摘要
在电容成像(E lectrica l C apac itance T om ography,ECT)中,为充分利用多次量测信息以提高电容成像图像重建质量,提出一种基于K a lm an滤波的电容成像图像重建算法。该算法重点考虑了测量噪声的影响,利用对流型一系列多次测量中获得的新息不断进行最优加权以获得重建图像的最小方差估计。针对3种典型介电常数分布进行了仿真,结果表明K a lm an滤波应用于ECT图像重建的可行性和有效性。提出了提高该算法运算速度的方案,分析和仿真结果表明通过预先计算最优滤波增益,并寻找合适的迭代次数,算法可快速地获得满意的图像重建结果。
The quality of reconstructed images for electrical eapacitance tomography is improved by a reconstruction algorithm based on Kalman filtering. Compared with the existing ECT image reconstruction algorithms which use only one frame of measured data, the new algorithm uses a series of data frames to obtain the minimum-variance estimate of the reconstructed image. The measurement noise is also considered. Simulations with three typical permittivity distributions show the feasibility and efficiency of the Kalman filtering-based image reconstruction algorithm for ECT. Several methods are also given to reduce the computation load. Analysis and simulations show that pre-calculating the Kalman gain and using the optimal iteration step provide fast, satisfactory reconstructed images.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2005年第10期1332-1334,1351,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家自然科学基金资助项目(60204003)
关键词
电容成像
图像重建
KALMAN滤波
最小方差估计
electrical capacitance tomography
image reconstruction
Kalman filtering
minimum-variance estimation
