ОБЗОР МЕТОДОВ ОБРАБОТКИ ВИДЕОДАННЫХ ДЛЯ СИСТЕМЫ ОЦЕНКИ ДЕБИТА СТРУИ РАСПЛАВА МЕТАЛЛУРГИЧЕСКОЙ ПЕЧИ

The article deals with the problem of using advanced computer technologies in development of industrial machine-vision system for estimation of molten metal discharge from furnace. The article tells about cardinal issues connected with implementation of the system and suggests basic ways for their solution. As shown by the analysis carried out the efficient implementation of algorithm for video images processing intended to determine the molten metal flow rate requires the use of rapid image registration methods. To increase performance and speed of response of subsystem for video data processing it is suggested to use the computing power of up-to-date graphic accelerators. The article also concerns multiflow exchange of information between the central processor and the graphic processor and its effectiveness is analyzed under various aspects of data processing. It is very important for the system development that respective equipment to measure temperature and space position of molten metal flow is chosen. The hardware interaction diagram suggested in this work proved to be good in service under heavy-load conditions of an industrial enterprise.