Exploiting fundamental properties of SAR data for compression of tactical SAR imagery

This paper introduces a technique that exploits the statistical properties of complex synthetic aperture radar (SAR) data to effectively compress tactical SAR imagery. The technique combines Fourier transforms and analysis/synthesis algorithms into a hybrid compression architecture that exploits the temporal correlation of the complex radar returns. In addition the technique incorporates the fact that the viewable SAR image is based on the magnitude of the complex imagery. This enables a vocoding technique that can preserve the frequency information and discard the phase information with little perceptual degradation in the image content. Finally the holographic property of SAR is also exploited. The performance of the new algorithm is compared with the JPEG compression standard. The results show that compression ratios of over 100:1 are achievable and far surpass the compression capabilities of the conventional JPEG algorithm