Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network
Energy spectrum computed tomography can provide quantitative information of scanned objects and realize material decomposition. At present, the material decomposition method based on neural networks overcomes the limited decomposition effect of traditional iterative algorithms. However, the performa...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Editorial Office of Computerized Tomography Theory and Application
2025-01-01
|
| Series: | CT Lilun yu yingyong yanjiu |
| Subjects: | |
| Online Access: | https://www.cttacn.org.cn/cn/article/doi/10.15953/j.ctta.2024.131 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832592382336434176 |
|---|---|
| author | Ran ZHANG Huihua KONG Jiaxin LI Yijiao SONG |
| author_facet | Ran ZHANG Huihua KONG Jiaxin LI Yijiao SONG |
| author_sort | Ran ZHANG |
| collection | DOAJ |
| description | Energy spectrum computed tomography can provide quantitative information of scanned objects and realize material decomposition. At present, the material decomposition method based on neural networks overcomes the limited decomposition effect of traditional iterative algorithms. However, the performance of traditional neural networks in feature detail recovery is still not satisfactory. To improve the material decomposition accuracy, a material decomposition method based on a Resnet and Squeeze excitation network (RS-Net) is proposed. The proposed method uses the structure of the U-Net network and Resnet-152 as the backbone network to extract multi-scale features. Parallel asymmetric convolution is used to complete the large kernel convolution, which reduces the number of parameters and computation of the network. The HD-SE attention mechanism is introduced in the decoder part to help the network recover the image features. Hybrid loss supervised network learning is used to improve the decomposition accuracy of the network. The feasibility of this method is verified on simulated rock and artificial sandstone datasets. The simulation and experimental results show that RS-Net combined with mixing loss can retain more internal details of the image, the decomposed image edge is clearer, and the image quality is higher. |
| format | Article |
| id | doaj-art-d303430c126c46ad8c7198cab9cb8e79 |
| institution | Kabale University |
| issn | 1004-4140 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Editorial Office of Computerized Tomography Theory and Application |
| record_format | Article |
| series | CT Lilun yu yingyong yanjiu |
| spelling | doaj-art-d303430c126c46ad8c7198cab9cb8e792025-01-21T09:14:43ZengEditorial Office of Computerized Tomography Theory and ApplicationCT Lilun yu yingyong yanjiu1004-41402025-01-0134111712810.15953/j.ctta.2024.1312024.131Dense Sandstone Material Decomposition Based on Improved Convolutional Neural NetworkRan ZHANG0Huihua KONG1Jiaxin LI2Yijiao SONG3School of Mathematics, North University of China, Taiyuan 030051, ChinaSchool of Mathematics, North University of China, Taiyuan 030051, ChinaSchool of Mathematics, North University of China, Taiyuan 030051, ChinaSchool of Mathematics, North University of China, Taiyuan 030051, ChinaEnergy spectrum computed tomography can provide quantitative information of scanned objects and realize material decomposition. At present, the material decomposition method based on neural networks overcomes the limited decomposition effect of traditional iterative algorithms. However, the performance of traditional neural networks in feature detail recovery is still not satisfactory. To improve the material decomposition accuracy, a material decomposition method based on a Resnet and Squeeze excitation network (RS-Net) is proposed. The proposed method uses the structure of the U-Net network and Resnet-152 as the backbone network to extract multi-scale features. Parallel asymmetric convolution is used to complete the large kernel convolution, which reduces the number of parameters and computation of the network. The HD-SE attention mechanism is introduced in the decoder part to help the network recover the image features. Hybrid loss supervised network learning is used to improve the decomposition accuracy of the network. The feasibility of this method is verified on simulated rock and artificial sandstone datasets. The simulation and experimental results show that RS-Net combined with mixing loss can retain more internal details of the image, the decomposed image edge is clearer, and the image quality is higher.https://www.cttacn.org.cn/cn/article/doi/10.15953/j.ctta.2024.131spectrum ctconvolutional neural networkmaterial decompositionattention mechanisms |
| spellingShingle | Ran ZHANG Huihua KONG Jiaxin LI Yijiao SONG Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network CT Lilun yu yingyong yanjiu spectrum ct convolutional neural network material decomposition attention mechanisms |
| title | Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network |
| title_full | Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network |
| title_fullStr | Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network |
| title_full_unstemmed | Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network |
| title_short | Dense Sandstone Material Decomposition Based on Improved Convolutional Neural Network |
| title_sort | dense sandstone material decomposition based on improved convolutional neural network |
| topic | spectrum ct convolutional neural network material decomposition attention mechanisms |
| url | https://www.cttacn.org.cn/cn/article/doi/10.15953/j.ctta.2024.131 |
| work_keys_str_mv | AT ranzhang densesandstonematerialdecompositionbasedonimprovedconvolutionalneuralnetwork AT huihuakong densesandstonematerialdecompositionbasedonimprovedconvolutionalneuralnetwork AT jiaxinli densesandstonematerialdecompositionbasedonimprovedconvolutionalneuralnetwork AT yijiaosong densesandstonematerialdecompositionbasedonimprovedconvolutionalneuralnetwork |