利用Kalman滤波算法进行电容层析成像图像重建,通过不断更新测量信息提高重建图像质量.在滤波初始,需要确定重建图像灰度及估计误差方差矩阵的初值.为研究不同初值组合对重建图像质量的影响,选择3流型进行了仿真实验,获得了最佳的初值组合,并在电容层析成像实验装置上进行了静态测试,实验结果表明,与Landweber算法相比,使用最佳初值组合的Kalman滤波算法可获得更为接近真实分布的重建图像.%Kalman filter was applied to improve the quality of reconstructed images for electrical capacitance tomography(ECT) by updating the measurement information.The initial values of reconstruction image gray and estimation of the error variance matrix needed to be determined during the process of Kalman filter.To study the influence of initial values on the quality of reconstructed image,three flow patterns were selected for simulation experiments and then the optimal initial values were obtained.At the same time,static tests were carried out using ECT device.Experimental results showed that the reconstructed images which are more close to the real distributions can be obtained based on Kalman filter algorithm using the optimal initial values compared with Landweber algorithm.
展开▼
