We present the follow-up of our previously published work, where we described a wavelet-based method to characterize the sea surface backscatter structures in Synthetic Aperture Radar (SAR) images. The method relies on the ability of the 2-D continuous wavelet technique to detect the spatial structure of the Marine Atmospheric Boundary Layer (MABL) and to isolate wind-related cells and features. The analysis of the cells' geometry, molded by the radiometric characteristics of the sea surface, permits the identification of the wind direction inside the cells, due to the along-wind asymmetry of backscatter structures, and thus the computation of the wind speed through standard algorithms. Twenty-one SAR images (ERS-2 and Envisat ASAR Wide Swath) over the Mediterranean Sea have been analyzed, and the results are compared with satellite wind fields. The images cover a range of meteorological conditions from low to moderate winds. Comparison of the SAR-derived wind fields with those provided by satellite scatterometers indicates a good score of success (roughly 70%-80%). The developed methodology, once tested over an adequate number of images to derive statistically reliable results, could be routinely used to enrich SAR images with the wind field as well as to characterize the MABL in terms of size, distribution, and shape of the backscatter cells.