Wavelet Spectrum Extracted from Coastal Marine Radar Images

Ocean waves are extremely random and are directly and indirectly dependent on meteorological, hydrological, oceanographic and topographical factors. Field measurements must be performed to increase the knowledge of ocean waves. Remote sensing techniques have become quite popular for ocean wave measurement applications. Microwave RADAR that can provide all weather monitoring during day and night and is less affected by cloud and rain is now the most popular sensor of remote sensing techniques among others. The spectrum represents the distribution of wave energy. The wave parameters such as wave height, period, direction and wavelength are derived from wave spectrum. It is therefore important to derive the spectral information from remote sensing images. However, the nearshore images feature non-homogeneity. It is therefore the purpose of this study to develop a non-homogeneous image spectrum analysis method in order to extract the spectral information. This method is developed based on two-dimensional Wavelet transform which has space-frequency representation. The derived image spectrum is called the Wavelet Spectrum. This paper presents the derivation and validation of this method.