Development Of On-line Monitoring Device For Quality Of Wheel Hub Finishing Based On Machine Vision

Hub is the cylindrical magnesium aluminum alloy firmware used to support the tires inside the motorcycle tires.As the load portion of the motorcycle,the tires have the function of supporting the overall weight of the motorcycle and driving the motorcycle forward.During the operation of the motorcycle,not only the interaction force between the wheels and the ground but also the moment of the motorcycle's movement is achieved through the hub.Therefore,the integrity of the hub is an important factor in the stable and reliable operation of the motorcycle.During the production process,due to improper processing of the motorcycle wheel,holes on the surface of the motorcycle's hub,loosening,reinforcing ribs,and bumps will cause defects in the performance of the motorcycle hub,which will affect the overall product quality of the hub.Thus,the quality measurement of the surface processing of the motorcycle wheel hub is of great significance.However,the detection of the surface defects of the manufactured hubs by the manufacturers of motorcycle wheel hubs still stays at the stage of manual visual inspection,which makes the accuracy of detection and the speed of detection unsatisfactory.If the working hours are slightly longer,there will be missed detection.The situation is inefficient,and the cost is high.The manual detection is completed by human subjective delusion,subject to subjective factors,making wheel hub quality inspection results often difficult to obtain sufficient assurance.In view of the above-mentioned problems,combined with the structural characteristics of the motorcycle's wheel hub and the features of defects in the machining surface of the motorcycle's hub,this paper proposes a complete set of on-line inspection system for the surface defects of the motorcycle hub machining surface,which can achieve the surface processing of various hub machining surfaces.The existing defects are effectively identified.The development of this paper includes the design of the hardware experiment platform,the selection of experimental equipment,and the design and processing of the defect image detection algorithm.Firstly,a rotating experimental platform was designed based on the motorcycle hub features and surface defects.Based on the imaging principle analysis and specific data,the construction methods and models of industrial cameras,lenses,and illumination light sources were determined.A complete on-line inspection platform for acquiring images of various processing surfaces of the hub can be utilized.Secondly,the image processing algorithm is designed to achieve effective extraction of defect areas and quality level division,including the use of Gamma technology for image enhancement,self-adaptive local thresholds to achieve defect extraction,Sobel algorithm and Steger peak search algorithm combined with defect texture extraction,and a series of detection algorithms such as the offset difference method to extract defects.According to the defect's unique shape features and gray features,the defects and interference are divided and the detection is completed.Finally,based on the Visual Studio 2012 software development environment,the overall framework is designed,written,tested and implemented.The multi-threaded parallel operation is used to implement multi-camera communication and multi-image processing.At the same time,the friendly interaction interface between PC and human is completed as an interaction media.According to the design of the prototype of the automatic detection system,and through several tests and examinations at the site,it is concluded that the system is in compliance with the defect detection of motorcycle wheels,and its accuracy and speed are much higher than that of manual detection.