Research On Image Processing Technology For Feature Extraction Of Flotation Foam

Flotation in metal production is an important link,under certain conditions,by adding an appropriate amount of reagent and the amount of air,and then you can start stirring,stirring process will appear a lot of bubbles,finally through to the recycling of these bubbles,to extract the ore from the bubble.Flotation froth characteristics such as size,shape,color.Flotation bubble size,flotation foam between adhesion sex is stronger,complex surface texture.Flotation bubble flotation process mostly rely on artificial observation,but also can cause unnecessary harm to human body,therefore,result in flotation process is difficult to optimize.This paper mainly introduces the artificial control of flotation can cause a waste of resources,the harm of human body,and to reduce the resource recovery and other issues,in order to solve these problems,we consider to replace the manual control with computer image processing technology,which requires to apply image processing technique to the flotation foam image processing.For flotation froth images,we visited the flotation workshop,the acquisition by flotation bubble image,at the same time introduces the flotation foam of the complexity of the process of image processing,and bubble image processing difficulty,if successful,will greatly improve the metal recovery,reduce the production cost,reduce waste of resources,and reduce the demand for artificial,this also can avoid flotation workshop environment,the harm of human body to improve overall production efficiency,improve the flotation,non-ferrous metal recycling industry production efficiency.So as to promote the development of the whole national economy,and also can enlarge the application of computer image processing technology development space,promote advanced computer image processing technology.In the thesis we mainly from the flotation foam image acquisition,the flotation foam image denoising,followed by then the flotation foam flotation bubble image segmentation image feature extraction has carried on the detailed description of this process,and in view of the flotation foam image feature extraction of three aspects:size,shape,color to make detailed explanation.The flotation process requirement for real time again at the same time,we consider introducing GPU and threads to improve the efficiency of flotation bubble image processing technology.After two years of efforts,with the help of the teachers and students mainly implements the following work:In this paper,according to the characteristics of the flotation foam,analysis to extract the size of flotation bubble,shape and color.This thesis mainly research work are as follows:(1)through and multi-scale wavelet analysis theory,the characteristics of the size of flotation bubble is extracted.(2)by studying the flotation foam image segmentation technology,watershed segmentation algorithm,for example,to extract the flotation image shape feature.(3)based on the analysis comparison of flotation image RGB color value distribution,image color feature extraction of flotation.(4)optimize the relevant procedure,using the GPU based on CUDA platform to enhance the real-time image processing.