Research And Implementation Of Wheel Image Segmentation And Defect Recognition Algorithm

With the continuous development of the automobile industry,the safety problem of automobile driving is increasingly concerned.In the aspect of safe driving,the quality of wheel hub plays an important role.At present,most of the domestic aluminum alloy hub manufacturing adopts casting technology,and most of them are integrated casting,with relatively low cost and strong ability to resist metal fatigue.The production process is to design the mold first,and then cast the molten alloy into the mold.However,due to the influence of cooling system,process parameters and other factors in the forming process,it is easy to produce internal defects,thus reducing the quality of products.In the process of automatic inspection of wheel hub,X-ray non-destructive inspection is a kind of defect identification method with good effect,and the automatic segmentation and recognition algorithm of inspection image plays an important role.In this paper,aiming at the problem of automatic segmentation and recognition of hub defects,the application of traditional algorithm and neural network algorithm in hub defect recognition is investigated.In this paper,the parameters of the traditional algorithm are designed and tested,the best parameters are found and the traditional algorithm is realized to detect the wheel hub defects.At the same time,the neural network model is trained to automatically identify the wheel hub defects,and the recognition effect of the two algorithms is compared and analyzed.Finally,the image defect recognition system of automobile hub is constructed,which mainly includes manual recognition,automatic recognition,calculation of defect area and other functions.In this paper,the traditional methods based on mathematical morphology and neural network technology are used to segment and identify the defects in the image obtained from the X-ray inspection of automobile wheel hub,which will provide some reference for the future development of the field of automatic identification and detection of automobile wheel hub defects.At the same time,this paper will build a car hub image defect recognition system,which can help the car hub image defect detection personnel to detect the car hub image defects faster and better.