Font Size: a A A

Defect Detection Of Sheet Metal Parts Based On Machine Vision

Posted on:2022-10-07Degree:MasterType:Thesis
Country:ChinaCandidate:Z H TangFull Text:PDF
GTID:2492306731975739Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
With the advent of intelligent manufacturing 2025,Machine vision is more and more applied to industrial inspection,as an important part of automobile,the quality of sheet metal is very important for the quality of sheet metal,in the production process,the output of sheet metal parts is large,and the vast majority of detection are based on manual visual inspection,the traditional detecti on methods can not meet the needs of modern industry,as a new detection method,machine vision can achieve non-destructive and rapid detection,which has attracted more and more attention and research of researchers.In this paper,the surface defects of sheet metal parts are detected by machine vision technology,and designed a set of user interaction detection system,the main contents are as follows:First,the hardware of the detection system is built,mainly on the selection of camera,lens and light source,we get a set of hardware equipment with moderate cost and can meet the requirements of sheet metal parts detection.In software aspect,the detection algorithm is designed by using the Halcon software,and the user interface system is written by C # language in vs2019.This paper studies the language compatibility between the operators in Halcon and the software system developed by C #,the availability of the system is guaranteed.Second,this paper mainly studies the scratch and stain defects of s heet metal parts,the image filtering algorithm is used to denoise the sheet metal surface,for scratch defects,in order to reduce the impact of surface reflection on detection,the improved diffusion algorithm and image difference algorithm are used to e xtract defects,using corrosion expansion operation to remove the influence of noise;The local adaptive threshold segmentation algorithm is used to detect the spot defect.The test results show that the method is effective and feasible.Finally,In order to solve the problem that image acquisition and image detection can not be carried out at the same time,the software system is improved by using thread pool technology.The experimental results show that the software system is stable enough,which has good practical value.
Keywords/Search Tags:Sheet metal parts, Machine Vision, Defect Detection, Thread pool
PDF Full Text Request
Related items