposed algorithm is effective for retrieving ocean wave information.