Ultra-Wideband Discrimination from a New Fused Super-Resolution, Two-Channel, Two-Dimensional Linear Prediction Algorithm

"Combining S- and X-band radar data provides a means for enhancing radar imagery by joint spatial information fusion of the radar scene for two different radar frequencies, and provides an opportunity for achieving high-resolution ultra-widebanddiscrimination between warheads and decoys. Prior to combining the two bands, preprocessing is required to account for the different data rates, co-registration of the imagery from the two radar bands, and an algorithm that is robust to differences intarget coherence times between the two bands must be developed.ORINCON proposes a new technique to combine S- and X-band radar after the data is preconditioned, interpolated and registration has been performed on the images. This technique involves a novel fused super-resolution two-channel, two-dimension linearprediction (FSRLP) algorithm, whose output is sharpened auto-spectral images of the two bands, and super resolved cross-spectral images of the fused bands. The super-resolution image is expected to provide the fidelity required to differentiate amongwarheads, cones, spheres, and cylinders.This proposal details the development of the FSRLP algorithm, which requires solving for the linear prediction coefficients, and generates the algorithm for the highly sharpened joint auto- and cross-spectral images. We believe that by adjustments in theFSRLP orders (coefficients), we can account for the target coherence times of the combined data. Comb