Phase unwrapping is the key problem in building the elevation map of a scen
e from interferometric synthetic aperture radar (SAR) system data. Phase un
wrapping consists in the reconstruction of the phase difference of the radi
ation received by two SAR systems as a function of the azimuth and slant ra
nge coordinates. The data available to reconstruct the phase difference are
a measure of the difference module 2 pi, We propose a phase unwrapping met
hod that makes use of the equivalent, in a discrete space, of the irrotatio
nal property of a gradient vector field. This property if used first to loc
ate the areas where the discrete vector held estimated from the available d
ata must be corrected, and then, with the knowledge of some a priori inform
ation, to perform the correction needed to obtain a useful estimate of the
discrete gradient of the phase difference function, from which the phase di
fference function is reconstructed. The use of the fast Fourier transform m
akes it possible to have a fast algorithm, that is to process an image of N
pixel in 0(N log N) elementary operations. Tests of the method proposed he
re on real and simulated data are presented.