Research On Image Recognition Of Flotation Forth About Concentrated Mineral's Grade

The development and application of digital image technology have been summarized in this dissertation. The application and algorithmic research of foreign and domestic digital image processing in flotation operation have been introduced and predicted.The Flotation Froth Image Recognition System (FFIRS) has been exploited under Microsoft Windows environment, which is based on the object-oriented programming method and the tool of Visual C++ 6.0 language. The software system mainly consists of distributed check, point operation, orthogonality transform, edge detection, lineament extraction, image operation, physical parameter, recognition result. An extendible and open code is adapted to the system. The limitation of conventional memory for large image is broken through by inheriting base classes of Visual C++ and overriding relevant virtual functions.The images of flotation forth and extractive mineral obtained from Tungsten Flotation Factory in XiangAn have been analyzed, and the texture parameters (entropy, energy, moments and average of gray level) have been distilled. We have established mathematic models between the texture parameters in the image and the grade of concentrated mineral in flotation forth with MATLAB language. There exists linear relation between the texture parameters and grade of concentrated mineral by using regression analysis.Detecting the flotation froth images of concentrated mineral whose grade is higher than 5 percent by using established mathematic model, the relative error between forecasted grade and original grade is no higher than 1.45 percent and the average relative error is 0.648 percent. Detecting microphotograph of concentrated mineral by using established mathematic model, you will find that the effect of recognition of microphotograph in-concentrated mineral is better than the effect of recognition of flotation froth. The reason why there is an error between forecasted grade and original grade has been analyzed. The reason why the effect of recognition of microphotograph in concentrated mineral is better than the effect of recognition of flotation froth has been found.FFIRS can elementarily detect grade of concentrated mineral online or outline. It has been proved by using image recognition of flotation froth and image recognition of microphotograph in concentrated mineral.