The Study Of Needle Defect Detection System For Needle Roller System Based On Machine Vision

Machine vision technology has a strong anti-jamming capability, high speed,good stability, and non-contact, etc. With the development of science, the testingrequirements for products are becoming more automated, but the requirement for thequality of product is continuously improving. Traditional detection methods can notmeet the requirement of the modern enterprise detection, machine vision, as a newdetection method, is being more and more widely applied in industry.This paper focuses on the needle leakage phenomenon during the needle rollerbearing assembly process, developing a kind of bearing needle leakage detectionmethod based on the machine vision and digital image processing technology,achieving the leakage detection judgment of needle bearing, ensuring the qualified rateof products, improving production efficiency and quality of bearing.The method adopts graphical programming software LabVIEW as a designplatform, the circular low Angle LED red light is available for the hardware platformas lighting devices, using the low illumination1/3"VS-250DM interlaced scanningCCD black and white camera of Micro-vision, VS-0660M optical lens and MV-U2000portable USB bus image acquisition card as image input device, with a computer forprocessing and as the output device. With the auxiliary of lighting devices, imageinformation of needle roller bearing is collected by CCD camera, and beingtransmitted to the image acquisition card, and then through the A/D conversion totransfer the image data information into the computer, after the settlement and analysisof the image information, output the distinguishing results. The procedure adopts thevisual function Vision8.2development kit of the LabVIEW software for imageacquisition of needle roller bearing, processing and recognition, and so on. Theprocess conducts pretreatment such as noise filtration of the collected images, linearenhancement, histogram equalization, bearing location, and finalization, etc. and thenapplies the improved template matching algorithm, based on a single needle image applies the improved template matching algorithm, based on a single needle image template, finish the needle image matching in the designated area, and output matching results, according to the results, to judge whether there’s needle defect of bearing, and occupies functions of needle defect alarm and calculating the needle defect number etc.The experiment verifies the feasibility of systematic procedure. The needle roller bearing lacked needle detecting method presented in this paper, when it tests the needle roller bearing, owing the software friendly interface, easy operation, accurate test results, quick speed, and good stability. In addition, it overcomes the disadvantages of traditional detection methods, and improves the testing efficiency and accuracy.