Based On The Infrared Video Image Of The Slag Detection System

In this dissertation, combined with the status quo of internal steel industry and deep investigation to internal and oversea slag detect technologies, the slag detection method based on infrared monitor has been developed. After deeply investigating the video character around the pouring place on the surface of the molten in the tundish, we designed a set of methods for the recognition and analysis of slag video characteristic parameters, and have developed the antitype system.This system uses infrared camera and video card to collect the still image. And we identify characteristic parameters by processing and analyzing those images. Then we convert those original parameters to network parameters as the input to the BP-ANN. Through the parameters analysis with the ANN algorithm, system automatically perform precise judgment of appearance of slag. The software development of this system is on the Win2000 OS platform, adopting OOD to develop all of the antitype system. In addition, we used the DirectX, COM and DLL technologies to separate the collecting part and algorithm component from whole architecture of the system for conveniences of future maintenance and probably transplanting.This system has passed tests in the laboratory, and been testing in the locale of foundry. As far as it goes, the system has achieved preferable result, and can meet the expectant goal.