A new automatic internal wave detection and characterization method for SAR images

An efficient method for automatic extraction and characterization of internal wave (IW) packets present in SAR images from the ocean surface is described in this paper. The utility of wavelet analysis as a tool for the detection of mesoscale oceanographic phenomena is examined using the wavelet transform (WT) based on the multiscale edge detection method. We show that the evolution of local maxima of the WT coefficients across scales characterize the local shape of these quasi-linear and periodic structures. Radon transform (RT) techniques are used to detect and orientate IW features since they seem to be well suited for extracting quasi-linear features even at low signal-to-noise levels. Results from this study show that wavelet and Radon transfom techniques are excellent tools to detect, orientate, and characterize IW features from SAR images against IW look-alikes.