The Research Of Recognition And Classification System For Cast Aluminum Alloy Wheel Hub

At the present,China's automobile ownership has been in the forefront of the world,the annual automobile production volume reached 28 million units in 2016,so,the wheel hub is in great demand.In China,most of aluminum alloy wheel hub for automobile production still use manual classification method,which is low efficiency and difficult to realize the informationization,can not meet the requirement of modern management,so we need to use automatic identification technology to eliminate the classification process of manual intervention,improve the classification efficiency,and obtain information data.By using machine vision and image processing technology,we can achieve automatic recognition and acquire the digital information of wheel hub.According to the hub production factory in Hubei,this paper first uses 3D design software Pro/Engineer to design the automatic recognition device,and select the appropriate MCU to control the hardware system,ensure the stable operation of the device;Secondly,we develope a hub casting automatic recognition system on the basis of the Windows Visual Studio software platform and the image processing library OpenCV.The median filtering,histogram equalization,morphological processing pretreatment method eliminate the influence of wheel hub blank which has large burr,noise and some illumination.Then we extract the outline of the radius of the circle,the spokes root number,center hole,bolt hole number,spoke window shape feature,the spoke window polygon simplification method is a good solution to the problem of some hub castings with large burr;At last,a comprehensive classification sequence file name recognition based on shape feature has been put forward for hundreds of different types of wheel hub production plant,by means of establishing appropriate rules for hub image's template library,the system automatically generates file name in the serial number for the images,when traversing the hub type template library file names in the directory,we can determine the target wheelmodel,which make the recognition process simple,efficient and practical.The experiment shows that the self-developed aluminum alloy wheel hub automatic recognition system can recognize different types of wheel castings,obtain the digital information in the casting,in order to follow the different types of wheels with different treatment,this method can greatly reduce the artificial error rate and provide high efficiency for the mass production process,meet the needs of automation and intelligence.