Research On Vision Inspection System For Inner Wall Defects Of Aluminum Alloy Seamless Gas

With the rapid development of economy,the gas industry has developed rapidly.Aluminum alloy gas cylinders have been used in thousands of households,which has promoted economic construction and improved people’s quality of life,but the following safety problems should be paid enough attention to.Many accidents occur not because of operational errors,but because gas cylinders are not inspected regularly in strict accordance with inspection standards,leading to enormous losses which destroy people’s lives and property.At present,most aluminum alloy seamless gas cylinder manufacturers still use the traditional mainstream method of human eye detection to detect gas cylinder defects.This traditional inspection method is inefficient and difficult to meet the production requirements.The rise of machine vision inspection technology provides a new solution for the inspection method of inner wall defects of aluminum alloy seamless gas cylinders.According to the sample gas cylinders provided by the manufacturer and the national quality inspection standard GB 11640-2011 "Seamless Aluminum Alloy Gas Cylinders",this paper designs a visual inspection system for the inner wall defects of aluminum alloy seamless gas cylinders to meet the requirements of existing productivity.The research main points of this paper concentrates on the following phases:(1)Introduce several defects on the inner wall of gas cylinders in industrial production,including common bumping injuries,scratches and stretching injuries.This paper analyzes and summarizes several methods for detecting the inner wall defects of aluminum alloy seamless gas cylinders in the market,and designs a complete visual inspection scheme for the inner wall defects of aluminum alloy seamless gas cylinders;(2)In the hardware design of the system,the gas cylinder support system,the linear array imaging system and the CIS image sensor support system are mainly designed,including hardware selection,motion mechanism design,circuit design,etc.;(3)In the research of defect detection algorithm,Firstly,image mosaic and image denoising are carried out on the collected inner images of gas cylinders,Secondly,comparing the threshold segmentation method based on amplitude and Canny edge detection method,this paper adopts the image segmentation method based on relative amplitude to extract the contour image of the defects on the inner wall of the gas cylinder,and then through a series of methods such as image pyramid,threshold segmentation,condition screening,morphological operation,etc.,the defects are screened out and marked in the original image;(4)In the design of the system software,the man-machine interface is written.Through the parameter regulation on the interface,the movement of the mechanical platform system is controlled,so that the imaging system moves stably,and the collected image of the inner wall of the gas cylinder and the position of the detected defect can be displayed instantly at the interface end.After building an experimental platform and testing the existing gas cylinder samples,the results show that the visual inspection system for the inner wall defects of aluminum alloy seamless gas cylinders studied in this paper can realize real-time online inspection of bumps,scratches and stretching defects on the inner wall of gas cylinders,and has certain practical value.