Adaptive phase-singular-unit restoration with entire-spectrum-processing complex-valued neural networks in interferometric SAR
A singular-unit restoration filter based on complex-valued neural networks that deal with spatial spectrum in interferometric synthetic aperture radar is proposed. This filter utilises more neural generalisation ability than conventional methods. In experiments, it shows a higher accuracy as well as shorter processing time than conventional real-space filters and shorter learning time than previous spectral-domain learning filters.