Research On Detection System Of Clutch Release Bearing Unit Based On Machine Vision

The clutch release bearing unit is the hub between the automobile engine and the transmission system,and is a key component to ensure the smooth start,stop and shift of the vehicle.In the process of separating the production of bearing units,enterprises have strict requirements for the quality inspection.The traditional manual detection methods not only have low efficiency,low precision,but also cost a lot of labor costs,which can no longer meet the requirements of modern industrial production.Machine vision-based inspection technology is increasingly applied to the inspection of automotive parts due to its non-contact,high efficiency,high precision and easy to automate production.Based on the industry standard and enterprise requirements of clutch release bearing unit,this paper proposes an automatic recognition and classification detection system based on machine vision.The main research work is as follows:Firstly,the defects of the four separate stations for the character of the clutch release bearing unit,the single inner diameter(abbreviated as the inner diameter),the annular surface of the upper end surface in contact with the clutch platen(referred to as the upper ring surface),and the diameter of the contact circle and the width of the unit are designed.The test is performed and the system structure diagram is given.In thehardware design of the system,the selection of camera,lens and light source is elaborated to ensure a stable and convenient image recognition imaging system;Then,in the image detection of the clutch release bearing unit,the image recognition operator in HALCON is combined with the basic theoretical knowledge of the correlation function to carry out detailed algorithm design for different detection contents.Using the Visual Studio programming platform,the c# and HALCON mixed programming method is used to design a human-computer interaction interface that is easy to operate and display.Finally,the test content of each station was tested experimentally.The test results show that the detection accuracy or detection accuracy of each station meets the requirements of the enterprise.