A wavelet based method for the extraction of sea wave orientation

In this paper, a method for determining sea wave orientation from image data is presented. Instead of focusing on images that cover vast distances such as those obtained from satellite synthetic aperture radar (SAR), this research is concerned with images acquired on a local level such as those obtained from a ship based, mast mounted camera. The developed method is based on the two-dimensional wavelet transform and its ability to isolate (or preserve) directional features contained in the root image. Another important property associated with wavelets is the notion of scale or resolution. The dominant wave features of the image are extracted through the use of a multi-scale wavelet analysis. These features are then used to determine the overall wave orientation for the image. Results are shown for images taken from both traditional visual cameras and forward-looking infrared (FLIR) based cameras.