基于预优化变换的电容层析成像图像重建算法

Image reconstruction algorithm for electrical capacitance tomography based on pre-optimization transition

本文提出了一种新型快速电容层析成像的图像重建算法。该方法首先将图像重建的病态问题转化为泛函极小化问题,然后采用Fletcher-Reeves算法(FR)有效地求解该泛函,并在求解的过程中引入物理意义上的约束使获得的解更符合实际情况。数值实验表明该算法的图像重建效果优于线性反投影法、Tikhonov正则法和Landweber迭代法,而且该算法具有较强的抗噪声能力。

A novel fast algorithm for electrical capacitance tomography is presented. Firstly, the method changes the ill-posed equation of image reconstruction into function minimization problem, and then solves the function effectively using Fletcher-Reeves algorithm （FR）. During the process of solving the problem, some physical constraints are considered; as a result the solution is more consistent with the real circumstance. Numerical experiments indicate that the quality of the reconstructed image using proposed algorithm is better than the ones using image reconstruction algorithms such as LBP, Tikhonov and Landweber algorithm; moreover, the algorithm also has excellent noise rejection capability.