Fracture identification in well logging images: two-stage adaptive network
Automatic fracture identification and segmentation in well logging images is increasingly critical but arduous because of extensive exploration for oil and gas. Domain adaptation for semantic segmentation is an appealing alternative to the lack of semantic fracture annotations. Nevertheless, previou...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/163781 |
| _version_ | 1824454237395877888 |
|---|---|
| author | Zhang, Wei Li, Zhipeng Wu, Tong Yao, Zhenqiu Qiu, Ao Li, Yanjun Shi, Yibing |
| author2 | School of Electrical and Electronic Engineering |
| author_facet | School of Electrical and Electronic Engineering Zhang, Wei Li, Zhipeng Wu, Tong Yao, Zhenqiu Qiu, Ao Li, Yanjun Shi, Yibing |
| author_sort | Zhang, Wei |
| collection | NTU |
| description | Automatic fracture identification and segmentation in well logging images is increasingly critical but arduous because of extensive exploration for oil and gas. Domain adaptation for semantic segmentation is an appealing alternative to the lack of semantic fracture annotations. Nevertheless, previous domain adaptation strategies merely focus on a single level (e.g., input, intermediate features, or output), which could result in a large generalization error on new well logging images. In this article, a two-stage network architecture is proposed. In stage 1, a pattern transfer network (PTN) model is utilized to carry out a transformation from one domain to the other domain, making images visually similar between two datasets. Sequentially, in stage 2, an adversarial learning model with a generator and discriminator is introduced to make two domain images generated via PTN producing similar semantic segmentation. Furthermore, an attention module was embedded in the generator to filter irrelevant noise in semantic segmentation results. Real logging images were collected as the tested datasets by an ultrasonic imaging logging instrument. Compared to earlier conventional algorithms and domain adaptation methods, the proposed method is shown to obtain better visual quality and accurate segmentation outcomes. |
| first_indexed | 2025-02-19T03:19:07Z |
| format | Journal Article |
| id | ntu-10356/163781 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-02-19T03:19:07Z |
| publishDate | 2022 |
| record_format | dspace |
| spelling | ntu-10356/1637812022-12-16T06:55:45Z Fracture identification in well logging images: two-stage adaptive network Zhang, Wei Li, Zhipeng Wu, Tong Yao, Zhenqiu Qiu, Ao Li, Yanjun Shi, Yibing School of Electrical and Electronic Engineering Engineering::Electrical and electronic engineering Domain Adaptation Fracture Identification Automatic fracture identification and segmentation in well logging images is increasingly critical but arduous because of extensive exploration for oil and gas. Domain adaptation for semantic segmentation is an appealing alternative to the lack of semantic fracture annotations. Nevertheless, previous domain adaptation strategies merely focus on a single level (e.g., input, intermediate features, or output), which could result in a large generalization error on new well logging images. In this article, a two-stage network architecture is proposed. In stage 1, a pattern transfer network (PTN) model is utilized to carry out a transformation from one domain to the other domain, making images visually similar between two datasets. Sequentially, in stage 2, an adversarial learning model with a generator and discriminator is introduced to make two domain images generated via PTN producing similar semantic segmentation. Furthermore, an attention module was embedded in the generator to filter irrelevant noise in semantic segmentation results. Real logging images were collected as the tested datasets by an ultrasonic imaging logging instrument. Compared to earlier conventional algorithms and domain adaptation methods, the proposed method is shown to obtain better visual quality and accurate segmentation outcomes. This work was supported by the Science and Technology Project of China National Offshore Oil Corporation (CNOOC—Development and Industrial Application of Ultrahigh Temperature and High-Pressure Wireline Logging System) under Grant CNOOC-KJ ZDHXJSGG YF 2019-02. 2022-12-16T06:55:44Z 2022-12-16T06:55:44Z 2021 Journal Article Zhang, W., Li, Z., Wu, T., Yao, Z., Qiu, A., Li, Y. & Shi, Y. (2021). Fracture identification in well logging images: two-stage adaptive network. IEEE Transactions On Instrumentation and Measurement, 71, 5003112-. https://dx.doi.org/10.1109/TIM.2021.3130671 0018-9456 https://hdl.handle.net/10356/163781 10.1109/TIM.2021.3130671 2-s2.0-85120546273 71 5003112 en IEEE Transactions on Instrumentation and Measurement © 2021 IEEE. All rights reserved. |
| spellingShingle | Engineering::Electrical and electronic engineering Domain Adaptation Fracture Identification Zhang, Wei Li, Zhipeng Wu, Tong Yao, Zhenqiu Qiu, Ao Li, Yanjun Shi, Yibing Fracture identification in well logging images: two-stage adaptive network |
| title | Fracture identification in well logging images: two-stage adaptive network |
| title_full | Fracture identification in well logging images: two-stage adaptive network |
| title_fullStr | Fracture identification in well logging images: two-stage adaptive network |
| title_full_unstemmed | Fracture identification in well logging images: two-stage adaptive network |
| title_short | Fracture identification in well logging images: two-stage adaptive network |
| title_sort | fracture identification in well logging images two stage adaptive network |
| topic | Engineering::Electrical and electronic engineering Domain Adaptation Fracture Identification |
| url | https://hdl.handle.net/10356/163781 |
| work_keys_str_mv | AT zhangwei fractureidentificationinwellloggingimagestwostageadaptivenetwork AT lizhipeng fractureidentificationinwellloggingimagestwostageadaptivenetwork AT wutong fractureidentificationinwellloggingimagestwostageadaptivenetwork AT yaozhenqiu fractureidentificationinwellloggingimagestwostageadaptivenetwork AT qiuao fractureidentificationinwellloggingimagestwostageadaptivenetwork AT liyanjun fractureidentificationinwellloggingimagestwostageadaptivenetwork AT shiyibing fractureidentificationinwellloggingimagestwostageadaptivenetwork |