Data compression method and apparatus for radar image formation and like

Abstract

Synthetic aperture radar (SAR) images are formed by storing in electronic memory possible outputs of a lossy data compressor, f.sub.1,1 [m], f.sub.1,2 [m], . . . f.sub.1,P [m], where f.sub.1,i [m] represents a segment of the unprocessed radar return; by storing range-correlated segments of the SAR data, g.sub.1,1 [m], g.sub.1,2 [m], . . . g.sub.1,P [m], where g.sub.1,i [m] is the range-correlated version of f.sub.1,i [m]; by storing all possible outputs of a second lossy data compressor, f.sub.2,1 [n], f.sub.2,2 [n], . . . f.sub.2,Q [n], where f.sub.2,i [n] represents a segment of the range-correlated SAR image; and by storing segments of the range-azimuth focussed SAR image data, g.sub.2,1,m [n],g.sub.2,2,m [n], . . . g.sub.2,Q,m [n], where g.sub.2,i,m [n] represents a segment of the SAR range-azimuth focussed image. Thereafter, a sequence of separate radar pulses are transmitted from an aircraft and the return signal is sampled and placed in an array. Vectors of the array are compared with segments f.sub.1,i [m], and the indices of the segments which most closely match the vectors are stored in an array. An intermediate image is formed by adding together segments, g.sub.1,i [m] whose indices match those in the array, while maintaining a predetermined offset. Vectors of this intermediate image are compared with segments f.sub.2,i [n] and the indices of these segments which most closely match the intermediate range vectors are stored in a second array. The final image is formed by adding together segments, g.sub.2,i,m [n], whose indices match those in the second array, while maintaining a predetermined offset between each segment during the addition.