@inproceedings{6c0d672fb37a4810a3141c117b8c193e,
title = "An improved segmentation method for porous transducer CT images",
abstract = "The paper presents an improved image segmentation method with a straightforward workflow for porous transducer CT images, which can be used to establish porous transducer three-dimensional model and further study its characteristics. Data distribution of CT images is firstly analyzed and Gaussian filtering is conducted to reduce divergence of CT images. An improved fully convolutional neural network model based on U-Net, for which multi-channel images are set as network input, is trained using training set. The proposed method improves pore connectivity of the segmentation results. Improvement of porosity and permeability relative errors as well as MIOU on test set shows that the proposed method is an effective and generic two-phase segmentation method for porous transducer CT images without need of adjusting any parameters.",
keywords = "Fully convolutional neural network, Image segmentation, Permeability, Pore connectivity, Porous transducer",
author = "Meiling Wang and Ruoyu Guo and Ke Ning and Li Ming",
note = "Publisher Copyright: {\textcopyright} COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.; 11th International Conference on Digital Image Processing, ICDIP 2019 ; Conference date: 10-05-2019 Through 13-05-2019",
year = "2019",
doi = "10.1117/12.2539604",
language = "English",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "SPIE",
editor = "Jenq-Neng Hwang and Xudong Jiang",
booktitle = "Eleventh International Conference on Digital Image Processing, ICDIP 2019",
address = "United States",
}