Computational strategies for synthetic aperture imaging

Synthetic aperture radar (SAR) image reconstruction based on a convolution backpropagation (CBP) algorithm is discussed. The authors describe how the CBP algorithm can be modified and applied within the framework of constraints imposed by narrowband spotlight-mode SAR. With this approach, it is possible to begin producing the image as soon as the first return (projection) is recorded. An important advantage is that only 1-D interpolation is required, as compared to 2-D interpolation that is required with direct Fourier methods. Also, it is easy to produce a partial spatial domain image if it is determined part way through the processing that all targets of interest lie in a subregion of the image plane, or to produce a reduced resolution image while saving computation. The CBP SAR processing algorithm is demonstrated on real SAR images consisting of corner reflectors, and the resulting images are compared with equivalent Fourier domain reconstruction