[1]肖红,陈希球.基于图像角点匹配的机械加工零件表面缺陷检测[J].机械与电子,2022,(08):62-65.
　XIAO Hong,CHEN Xiqiu.Surface Defect Detection of Machined Parts Based on Image Corner Matching[J].Machinery & Electronics,2022,(08):62-65.
点击复制

基于图像角点匹配的机械加工零件表面缺陷检测()

分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

Title:: Surface Defect Detection of Machined Parts Based on Image Corner Matching

Author(s):: XIAO Hong¹ ; CHEN Xiqiu²; (1.Department of Information Technology , Guangzhou Sontan Polytechnic College , Guangzhou 511300 , China ; 2.Changjiang Institute of Technology , Wuhan 430212 , China )

Keywords:: image corner matching ; machined parts ; surface defect detection ; threshold segmentation

摘要:: 当前在对机械加工零件表面缺陷检测时,存在表面缺陷检测精度差的问题,导致检测结果不理想。提出一种基于图像角点匹配的机械加工零件表面缺陷检测方法,利用曲率空间检测零件图像的角点,采用泰勒级数删除伪角点。对特征点邻域梯度方向进行角度限度和就近投影,同时借助双向匹配方法进行机械加工零件图像角点匹配。在上述操作的基础上,利用一维直方图的阈值分割对零件图像进行分割处理,最终实现机械加工零件表面缺陷检测。实验结果表明,该方法能够获取高精度的零件表面缺陷检测结果,且对加工零件缺陷厚度、孔洞缺陷及缺陷最大边界距离的测量均较为准确。

Abstract:: At present , when detecting surface defects of machined parts , there is a problem of poor surface defect detection accuracy , which leads to unsatisfactory detection results.A surface defect detection method for machined parts based on image corner matching is proposed.The corner points of the part image are detected by the curvature space , and the pseudo corner points are deleted by using Taylor series. The angle limit and the nearest projection are performed on the gradient direction of the feature point neighborhood , and the corner points of the machined parts image are matched by the two-way matching method.On the basis of the above operation , the part image is segmented by the threshold segmentation of the one-dimensional histogram , and finally the surface defectdetectio of the machined part is realized. The experimental results show that the proposed method can obtain high-precision parts surface defect detection results , and the measurement of defect thickness , hole defects and maximum boundary distance of the machined parts are relatively accurate.

[ 1 ] 李梓瑞,王慧琴,胡燕,等 . 基于深度学习和最大相关最小冗余的火焰图像检测方法[ J ] . 激光与光电子学进展,2020 , 57 ( 10 ): 160-170.

[ 2 ] 包能胜,方海涛 . 基于蚁群算法的辊式涂布涂层厚度图像检测[ J ] . 应用光学, 2020 , 41 ( 3 ): 516-522.

[ 3 ] 舒威,杨贤昭,杨艳华,等 . 基于无人机视觉的储罐表面缺陷检测方法 [ J ] . 高技术通讯, 2019 , 29 ( 8 ):799-807.

[ 4 ] 周友行,马逐曦,石弦韦,等 .HSI 颜色空间下的直线导轨表面缺陷检测方法 [ J ] . 中国机械工程, 2019 ,30( 18 ): 2179-2184.

[ 5 ] 骆荣坤,刘桂礼,孔全存,等 . 基于四邻域对角线上灰度约束的亚像素角点检测[ J ] . 计算机工程与设计, 2019 ,40 ( 1 ): 77-83.

[ 6 ] 董立红,彭业勋,符立梅 . 基于 Sobel 边缘检测的圆周Harris 角点检测算法[ J ] . 西安科技大学学报, 2019 , 39( 2 ): 374-380.

[ 7 ] 吴禄慎,万尧,陈华伟,等 . 基于 CSS 角点检测的快速匹配算法[ J ] . 电光与控制, 2019 , 26 ( 2 ): 28-31.

[ 8 ] 李云红,何亚瑞,章为川,等 . 利用点弦距离递归的图像角点检测算法[ J ] . 中国图象图形学报,2019 , 24 ( 7 ):1148-1159.

[ 9 ] 谢敏,周凯,黄永禄,等 . 一种基于短时奇异值分解的局部放电白噪声抑制方法[ J ] . 中国电机工程学报, 2019 ,39 ( 3 ): 915-922 , 970.

[ 10 ] 周凯,黄永禄,谢敏,等 . 短时奇异值分解用于局放信号混合噪声抑制[ J ] . 电工技术学报,2019 , 34 ( 11 ):2435-2443.

[ 11 ] 赵小强,张源峰 . 基于 Harris 角点与 SIFT 的快速图像匹配算法[ J ] . 兰州理工大学学报,2019 , 45 ( 1 ):101-106.

[ 12] 赵明富,陈兵,宋涛,等 . 融合特征光流与角点特征的图像特征匹配算法研究 [ J ] . 半导体光电,2019 , 40( 4 ): 539-545 , 549.

[ 13 ] 钟佩珂,江万寿,李金龙,等 . 基于相位一致性的红外多光谱影像波段间特征匹配[ J ] . 地理与地理信息科学,2020 , 36 ( 3 ): 42-48 , 62.

[ 14 ] 蔡欣展,刘艳艳 . 基于自适应筛选 Harris 角点检测的快速图像配准算法 [ J ] . 半导体光电,2020 , 41 ( 6 ):875-878.

[ 15 ] 向玉开,雷林建,张悦 . 基于计算机视觉的注塑零件表面缺陷在线检测研究[ J ] . 塑料工业,2019 , 47 (增刊1 ): 87-90 , 97.

备注/Memo:: 收稿日期: 2021-11-11
基金项目:中国高校产学研创新基金( 2019ITA02017 )
作者简介:肖红 ( 1977- ),男,湖北天门人,硕士,讲师,研究方向为图像处理、模式识别;陈希球 ( 1962- ),男,湖北崇阳人,硕士,教授,研究方向为信息技术应用。

更新日期/Last Update: 2022-10-04