摘要
为了提高电阻层析成像重建图像分辨率,提出一种改进牛顿-拉夫逊绝对图像重建算法.在算法迭代过程中,采取灵敏度矩阵自动更新的策略,同时结合图像重建结果自动选取阈值,并根据阈值修正电阻率.仿真实验结果表明:相比单纯采用灵敏度矩阵自动更新策略的改进牛顿-拉夫逊算法,在无噪声干扰下,新算法图像相对误差平均降低了35.620%,在1%幅度噪声干扰下,新算法图像相对误差平均降低了31.962%,在实现绝对图像重建的同时,有效提高了图像重建质量.
A modified Newton-Raphson absolute image reconstruction algorithm was proposed to improve the space resolution of the reconstructed image in electrical resistance tomography（ERT）. In the iteration process,the sensitiv-ity matrix was updated automatically;meanwhile,the resistivity was corrected according to a threshold determined by referring to the reconstruction result. Simulation experimentalresults demonstrate that compared to the modified Newton-Raphson method with sensitivity matrix updating,the proposed method reduces the relative image error by 35.620% in the noise-free case. With the existence of 1% noise,the proposed method reduces the relative image error by 31.962%. The proposed method enhances the imaging quality effectively while realizing absolute image reconstruc-tion.
出处
《天津大学学报（自然科学与工程技术版）》
EI
CAS
CSCD
北大核心
2015年第8期734-741,共8页
Journal of Tianjin University：Science and Technology
基金
国家自然科学基金资助项目(61201350
61302122)
徐州工程学院校科研资助项目(XKY2013318)
关键词
电阻层析成像
绝对图像重建
灵敏度矩阵
阈值
图像相对误差
electrical resistance tomography
absolute image reconstruction
sensitivity matrix
threshold
relative image error
