Research Of Seamless Steel Tubes Diameter Measurement And Surface Defect Detection System

Seamless steel tubes are a highly practical section steel,which has a great demand in production and life,such as petroleum,boiler,power station,ship,machinery manufacturing,automobile,aviation,aerospace,energy,geology,construction and military.The diameter and surface defect information of seamless steel tubes is an important criterion for judging the quality of seamless steel tubes and is the most important part of enterprises and customers.Since the 13 th Five-Year Plan,the market's demand for seamless steel tubes still growing,and the seamless steel tubes Detection methods are hard to meet the production requirements of enterprises.In this paper,a set of equipment for image acquisition of cylindrical seamless steel tubes using a single area array camera combined with a rotary advancement scheme is designed,and can be completed simultaneously by a single acquisition for the diameter measurement and surface defect detection.First of all,design of the Image acquisition device.Through the analysis of the shape characteristics of the seamless steel tubes and the requirements of the inspection task,a rotating forward image acquisition device is designed for the seamless steel tubes.The problem of surface reflection and retaining boundary information of the steel tubes is solved by using a combined light source composed of a double strip light source and a parallel light source,and the stability and expandability of the image acquisition device are also increased.Second,hardware selection of the Image acquisition device.The work of hardware selection mainly considers the cost problem in actual detection,and should to reduce the hardware cost in the detection requirements are met.The model of the image acquisition hardware is determined by analyzing the main parameters of the camera,lens and light source.Then,positioning boundaries and diameter measurement of seamless steel tubes.The edge positioning of the seamless steel tubes in the original image is the basis for the diameter measurement and surface defect detection.The boundary of the seamless steel tubes is completed by algorithms such as image segmentation,gradient calculation and straight line fitting,and the exact diameter of the seamless steel tubes is further obtained by the boundary equation,and the measurement error is less than 0.01 mm.Finally,detection of seamless steel tubes surface defects.Due to the different surface roughness of the original image,it is difficult to directly separate the two types of defects on the surface.In this paper,the surface defect detection of seamless steel tubes is completed by extracting the suspected defect area and using the support vector machine classifier to classify the suspected area,and the correct rate of defect detection is 96%.The testing standard for enterprise production requirements is that the measurement error of diameter is less than 0.01 mm,and the detection rate of surface defects of 0.2mm×0.2mm or more is higher than 95%.The detection algorithm of this paper has met enterprise's requirements for the measurement of the diameter of seamless steel tubes and the detection of surface defects.