Research Into The Achievement Of Class Of Flotation Froth Image Through Image Processing

In the first instance, the characteristics and the problems of flotation technics are researched and analyzed, the feasibility of the application of digital image process in flotation control is explained and the research status quo in the aspect is showed. Authors indicate the meaning of the paper. Subsequently, based on the analyzation of the image processing technology and the froth image, the pretreatment is done to the flotation image. Authors select several characteristics regarded as input variable from image characteristcses. To deign a viable froth image recognition system, engineers has to consider the practical situation of production process. So this paper introduces rough set and LVQ neural net to build the froth image recognition model to identify the states of flotation. In the system, the application of rough set theory is to predigest the decision table and the application of LVQ neural net is to recognition all sorts of images belonging to different types through studying rough set and LVQ neural net arithmetic. At last, the system can content the requirement of production control.