Research On On-line Recognition System Of Wheel Hub Type

It is very important for the development of wheel hub enterprises to realize the automatic recognition of wheel hub models on the production line.At present,the production of wheel hub is artificial recognition,but the artificial energy is limited and the recognition efficiency is not high,which can not meet the production needs of wheel hub.Therefore,it is very necessary to develop an efficient and accurate on-line identification system for wheel hub models.Machine vision recognition technology has many advantages,such as non-contact,on-line real-time,strong anti-interference ability.Its biggest characteristics are fast speed,large amount of information and many functions.It can realize the precise classification of products,adapts to the progress and development of modern industry,and shows a broad application prospect in practice.Based on this,this paper combines machine vision technology and image processing technology to achieve various types of car hub classification matching for the research object.The design idea of the wheel type on-line recognition system is to combine the machine vision technology and image processing technology to realize the accurate recognition of many kinds of automobile wheel models,and design the wheel type recognition system from both hardware and software aspects.The hardware part of the system mainly realizes the image acquisition function,using the CCD camera,photoelectric switch,industrial computer and image acquisition card to form the hardware system to obtain the wheel hub image.The hardware system circuit is designed to realize the control of the wheel hub position and the remote image transmission.The image processing part uses the mean filter and the median filter to remove the noise,the gray stretch method to enhance the image,and the OTSU method to segment image.The edge detection Canny operator is improved by using the optimized Canny operator to extract the edge feature of hub.At the same time,the features of wheel hub image are extracted.Template matching algorithm is used to recognize the wheel hub model by using the feature information extracted above.Compared with the traditional recognition technology,it improves the precision of recognition and improves the level of production automation to a great extent.The system software is developed in VC++ 6.0 environment.The function of image acquisition,image processing,output display and printing and auxiliary function are realized with modular structure,respectively.Using different types of wheels hub for identification tesing.The test results show that the system can complete the on-line high-efficient and accurate identification of hub models,solve the problems existing in manual recognition,and improve the production efficiency and production automation level of the hub.