摘要
由于传统的电阻抗断层成像算法成像速度慢、且在某些情况下得到的图像无法反映敏感场的真实阻抗分布,本文提出了两种基于不同神经网络(BP神经网络和RBF神经网络)的电阻抗图像重建模型及其成像算法。结果表明,基于神经网络模型的成像算法从成像质量和速度上均优于传统的成像算法。
Duo to slow imaging speed and images got being unable to represent real impedance distribution of sensitive field in some case by traditional electrical impedance tomography algorithms, two models base on different neural networks (BP neural network and RBF neural network) and their image reconstruction algorithms for electrical impedance tomography are presented in this paper. Result illustrate that image reconstruction algorithms based on neural network model are superior in speed and quality to traditional algorithms.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2004年第1期94-97,共4页
Pattern Recognition and Artificial Intelligence
