基于电磁层析成像的金属缺陷稀疏成像方法

发布时间：2018-05-29 09:42

本文选题：金属缺陷 + 电磁层析成像　；参考：《仪器仪表学报》2017年09期

【摘要】：采用电磁层析成像(electromagnetic tomography,EMT)技术实现对金属缺陷的可视化,克服了传统的检测技术无法对缺陷进行可视化的不足。首先设计了一种新型的平面EMT传感器,其次根据缺陷分布的稀疏性,提出了l1正则化稀疏成像算法。该算法能够有效避免传统的l2正则化算法带来的过度光滑的问题,成像更加精确。最后为证明该算法相对于l2正则化算法的优越性,进行了仿真和实验。仿真和实验结果均表明l1正则化稀疏成像算法能够有效提高缺陷图像的重建质量和精度。
[Abstract]:The electromagnetic tomography (EMT) technique is used to visualize metal defects, which overcomes the deficiency that the traditional detection technology can not visualize the defects. Firstly, a new planar EMT sensor is designed, and then, according to the sparsity of defect distribution, a L 1 regularized sparse imaging algorithm is proposed. The algorithm can effectively avoid the problem of excessive smoothness caused by the traditional l2 regularization algorithm and make the imaging more accurate. Finally, in order to prove the superiority of the algorithm compared with the L 2 regularization algorithm, the simulation and experiment are carried out. The simulation and experimental results show that the l _ 1 regularized sparse imaging algorithm can effectively improve the reconstruction quality and accuracy of defective images.
【作者单位】：天津工业大学电子与信息工程学院;天津市光电检测技术与系统重点实验室;天津大学电气与自动化工程学院;
【基金】：国家自然科学基金(61373104,61402330,61405143,61601324) 天津市应用基础与前沿技术研究计划(15JCQNJC01500)项目资助
【分类号】：TP212;TP391.41

【相似文献】