A convolutional image reconstruction network with spatial pyramid pooling for EIT

Feng Li^*, Yongwei Li, Zhimin Qiao, Yuqing Niu

Taiyuan Institute of Technology, Intelligent Detection and Control Engineering Research Center, Taiyuan 030008, China

^*Email: fengli@tju.edu.cn

ABSTRACT

Electrical impedance tomography (EIT) technology provides an effective approach for visualizing the distribution of multiple medium in multiphase flow. To address the complex nonlinear reconstruction problem of EIT, combined with the deep learning data-driven modeling method, a convolutional reconstruction network with spatial pyramid pooling is proposed. The nonlinear relationship between the measurement information obtained from the electrical impedance tomography system and the pixel information of the measured space is implemented by a fully connected block. The convolution block and the spatial pyramid pooling block increase the dimension and diversity of features, and the ill-posedness of image reconstruction is improved. The simulation and experimental results show that the proposed method has a significant improvement in reconstruction accuracy, compared with the reconstruction methods of Tikhonov regularization (TR) , tobal variation (TV) and V-Net.

Keywords: Multiphase flow, electrical impedance tomography, image reconstruction, convolutional network, spatial pyramid pooling

Industrial Application: Multiphase flow industry