基于方向梯度直方图的螺纹连接件端面缺陷检测方法
PDF下载 (228)戚嘉辉,赵一鸣,王 壮,胡燕海.基于方向梯度直方图的螺纹连接件端面缺陷检测方法[J].宁波大学学报(理工版),2022,35(5):86-90.DOI:
QI Jiahui,ZHAO Yiming,WANG Zhuang,HU Yanhai.Detection method of end face defects of threaded connectors based on histogram of oriented gradient[J].Journal of Ningbo University(Natural Science & Engineering Edition),2022,35(5):86-90.DOI:
| Title: | Detection method of end face defects of threaded connectors based on histogram of oriented gradient |
| 作者: | 戚嘉辉, 赵一鸣, 王 壮, 胡燕海 |
| Author(s): | QI Jiahui, ZHAO Yiming, WANG Zhuang, HU Yanhai |
| 关键词: | 机器视觉; 图像处理; 梯度幅值; 边缘检测; 轮廓提取 |
| Keywords: | machine vision; image processing; gradient magnitude; edge detection; contour extraction |
| 分类号: | TP391 |
| 文献标识码: | A |
| 摘要: | 针对螺纹连接件端面表面的反光性、阈值选取的不确定性, 提出了一套图像处理方法. 使用中值滤波和图像锐化增强等方法处理图片, 解决背景与前景差距较小的问题; 选取利用方向梯度直方图确定图像上的边缘方向, 实现对目标的直线轮廓提取, 与已知目标的轮廓模型结合实现对非直线目标轮廓的提取. 该处理方法可以自动提取目标的轮廓, 而不需要设定阈值, 且对亮度、对比度和噪声变化的适应性较好. 对2500个报警器底盘做的检测实验结果表明, 采用该方法能去除噪声, 与Canny轮廓提取法相比, 能更准确且稳定地提取出螺纹连接件端面的边缘特征; 对螺纹连接件端面磨损和歪斜的检测准确率分别达到了99.4%和98.6%, 满足工业的准确率和实时性的需求; 同时可减少人工成本, 提高检测速度. |
| Abstract: | To tackle the issues of the surface reflection of threaded connectors and the uncertainty of threshold selection, a set of image processing methods are proposed, whereby median filtering and image sharpening are used to process the image to solve the problem of small gap between background and foreground. The histogram of oriented gradient is applied to determine the direction of the upper edge of the image to extract the linear contour of the target, and the contour model of the known target is combined to extract the contour of the non-linear target. This method can automatically extract the contour of the target without setting a threshold, and has good adaptability to the changes of brightness, contrast and noise. In this paper, 2500 alarm chassis detection experiments are carried out. The experimental results show that this method can remove the noise and extract the edge features of threaded connector surfaces more accurately and more stably than canny contour extraction. The accuracies of the proposed method for threaded connector’s end face wear and skew detection are 99.4% and 98.6% respectively, which meet the requirements of real-time industrial accuracy. In addition, it can reduce the labor cost and improve the detection speed. |
| 参考文献 /References: | [1] 张一明. 金属零部件表面瑕疵缺陷的视觉检测技术研究[D]. 沈阳: 沈阳理工大学, 2019. [2] 李佳伟. 基于机器视觉的小型机械零部件质量检测系统研究[D]. 常州: 江苏理工学院, 2018. [3] Choi D C, Yun J P, Jeon Y J, et al. Pinhole detection in steel slab images using machine vision[J]. IFAC Proceedings Volumes, 2009, 42(19):397-401. [4] 李顺. 基于机器视觉和深度学习的螺丝综合检测方法研究[D]. 广州: 华南理工大学, 2020. [5] 陈守恒, 谢珩, 姚启晨, 等. 基于LabVIEW的灯座缺陷检测方法[J]. 电子测量技术, 2020, 43(14):127-131. [6] 尚玉廷. 包装箱型号标记缺陷检测系统设计与实现[J]. 包装工程, 2021, 42(1):214-223. [7] 万林, 巫世晶, 谢福起, 等. 基于图像处理的输电线路耐张线夹监测系统[J]. 武汉大学学报(工学版), 2020, 53(12):1106-1111. [8] 项新建, 王乐乐, 曾航明. 基于机器视觉的推力轴承垫圈缺陷检测系统研究[J]. 中国测试, 2021, 47(2):133-139. [9] 代沁伶. 边缘保持图像滤波的应用研究[D]. 武汉: 武汉大学, 2018. [10] 葛振. 基于机器视觉的螺丝刀头自动检测系统设计[D]. 长沙: 湖南大学, 2019. [11] 廖益龙. 钢球表面缺陷检测仪控制技术及图像处理算法研究[D]. 哈尔滨: 哈尔滨理工大学, 2015. [12] Nieniewski M, Zaj?czkowski P. Comparison of ultrasound image filtering methods by means of multivariable kurtosis[J]. Image Analysis & Stereology, 2017, 36(2):79-94. [13] 周金和, 彭福堂. 一种有选择的图像灰度化方法[J]. 计算机工程, 2006, 32(20):198-200. [14] Park S, Yu S, Moon B, et al. Low-light image enhance- ment using variational optimization-based retinex model [J]. IEEE Transactions on Consumer Electronics, 2017, 63(2):178-184. [15] Tang J R, Mat I N A. Intensity exposure-based bi-histogram equalization for image enhancement[J]. Turkish Journal of Electrical Engineering & Computer Sciences, 2016, 24:3564-3585. [16] 孙光民, 孙凡. 基于双边缘模板匹配的类圆颗粒检测[J]. 软件工程, 2019, 22(4):6-9; 5. [17] 朱寒, 林丽, 王健华等. 基于改进模板匹配及图像差分法的PCB板缺陷多级检测方法[J]. 应用光学, 2020, 41(4):837-843. [18] 陈炳文, 王文伟, 杨文英. 基于图像块和边缘增强的运动目标检测[J]. 计算机工程, 2010, 36(17):192-194. [19] 张小虎, 李由, 李立春等. 一种基于梯度方向直方图的直线轮廓提取新方法[J]. 光学技术, 2006, 32(6):824-826. [20] Canny J. A computational approach to edge detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6):679-698. [21] Qu Y D, Cui C S, Chen S B, et al. A fast subpixel edge detection method using Sobel-Zernike moments operator [J]. Image and Vision Computing, 2005, 23(1):11-17. |
| 备注/Memo: | 收稿日期: 2021－06－03. 宁波大学学报（理工版）网址: http://journallg.nbu.edu.cn/ 基金项目: 国家自然科学基金（51705263）. 第一作者: 戚嘉辉（1997－）, 男, 浙江宁波人, 在读硕士研究生, 主要研究方向: 机电液产品开发. E-mail: 1005300267@qq.com 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/*通信作者: 胡燕海（1966－）, 男, 浙江宁波人, 教授, 主要研究方向: 机电液产品开发、生产计划与控制. E-mail: huyanhai@nbu.edu.cn |
