Shale-pore Semantic Segmentation Network Based on Pseudo-labeling
Shale pore structures contribute significantly to shale gas reservoirs, with their shape, size, connectivity, and development directly affecting storage. To achieve intelligent recognition and classification of shale pores, this study proposes a shale-pore semantic segmentation network based on a ps...
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.2024.039 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832592364651151360 |
|---|---|
| author | Chenzhang WANG Yanfei WANG Zhijing BAI |
| author_facet | Chenzhang WANG Yanfei WANG Zhijing BAI |
| author_sort | Chenzhang WANG |
| collection | DOAJ |
| description | Shale pore structures contribute significantly to shale gas reservoirs, with their shape, size, connectivity, and development directly affecting storage. To achieve intelligent recognition and classification of shale pores, this study proposes a shale-pore semantic segmentation network based on a pseudo-label method. A total of 251 scanning electron microscopy images of the Longmaxi Formation shale reservoir in Chongqing are used, and the Pyramid Scene Parsing Network is utilized for training. Additionally, pseudo-label generation is employed, which involves annotating only a few images and using the model’s segmentation results on unlabeled images for iterative training. Subsequently, ensemble learning is conducted to improve the model’s accuracy and generalizability. The iteratively trained model has a mean intersection-over-union (MIoU) score exceeding 0.70. Comparative experiments show that employing pseudo-labeling and ensemble learning increases the model’s MIoU by approximately 0.07. Furthermore, pseudo-labeling enhances the generalizability of neural networks while addressing the high time cost of manual annotation required in previous deep-learning segmentation methods, whereas ensemble learning stably increases the model’s accuracy. |
| format | Article |
| id | doaj-art-8c9ed94165b445c593210d6829969854 |
| 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-8c9ed94165b445c593210d68299698542025-01-21T09:14:43ZengEditorial Office of Computerized Tomography Theory and ApplicationCT Lilun yu yingyong yanjiu1004-41402025-01-01341899810.15953/j.ctta.2024.0392024.039Shale-pore Semantic Segmentation Network Based on Pseudo-labelingChenzhang WANG0Yanfei WANG1Zhijing BAI2State Key Laboratory of Deep Petroleum Intelligent Exploration and Development, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, ChinaState Key Laboratory of Deep Petroleum Intelligent Exploration and Development, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, ChinaState Key Laboratory of Deep Petroleum Intelligent Exploration and Development, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, ChinaShale pore structures contribute significantly to shale gas reservoirs, with their shape, size, connectivity, and development directly affecting storage. To achieve intelligent recognition and classification of shale pores, this study proposes a shale-pore semantic segmentation network based on a pseudo-label method. A total of 251 scanning electron microscopy images of the Longmaxi Formation shale reservoir in Chongqing are used, and the Pyramid Scene Parsing Network is utilized for training. Additionally, pseudo-label generation is employed, which involves annotating only a few images and using the model’s segmentation results on unlabeled images for iterative training. Subsequently, ensemble learning is conducted to improve the model’s accuracy and generalizability. The iteratively trained model has a mean intersection-over-union (MIoU) score exceeding 0.70. Comparative experiments show that employing pseudo-labeling and ensemble learning increases the model’s MIoU by approximately 0.07. Furthermore, pseudo-labeling enhances the generalizability of neural networks while addressing the high time cost of manual annotation required in previous deep-learning segmentation methods, whereas ensemble learning stably increases the model’s accuracy.https://www.cttacn.org.cn/cn/article/doi/10.15953/j.ctta.2024.039deep learningshale poressemantic segmentationpseudo-labelensemble learning |
| spellingShingle | Chenzhang WANG Yanfei WANG Zhijing BAI Shale-pore Semantic Segmentation Network Based on Pseudo-labeling CT Lilun yu yingyong yanjiu deep learning shale pores semantic segmentation pseudo-label ensemble learning |
| title | Shale-pore Semantic Segmentation Network Based on Pseudo-labeling |
| title_full | Shale-pore Semantic Segmentation Network Based on Pseudo-labeling |
| title_fullStr | Shale-pore Semantic Segmentation Network Based on Pseudo-labeling |
| title_full_unstemmed | Shale-pore Semantic Segmentation Network Based on Pseudo-labeling |
| title_short | Shale-pore Semantic Segmentation Network Based on Pseudo-labeling |
| title_sort | shale pore semantic segmentation network based on pseudo labeling |
| topic | deep learning shale pores semantic segmentation pseudo-label ensemble learning |
| url | https://www.cttacn.org.cn/cn/article/doi/10.15953/j.ctta.2024.039 |
| work_keys_str_mv | AT chenzhangwang shaleporesemanticsegmentationnetworkbasedonpseudolabeling AT yanfeiwang shaleporesemanticsegmentationnetworkbasedonpseudolabeling AT zhijingbai shaleporesemanticsegmentationnetworkbasedonpseudolabeling |