Forming of representative set of attribute vectors for training recognition system of aircrafts types by method for simulating of distributions of scattering coefficients of their surfaces

An article is set forth a method for forming representative set of attributes for training a recognition system of types of aircrafts on the range images to be registered in an optical wavelength band. The results of simulation of attribute vector sets obtained by solving the task of multivariate optimization of scattering coefficients on finite element models of surfaces of aircrafts, as well as the results of statistical tests of neuronet recognition systems are given.