Texture analysis approaches to extract ocean wavelengths, directions, and heights from ocean wave images

Texture analysis approaches for extracting ocean wavelengths, directions, and heights from ocean wave images are presented. In the frequency domain, we use Fourier power spectral features to construct the texture measures. In the spatial domain, we introduce three texture measures based on the distribution of local extrema, a gray-level co-occurrence matrix, and two-dimensional histograms of the absolute-relative peak heights to discriminate different kinds of ocean wavelengths, directions, and heights. The experimental results indicate that these approaches are powerful.