Analysis Of Ferrography Image Based On Level Set Method And Gray Target Theory

The wear of mechanical equipment occurs at every moment of equipment operation.The wear will make the failure of the parts of equipment,reduce manufacturing accuracy and severe wear may cause major accidents.The wear not only causes the waste of material and energy,but also may affect the personal safety.Therefore,in order to ensure the normal and safe operation of the equipment,the condition monitoring and fault diagnosis are essential.Ferrography analysis technology is an effective technique for condition monitoring and fault diagnosis.In order to improve the automation and intelligence level of the ferrography analysis technology.In this paper,The corresponding algorithm is proposed by studying the method of segmentation and analysis of Ferrography Image for the automatic segmentation of wear particles and wear state recognition.And it verified by the actual ferrography image and comparative analysis.The main work of this paper is as follows:1.According to the characteristics of Ferrography Image,the level set method is used to segment the wear particles segmentation.The Chan-Vese(CV)model in the level set method is studied,which is used to realize the accurate segmentation of single large wear particles in the Ferrography Image.But the model can only be used for the segmentation of uniform background images,which need many iterations and have slow speed.For the slow speed of CV model,Local and Global Fast Mode(LGF)model is constructed by local information is introduced into the energy functional,and the term of energy penalty are added to the regularization terms.The model solves the problem that the CV model needs to be re initialized and have slow speed of evolution.In order to realize the segmentation of Ferrography Image with complex background and various types,a new method of wear particles segmentation based on LGF model and CIE L*a*b* color space is proposed.In this method,the RGB color space image firstly is transformed into Lab color space image,then the LGF model is used in B channel image,so as to realize the automatic and accurate segmentation of wear particles.2.Wear condition recognition based on grey target theory is proposed.In this method,the accurate wear particle concentration in the range of the ferrography image is obtained by the segmentation algorithm;Grey Target Approaching Analysis is carried on pattern sequence,which is established by wear particle concentration;The contribution of each size range is calculated by the contribution analysis and the calculation of coefficients can be calculated by using the contribution degree;Finally,approaching degree be modified by the approaching degree coefficient calculation.Compared with the wear state of equipment,the wear state of the equipment is judged.In this paper,we use Microsoft Visual Studio 2010 as the programming software platform,C++ as the programming language,and call the OpenCV function library to implement the method.The experimental results show that the method proposed in this paper is feasible,fast and effective.