Characterization of hurricane eyes in RADARSAT-1 images with wavelet analysis

Striking examples of RADARSAT-1 synthetic aperture radar (SAR) images of hurricanes have been acquired over the past few years. The images show, with high resolution, the imprint of these storms on the ocean surface roughness, including structures associated with atmospheric processes such as boundary layer rolls, and details associated with the eye of the storm. In this paper, an image-processing procedure for quantitatively characterizing SAR images of hurricane eyes (HEs) is described. The procedure uses the edge detection properties of wavelets to estimate the scale and area of HEs. Procedures are also introduced to determine a reference ellipse, the location of the centre, and an elliptical index. All parameters are measured quantitatively and objectively. Provision of a universal characterization procedure for SAR images of HEs will promote the use of RADARSAT-1 SAR images for the study of hurricane morphology and dynamics.