|本期目录/Table of Contents|

[1]马振起,刘 泽,曹景铭,等.基于降维处理的快速EMT图像重建算法[J].工业仪表与自动化装置,2024,(04):92-97.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2024.04.018]
　MA Zhenqi,LIU Ze,CAO Jingming,et al.Fast electromagnetic tomography image reconstruction algorithm based on dimensionality reduction[J].Industrial Instrumentation & Automation,2024,(04):92-97.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2024.04.018]
点击复制

基于降维处理的快速EMT图像重建算法(PDF)

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2024年04期
页码:
92-97
栏目:
出版日期:
2024-08-15

文章信息/Info

Title:
Fast electromagnetic tomography image reconstruction algorithm based on dimensionality reduction
文章编号:
1000-0682（2024）04-0092-06
作者:
马振起; 刘 泽; 曹景铭; 等
（北京交通大学 自动化与智能学院，北京 100044）
Author(s):
MA Zhenqi; LIU Ze; CAO Jingming;  et al
（School of Automation and Intelligence, Beijing Jiaotong University, Beijing 100044, China)
关键词:
电磁层析成像; 图像重建算法; 数据降维; 核主成分分析; 病态性
Keywords:
electromagnetic tomography;  image reconstruction algorithm;  data dimensionality reduction;  kernel principal component analysis;  ill-conditioning
分类号:
TP23；TP212
DOI:
DOI:10.19950/j.cnki.cn61-1121/th.2024.04.018
文献标志码:
A
摘要:
电磁层析成像技术(EMT)具有非侵入、响应速度快、成本低等优点，在工业过程监测和多相流测量等领域有广泛的应用前景。该文针对电磁层析成像逆问题的病态性，提出了1种非迭代的、基于灵敏度矩阵降维的EMT图像重建算法，应用核主成分分析(KPCA)方法对灵敏度矩阵进行降维，有效降低了算法计算复杂度，同时降低了灵敏度矩阵的病态程度。为验证该算法的有效性，将该算法应用于平面EMT金属探伤，并将其与传统的线性反投影算法和Landweber迭代法进行比较。仿真和实验结果表明，该算法的成像质量远高于线性反投影算法，与Landweber迭代法相近，且该算法的计算耗时仅为Landweber迭代法的20%左右。
Abstract:
Electromagnetic tomography (EMT), which has the advantages of noninvasiveness, fast response and low cost, has the potential for extensive use in industrial process monitoring, multiphase flow measurement and other fields. To cope with the ill-conditioning of the inverse problem of EMT, a non-iterative EMT image reconstruction algorithm based on sensitivity matrix dimensionality reduction is proposed. Kernel principal component analysis (KPCA) is used to reduce the dimensionality of the sensitivity matrix, which effectively reduces the computational complexity of the algorithm and reduces the ill-conditioning of the sensitivity matrix. To verify its effectiveness, the proposed algorithm is applied to planar EMT metal flaw detection, and is compared with the traditional linear back projection (LBP) algorithm and the Landweber iteration method. Simulation and experimental results show that the imaging quality of the proposed algorithm is much higher than that of the LBP algorithm and is similar to that of the Landweber iteration method, and the calculation time of the proposed algorithm is only about 20% of that of the Landweber iteration method.

参考文献/References:

[1] XIANG J X, DONG Y G, YANG Y J. Multi-frequency electromagnetic tomography for acute stroke detection using frequency-constrained sparse bayesian learning[J]. IEEE Transactions on Medical Imaging,2020,39(12): 4102-4112.

[2] 胡亚峰.面向气液固三相流检测的快速电/磁层析成像系统设设计[D].天津:天津大学,2022.
[3] Semenov S. Electromagnetic tomographic cerebral angiography[J]. Scientific reports, 2024,14(1):1792-1792.
[4] 王成飞.高速列车空心轴电磁层析成像缺陷检测技术研究[D].北京:北京交通大学,2020.
[5] 霍继伟,刘泽,王亚东,等.平面电磁层析成像钢轨探伤[J].中国电机工程学报,2021,41(15):5351-5361.
[6] 刘泽,肖君,刘向龙,等.一种电磁层析图像快速重建算法[J].北京航空航天大学学报,2018,44(08):1569-1576.
[7] 王亚东,刘泽,霍继伟,等.基于SOA优化的电磁层析成像图像重建研究[J].计算机测量与控制,2020,28(12): 207-212.
[8] GUO R, HUANG T Y, LI M K, et al. Physics-Embedded Machine Learning for Electromagnetic Data Imaging: Examining three types of data-driven imaging methods[J]. IEEE Signal Processing Magazine,2023,40(2):18-31.
[9] LIU X, LIU Z, LI Y, et al. Research on direct 3D electromagnetic tomography technique[J].IEEE Sensors Journal, 2020,20(9):4758-4767.
[10] 霍继伟.高速电磁层析成像关键技术研究及应用[D].北京:北京交通大学,2021.
[11] 冯旭刚,文作银,章家岩,等.基于KPCA-AGRU神经网络的火电机组NO_x排放预测[J].重庆工商大学学报(自然科学版),2023,40(06):18-24.
[12] 王艳,颜华,王伊凡,等.基于降维处理的Tikhonov正则化ECT图像重建[J].中国电机工程学报,2022,42(24): 9118-9126.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期：2024-03-04基金项目：国家重点研发计划资助项目（2020YFC2200704）第一作者：马振起（1999—），男，山东青岛人，硕士研究生在读，研究方向为智能信息处理、电磁层析成像技术。E-mail: 21120232@bjtu.edu.cn通信作者：刘泽（1973—），男，内蒙古呼和浩特人，博士，教授，主要研究领域为电磁层析成像、精密测量、智能仪器、交通信息及控制。E-mail: zliu@bjtu.edu.cn
更新日期/Last Update: 1900-01-01