基于WNN的电阻层析成像图像重建算法的研究

本文对于电阻层析成像系统的图像重建提出了一种基于小波神经网络的图像重建算法,介绍了电阻层析成像技术的原理、数学模型及问题描述,给出了小波神经网络的结构、特点及图像重建的过程.利用MATLAB编程进行了图像重建的仿真实验.仿真结果表明,与线性反投影算法相比,小波神经网络算法提高了重建图像的质量.用横截面图像误差作为重建图像质量的评判标准,线性反投影算法的误差超过了10%,而小波神经网络的算法低于5%.

Research of the image reconstruction algorithm for electrical resistance tomography (ERT) based on wavelet neural network

This paper presents a new image reconstruct algorithm for electrical resistance tomography based on wavelet neural network,and introduces ERT Technology's principle,mathematic model and wavelet neural network's structure,character and the process of image reconstruction based on this algorithm.The computer program is proposed using MATLAB.Some simulation results of image reconstruction are given.The result shows that,comparing with linear back projection,the qualities of the image reconstructed by the proposed algorithm are improved.When using cross section image error as standard to appraise the qualities of the image reconstruct,the LBP's error is above 10 percent,while WNN's error is under 5 percent.