电阻层析成像技术正则化图像重建算法研究  

Image Reconstruction Based on Regularization Algorithm for Electrical Resistance Tomography

作　　者：丁永维[1] 董峰[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津300072

出　　处：《控制工程》2007年第S3期136-138,141,共4页Control Engineering of China

基　　金：国家自然科学基金重点资助项目(60532020);863计划资助项目(2006AA04Z167)

摘　　要：针对目前电阻层析成像(ERT)技术中图像重建算法存在成像精度低和成像速度慢的问题,提出了一种正则化的电阻层析成像图像重建算法。该方法利用灵敏度原理计算ERT系统正问题的观测算子矩阵,利用最小二乘最小范数解的办法来求解电阻率分布矩阵,通过正则化的方法对电阻率分布矩阵进行修正,以提高图像重建的质量。通过与线性反投影算法和灵敏度系数算法的成像效果对比,该方法能够明显改善图像的质量。正则化图像重建算法成像速度很快,实时性很好。In order to solve the problem of low reconstruction quality and low reconstruction speed that exist in image reconstruction for electri- cal resistance tomography technology,a regularization image reconstruction algorithm for electrical resistance tomography is proposed.This method calculates the observation operator matrix for the forward problem of ERT system by sensitivity coefficient theory,and the resistivity dis- tribution matrix by the minimal norm and the least square solutions.The energy constraint condition and the least square constraint condition to revise resistivity distribution matrix to improve the quality of the images are introduced.Through the comparison of the images reconstructed by linear back projection algorithm,sensitivity coefficient algorithm and the regularization optimization algorithm proposed,it is apparent that im- ages can be reconstructed dearly and quickly by proposed regularization algorithm.

关 键 词：电阻层析成像 图像重建 灵敏度系数 正则化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]