摘要
针对电阻抗成像空间分辨率低和对测量噪声敏感的问题,将传统Tikhonov正则化问题中目标函数的L2范数正则项修正为L1范数,将动态电阻抗图像中非均匀的电导率具有稀疏性作为先验信息添加到L1范数正则项中,由此提出一种电阻抗成像的稀疏重建算法。建立基于总变差法、正交匹配追踪法以及L1范数最小二乘法的电阻抗成像模型,并借助实验可知,新算法成像质量好,对测量噪声不敏感,且成像速度较快。
In this paper, an electrical impedance imaging algorithm based on sparse reconstruction was proposed in order to tackle the low spatial resolution and sensitive to measurement noise in traditional electrical impedance imaging. In this algorithm, L2-norm regularization in traditional Tikhonov regularization was corrected for Ll-norm, and L1-norm was then combined with the specialty which conductivity with sparse characteristics in the dynamic imaging was added to it as prior knowledge. An electrical impedance imaging model was constructed based on the total varia tion algorithm, the orthogonal matching pursuit algorithm and L1 norm least squares algorithm. Experimental results showed that the new algorithms not only have high-quality imaging and fast imaging, but also were insensitive to noise.
出处
《西安邮电学院学报》
2013年第2期92-96,110,共6页
Journal of Xi'an Institute of Posts and Telecommunications
基金
陕西省教育厅科学研究基金资助项目(12JK0735)
西安邮电大学青年教师科研基金资助项目(ZL2012-20)
关键词
电阻抗成像
稀疏重建
最优化
electrical impedance tomography （EIT）, sparse reconstruction, optimization
