[1]丛　军,张辉宇,谢 建,等.基于对数极坐标变换的X射线焊缝缺陷图像识别算法[J].焊管,2022,45(2):39-43.[doi:10.19291/j.cnki.1001-3938.2022.02.006]
　CONG Jun,ZHANG Huiyu,XIE Jian,et al.Algorithm for Detecting X-ray Welding Defects Image based on LPT[J].,2022,45(2):39-43.[doi:10.19291/j.cnki.1001-3938.2022.02.006]
点击复制

基于对数极坐标变换的X射线焊缝缺陷图像识别算法()
分享到：

《焊管》[ISSN:1001-3938/CN:61-1160/TE]

卷:
45
期数:
2022年第2期
页码:
39-43
栏目:
应用与开发
出版日期:
2022-02-28

文章信息/Info

Title:
Algorithm for Detecting X-ray Welding Defects Image based on LPT
文章编号:
10.19291/j.cnki.1001-3938.2022.02.006
作者:
丛　军; 张辉宇; 谢 建; 陈健飞; 高海涛; 张 宏; 韩曙光; 王 征
中石化胜利海上石油工程技术检验有限公司，山东 东营 257000
Author(s):
CONG Jun;  ZHANG Huiyu;  XIE Jian;  CHEN Jianfei;  GAO Haitao; ZHANG Hong;  HAN Shuguang;  WANG Zheng
Shengli Offshore Petroleum Engineering Technology Inspection Co., Ltd., Sinopec, Dongying 257000, Shandong, China
关键词:
缺陷识别; X射线; 压缩感知; 对数极坐标变换
Keywords:
defect recognition;  X-ray;  compressed sensing;  log polar coordinate transformation
分类号:
TG441.7
DOI:
10.19291/j.cnki.1001-3938.2022.02.006
文献标志码:
A
摘要:
为了提高X射线焊缝图像缺陷分类识别的成功率，提出利用对数极坐标变换（LPT）的距离及角度不变特性解决实际缺陷识别中因缺陷图像位置、形状及方向变换导致的识别准确率下降问题。同时，为降低噪声的影响，提出基于压缩传感技术的缺陷识别，将待检测图像视为一组样本图像的线性组合，通过求取系数向量确定待检测图像的类型。由于缺陷的信息存于一组系数向量中，而非依赖于某几个特定的特征值，因此任一图像的变化对最终识别结果不会造成影响。实际试验表明，LPT结合压缩传感技术可以实现较高的X射线焊缝图像缺陷识别率，极大地降低了噪声误报率。
Abstract:
In order to improve the success rate of X-ray welding defect identification, the distance and angle invariance of log polar coordinate transformation (LPT) is proposed to solve the problem of accuracy degradation caused by the transformation of defect image position, shape and direction in practical defect recognition. At the same time, in order to reduce the influence of noise, a defect recognition based on compressed sensor technology is proposed. The image to be detected is regarded as a linear combination of a group of sample images, and the type of image to be detected is determined by calculating the coefficient vector. Because the information of defect image is reflected by a group of coefficient and not by some specific characteristics, any change in one image has no influence on the recognition result. Experiments show that LPT combined with compress sensor can achieve high defect recognition rate of X-ray weld image and greatly reduce the noise false alarm rate.

相似文献/References:

[1]张树鹏,赵文生.X射线实时成像系统在压力容器无损检测中的应用[J].焊管,2007,30(6):45.[doi:1001－3938（2007）06－0045－03]
　ZHANG Shu-peng,ZHAO Wen-sheng.Research and Application of the Real-time Imagery System in Nondestructive Detection for the Pressure Vessel[J].,2007,30(2):45.[doi:1001－3938（2007）06－0045－03]
[2]徐　轲,浦蕾晔,段红红,等.焊缝无损检测中未焊透与未熔合的识别[J].焊管,2011,34(6):51.[doi:1001-3938（2011）06-0051-03]
　XU Ke,PU Lei-ye,DUAN Hong-hong,et al.Identification of Incomplete Penetration and Lack of Fusion in Weld Nondestructive Testing [J].,2011,34(2):51.[doi:1001-3938（2011）06-0051-03]
[3]董 斌,杨剑峰.X射线探伤技术对埋弧焊管焊缝检测中气孔缺陷的判定方法[J].焊管,2013,36(12):53.[doi:１００１－３９３８（２０１3）12-0053-03]
　DONG Bin,YANG Jianfeng.Determination Method of X-ray Radiographic Inspection for Gas Pocket in SAW Pipe Weld Detection[J].,2013,36(2):53.[doi:１００１－３９３８（２０１3）12-0053-03]
[4]刘 欣.X射线探臂升降装置及射线检测室的设计改造[J].焊管,2014,37(1):50.[doi:１００１－３９３８（２０１4）01-0050-03]
　LIU Xin.Design Modification of X-ray Lifting Device and the Ray Testing Room[J].,2014,37(2):50.[doi:１００１－３９３８（２０１4）01-0050-03]
[5]王长安,余国民,魏雅博.平板DR数字成像技术在焊管行业中的应用[J].焊管,2014,37(12):52.[doi:１００１－３９３８（２０１4）12-0052-04]
　WANG Changan,YU Guomin,et al.Application of Flat Panel Digital Radiography Imaging Technology in Welded Pipe Industry[J].,2014,37(2):52.[doi:１００１－３９３８（２０１4）12-0052-04]
[6]马 钢,白 瑞.基于EMD技术的管道漏磁内检测信号研究*[J].焊管,2018,41(5):38.[doi:10.19291/j.cnki.1001-3938.2018.05.007]
　MA Gang,BAI Rui.Research on Internal Detection Signal of Pipeline Magnetic Flux Leakage Based on EMD Technology[J].,2018,41(2):38.[doi:10.19291/j.cnki.1001-3938.2018.05.007]

备注/Memo

备注/Memo:
收稿日期：2021-08-23基金项目： 陕西省重点研发计划项目“冷阴极X射线焊缝缺陷自动检测系统”（项目编号2020GY-179）。作者简介：丛 军（1971—），男，硕士，高级工程师，主要从事无损检测研究。
更新日期/Last Update: 2022-03-03