To make better use of wind streak information from Synthetic Aperture Radar（ SAR） images used in turn to obtain high-precision sea surface wind fields

wind direction is retrieved using decimated wavelet transform associated with two-dimensional Fast Fourier Transform（ FFT） with different modes and Advanced Synthetic Aperture Radar（ ASAR） images of various p olarizations. Finally

the resulting wind direction is compared with those obtained using traditional FFT and the i mproved local gradient method. The results demonstrate the feasibility of applying the wavelet transform algorithm to sea surface wind direction inversion. Multi-resolution analysis of ASAR images with time-frequency window of wavelet transform enables more precise determination of wind direction compared with traditional Fast Fourier transform and improved local gradient method. In addition

the inversion results contain certain differences in diverse modes and various polarization data.