基于小波神经网络的电容层析成像图像重建算法

电容层析成像(electrical capacitance tomography，ECT)技术是一个复杂的非线性问题，针对图像重建问题的病态性，提出了基于小波神经网络的图像重建算法。利用主成分分析法对输入变量(电容测量值)进行降维处理，利用小波神经网络建立电容测量值与成像区域介电常数分布的非线性映射。小波神经网络的神经元激励函数由伸缩和平移因子决定的小波基函数，采用BP算法对网络进行训练，并引入学习率与动量因子的自适应调整方法以加快网络训练的收敛速度。实验结果表明，与典型的反投影及Landweber迭代算法相比，该算法所构图像质量有明显改善。

Image Reconstruction Algorithm for Electrical Capacitance Tomography Based on Wavelet Neural Networks

Electrical capacitance tomography (ECT) technique is a complex nonlinear problem. To solve the ill-posed image reconstruction problem, a new image reconstruction algorithm based on wavelet neural networks (WNN) was proposed. The principal component analysis (PCA) method was used to reduce the dimension of the input vectors (capacitance measurements). Then, a nonlinear map between capacitance measurements and the permittivity distribution in image region was built. The transfer functions of the neurons in the wavelet neural networks were wavelet base functions which were determined by retract and translation factors. BP algorithm was used to train the WNN, and self-adaptive learning rate and momentum coefficient were also used to accelerate the learning speed. Experimental results show the image quality has been improved markedly, compared with the typical linear back projection (LBP) algorithm and Landweber iteration algorithm.