摘要： 对采用RBF神经网络的8电极电容层析成像系统的图像重建的方法进行了探讨。该神经网络采用遗传算法结合传统的最近邻聚类方法进行学习。仿真实验结果表明，该方法的成像精度及成像实时性较好。

关键词: 电容层析成像, RBF, 遗传算法, 神经网络, 图像重建

Abstract: An image reconstruction method that is based on RBF neural network for 8-electrode electrical capacitance tomography system is discussed. A genetic algorithm combining with nearest neighbor clustering algorithm is proposed to train the RBF network. The simulation results show the method is effective and accurate for image reconstruction.

中图分类号: 

• TP391.41