Design And Implementation Of A Defect Inspection System For Smooth Product

The detection of product appearance defects is an important part of industrial production.It is not only related to the product's value but also the company's image.Our country has become the center of global manufacturing,at present,the production lines of most companies mainly rely on manual visual inspection to detect product appearance defects.Scratches,bumps,and other defects are inevitable on the surface of a smooth part during processing.According to the characteristics of the product and its defects,we designed a set of defect detection system based on machine vision.The defect detection system can detect defects in real time,classify defects,and measure various attributes of defects,and can also realize three-dimensional display of defects.This thesis has completed the following tasks:(1)By analyzing the defect detection requirements of the smooth part surface,According to the strong reflective characteristics of the object surface and the actual detection requirements,the appropriate light source and image acquisition equipment were selected to solve the problem of insufficient image acquisition uniformity.(2)A target region extraction algorithm was explored.According to the characteristics of the acquired image,firstly,the binarization operation was performed,then the morphological erosion operation is performed,the binary image and the edgedetected image were subjected to an operation to obtain a suspected defect.(3)Morphological operations was performed on the suspected target defects.According to the product classification table,the spurious points was filtered according to the size of the connected components,and a real defects image was obtained.(4)Defects classification algorithm was designed,calculating the geometrical characteristics of each connected component,classify the defect classification category,and calculate the area and length of the defect based on the defect classification category.this was the most important step.(5)3D reconstruction algorithm was designed,firstly,the panorama of the side of the product is obtained by splicing,the side panorama is mapped on the cylinder,and the front image is mapped onto the circle at the top of the cylinder to obtain the threedimensional model of the product,you can view the position of defects at various angles.(6)Building a defect detection system and design human-computer interaction interface to visualize the defect detection process.