Oleshchuk O. Analytic and informational models and methods of moving objects identification for control systems of freight flows

The thesis is dedicated to development of theoretical bases and facilities of construction of the automated system of railroad carriages authentication and registration (ASRCAR). It is shown that construction of ASRCAR on the base of receiving and processing of visual information about the vehicles is possible and economic justified. The analytical models of transformation of interlaced images to progressive image are developed. The edge sensing methods are improved due to the using of discovered statistical dependences of brightness correlations on the identification elements borders. The method of template-dependent processing is developed for recognition at the high level of noise. Hardware and software realization of ASRCAR is created on the basis of the offered models, methods and algorithms. An industrial prototype showed ASRCAR efficiency and availability of its introduction for automation of control systems of freight flows.