А.N. Detkov - Dr.Sc. (Eng.), Professor, Research Scientist, FGBU «Chief Research and Testing Center robototehnicks» Defense of Russian Federation. E-mail: detkov_an@mail.ru А.V. Nikonenko - Head of Department, FGBU «Chief Research and Testing Center robototehnicks» Defense of Russian Federation. E-mail: nikon-74@mail.ru D.А. Nitsak - Ph.D. (Eng.), Head of Department, VUNTS Air Force «VVA them. Zhukovsky and Yury Gagarin», Voronezh. E-mail: dima_nitsak@mail.ru

In perspective radar systems using automatic target detection, selection fluctuating small-sized targets on the background of clutter is a fairly complex task. Difference algorithm is proposed for processing vector-polarimetric observations of the SAR, which will surely detect fluctuating polarization of featured small targets against the background of fluctuating passive interference. It is shown that of all the possible FST can extract two specific (singular) of the target - the most visible and most hidden in the clutter, which generally improves the detection of small targets. The main limitation of the use of differencing algorithm technique polarimetric SAR remains a need for an adaptive assessment matrices coherence distinguished fluctuating purposes that involves considerable computational cost, especially at the initial stage of convergence.