首页|基于机器视觉的刚性罐道接头缝隙宽度测量算法

基于机器视觉的刚性罐道接头缝隙宽度测量算法

扫码查看
原文链接
  • 国家科技期刊平台
  • NETL
  • NSTL
  • 万方数据
  • 维普
针对罐道结构复杂、空间狭小、传统检测系统不便于安装,容易导致接头缝隙测量困难问题,提出一种基于机器视觉的非接触式测量接头缝隙宽度的方法.使用高斯滤波和CLAHE算法,去除噪声并提高图像对比度,经过阈值分割、图像形态学操作提取接头缝隙边缘特征;通过孤立点消除和透视变换去除图像畸变,使用SKE-骨架提取算法减少接头缝隙特征像素宽度,通过霍夫变换和最小二乘法计算接头缝隙宽度.测量了 4～12 mm接头缝隙宽度,结果表明:该方法测量误差控制在 0.8 mm以内,为立井提升系统安全稳定运行提供保障.
GAP WIDTH MEASUREMENT ALGORITHM FOR RIGID TANK JOINTS BASED ON MACHINE VISION
In view of the complex structure of the tank road,the narrow space,the inconvenient installation of the traditional detection system,and the difficulty in measuring the joint gap,a non-contact method for measuring the width of the joint gap based on machine vision is proposed.Use Gaussian filtering and CLAHE algorithm to remove noise and improve image contrast,extract joint gap edge features through threshold segmentation and image morphology operations;remove image distortion through isolated point elimination and perspective transformation,and use SKE-skeleton extraction algorithm to reduce joint gap feature pixel width,the joint gap width is calculated by Hough transform and least square method.The joint gap width of 4-12 mm is measured,and the results show that the measurement error of this method is controlled within 0.8 mm,which provides guarantee for the safe and stable operation of the shaft hoisting system.

machine visionrigid tank channelwidth measurementimage processing

马天兵、彭猛、杜菲、纵书棋、方佳欣

展开 >

安徽理工大学深部煤矿采动响应与灾害防控国家重点实验室,安徽,淮南 232001

安徽理工大学机械工程学院,安徽,淮南 232001

机器视觉 刚性罐道 宽度测量 图像处理

国家重点实验室资助项目安徽省自然科学基金面上项目矿山智能技术与装备科研创新团队项目安徽高校学科拔尖人才项目安徽高校协同创新项目安徽省重点研究与开发计划项目

SKLMRDPC20ZZ012008085ME1782022AH010052gxbjZD202020063GXXT-2022-019202104a07020005

2024

10.3969/j.issn.1674-8085.2024.02.012

井冈山大学学报(自然科学版)
井岗山大学

井冈山大学学报(自然科学版)

影响因子：0.298
ISSN：1674-8085
年,卷(期)：2024.45(2)
  • 17