Defect Detection Of Hub Bearing Unit Based On Machine Vision

Hub bearing unit is mainly used at the axle.Its main function is to bear the weight of the body and provide accurate guidance for the rotation of the hub.It not only bears the axle axial load force but also the radial load force.It is an important component of the automobile which integrates the load and rotation.Hub bearing unit belongs to the important parts of automobile industry.Its production,processing,factory quality and assembly requirements are very important.Once any link is wrong,it will lead to the damage of hub bearing unit.The traditional manual inspection of hub bearing defects is not only inefficient,but also greatly influenced by human factors.Moreover,it consumes a lot of manpower costs,and can no longer meet the needs of modern industrial production.Based on the modern machine vision inspection technology,with its high efficiency,high precision,non-contact,safety inspection means,easy to automate production characteristics,more and more applied to automotive parts inspection.In this paper,a defect detection system for hub bearing unit is designed based on the research of defect detection algorithm for hub bearing unit:(1)The overall design of image acquisition system.It is easy to collect the high definition inspection image of hub bearing unit,and carefully study the characteristics of industrial cameras,lenses and light sources by analyzing the specific requirements of actual inspection.Then a set of image acquisition system for hub bearing unit is designed to obtain high definition image.(2)Design and validation of algorithm for measuring roundness of inner and outer rings of hub bearing unit.By using 7 *7 calibration plate,the camera’s inner and outer parameter matrices are determined,and the image coordinates are converted to the world coordinates,so as toachieve the accuracy of measuring the inner and outer circles of hub bearings.(3)The arithmetic design of detecting toroidal characters on hub bearing unit.By collecting clear images,the polar coordinate transformation algorithm is used to detect and recognize.(4)Design of algorithm for detecting inner surface defects of hub bearings.With the rotating platform and motor driving the bearing to rotate,the clear image of hub bearing defects is acquired by camera,and the algorithm of spatial domain frequency conversion in Halcon image processing software is adopted.Combined with the basic theory knowledge of correlation function,the algorithm of detection content is designed in detail.To achieve the experimental purpose.(5)System integration and experimental analysis.Based on Halcon Image Processing Software Library and C# Programming Language,the system software is integrated and developed with modular programming thinking,thus the whole detection system is constructed.