Method of processing spotlight SAR raw data

A method of processing two-dimensional spotlight SAR raw data into image data, comprising sequential steps of:

• (3.1) dividing the spotlight SAR raw data into azimuth sub-aperture data;
• (3.2) transforming the sub-aperture data, via short azimuth FFTS, into azimuthal frequency data, whereby the raw data comprises two-dimensional time data;
• (3.3, 3.4) multiplying the two-dimensional data by a frequency-scaling function Hf(fa, tr;
• ro) defined by H r ⁡ ( f a , t r ;
• r o ) = ⁢ exp ⁡ [ j · 2 · π · k r c o · β ⁡ ( f a ) · t r 2 ] · ⁢ exp ⁡ [ j · 2 ⁢ π · β 2 - 1 c o 2 · β 3 · r o · λ · k r 2 · ( t r - 2 · r ref c o ) 2 ] ( 5 )
• wherein
• fa is azimuth frequency, tr is range time, r0 is a shortest distance from a target, kr is a modulation rate of the transmitted pulses in the range, rref is a reference range, fr is a range frequency, c0 is the speed of light, λ is wavelength, and wherein β is given by β ⁡ ( f a ) = 1 - ( f a ⁢ λ 2 ⁢ V ) 2 ( 7 )
• where V is speed,
• whereby a secondary range compression of the two-dimensional data is performed;
• (3.5) transforming the two-dimensional data, via short range FFTs, into two-dimensional frequency data (azimuthal and range frequency domain);
• (3.6, 3.7) multiplying the two-dimensional frequency data by a residual-video-phase correction function HRVP given by H RVP ⁡ ( f r ) = exp ⁡ [ - j · π · f r 2 β · k r ] ( 8 )
• (3.8) transforming the two-dimensional frequency data, via short range IFFTs, back into range time/azimuthal frequency domain;
• (3.9) multiplying the two-dimensional data by an inverse frequency-scaling function Hif given by H If ⁡ ( f a , t r ) = ⁢ exp ⁡ [ π · k r · ( β 2 - β ) · t r 2 ] · ⁢ exp ⁡ [ j · 4 · π · k r c o · r ref · ( 1 - β ) · t r ] ( 9 )
• whereby a block shift is performed;
• (3.11) transforming the two-dimensional data, via long range FFTs, into two-dimensional frequency domain;
• multiplying the two-dimensional frequency data by the phase-correction function Hkoor given by H korr ⁡ ( f a , f r ) = exp ⁡ [ - j · 4 · π · k r c o · 2 · r ref c o · ( 1 - β ) · ( r ref - f r · c o 2 · k r ) ] ( 11 )
• (3.15) subsequently multiplying the two-dimensional frequency data by an azimuth-scaling function Ha(fa, fr) given by H a ⁡ ( f a , f r ) = ⁢ exp ⁡ [ j ⁢ ⁢ 4 · π λ · ( r ref - f r · c o 2 · k r ) · ( β - 1 ) ] · ⁢ exp ⁡ [ j · π k a , scL · f a 2 ] ( 12 )
• wherein ka,scl is a scaling Doppler rate;
• (3.16) transforming the two-dimensional frequency data, via short azimuth FFTs, into the range frequency/azimuthal time domain;
• (3.17) re-assembling the azimuth sub-apertures;
• (3.18, 3.19) multiplying the two-dimensional data of the azimuth sub-apertures by a de-ramping function Hder(ta) given by
• Hder(ta)=exp[j·π·ka,scl·ta2] (13)
• wherein ta is a time in the azimuth direction, to obtain de-ramped data;
• and
• (3.20) transforming the de-ramped data, via long azimuth FFTs, into the image data.