A projected WRLA super-resolution algorithm for PMMW imaging

Images acquired from passive millimeter-wave (PMMW) radiometer have poor resolution due to limited aperture dimension. Therefore, super-resolution algorithm (SRA) is a key component for PMMW imager. SRA realizes spectrum extrapolation, however, spectrum in-band is deteriorated too. Wiener-Richardson-Lucy algorithm (WRLA) recovers spectrum in-band by Wiener filter, and extrapolates spectrum out-of-band by Richardson-Lucy algorithm (RLA). WRLA can avoid spectrum deterioration in passband, but its ability for spectrum extrapolation is limited also. We propose an improved algorithm to solve this problem. This algorithm, which has built constraints according to characteristics of PMMW imaging, uses projection-onto-convex-sets (POCS) technique to enhance the WRLApsilas performance, so it is termed as POCS-WRLA.