L.I. Dvoiris – Dr.Sc.(Eng.), Professor

K.D. Galev – Post-graduate Student

The article deals with a combination of a computing core (microprocessor) and a set of algorithms in the development of automated systems for detection and recognition of moving objects by seismic signals. The rationale for choice of the microprocessor will allow classifying them according to different detection and recognition algorithms as well as identifying the dependence of the influence of the type of architecture on the quality of detection and recognition of moving objects.

Currently, the developers of the detection (hereinafter referred to) much attention is paid to seismic. This is determined by a number of factors: the need for rapid blocking of local and extended areas, low cost and relatively high detection characteristics, the possibility of using microprocessors with low power consumption. The existing algorithms of information processing of seismic detection means are characterized by high energy costs, while to ensure the specified indicators of recognition quality during the entire period of operation of the tool requires a classification of sources of seismic effects. The construction of a microprocessor selection model in the development of detection and recognition subsystem is relevant for solving the problem of choosing the optimal (by a number of criteria) combination of detection, feature extraction and recognition algorithms together with one of the many available microprocessors. Justification of the choice will allow to classify the different microprocessors on the various algorithms of detection and recognition, and to identify the dependence of the influence of the type of architecture on the quality of detection and recognition of moving objects.