Filtering methods for ISAR image processing

Summary form only given. This paper deals with filtering methods and algorithms for processing of images obtained from microwave hologram registered by inverse synthetic aperture radar (ISAR). The following image reconstruction filters and filtering procedure are developed: one dimensional median filter; two dimensional cross form median filter; filter with double median filtration (filtration of the rest); filter using average procedure; differential filter; edge or detailing filter; threshold processing with fixed threshold rate; and threshold processing with determination of the threshold rate by discrimination criteria. The computational procedure for root derivation of spectral coefficients and cepstral processing are worked out. The combinations of filtering procedures are employed: circumscription of a object image by root derivation of spectral coefficient; edge detection; and threshold processing. The same filtering succession is applied without the root derivation procedure. To depress pulse clutter a filter using the average procedure, and the detailing filter with threshold processing are utilized. To underline the silhouette of the object the procedure of root derivation of the spectral coefficient, threshold processing and two dimensional median filtering are applied. The numerical experiments of image reconstruction procedures are accomplished