Research On Magnetic Partical Detection Technology Of Bearing Rings Based On Image Method

Bearing is an important component in the industrial field,whose quality directly determines the performance and stability of equipments.As one of the most effective nondestructive detection technologies for the surface crack detection of bearing rings,magnetic partical detection has the advantages of simple operation and high sensitivity.However,the current magnetic partical detection equipments for bearing rings stay in the stage that the recognition results must be identified by manual observation,which causes unstable accuracy and low efficiency.In view of the above deficiencies,a set of automatic magnetic partical detection system for bearing rings is designed,in which the machine vision and image processing technology are adopted to achieve automatic recognition.The main work is summarized as follows:1.The design of overall system.This paper analyzes the research status of the current automatic recognition system of magnetic partical detection,and explains the shortcomings of the existing equipments with bearing rings as the research object.According to the requirements of bearing ring manufactures,the overall system architecture based on magnetic particle detection and image processing is designed.2.Image processing.The images collected by CCD camera are processed by image smoothing,image enhancement and edge detection in order to extract the target information area.Because the target area is mixed with other interference factors,the mathematical morphology method is adopted to eliminate the interference factors.Then the characteristic parameters of suspected crack information are extracted.3.Crack recognition.Aimed at the problem that characteristic parameters and computation are large,the principal component analysis method is adopted to reduce the characteristic parameters from 8 to 3 dimensions.Then the support vector machine learns and classifies the three principal components to judge whether the parts have crack defects.4.The design of software platform.Based on the Halcon and VS2010 compilation environment,a crack detection software is developed,and the crack defection accuracy test has been done.This crack recognition system is debugged on the SCT-2000 magnetic partical detection experimental platform.The results show that this system is stable,which meets the needs of manufactures and make contributions to the development of automatic magnetic partical detection technology for bearing ring.