EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines
Quantitative analysis of the magnetic leakage signal is crucial for evaluating the magnitude of pipeline damage after fault detection. To address these issues, it is necessary to establish a correlation between the leakage signal and the magnitude of the defect to make accurate predictions through r...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10820322/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832592921081151488 |
|---|---|
| author | Di Wu Yong Hong Jie Wang Shaojun Wu Zhihao Zhang Yizhang Liu |
| author_facet | Di Wu Yong Hong Jie Wang Shaojun Wu Zhihao Zhang Yizhang Liu |
| author_sort | Di Wu |
| collection | DOAJ |
| description | Quantitative analysis of the magnetic leakage signal is crucial for evaluating the magnitude of pipeline damage after fault detection. To address these issues, it is necessary to establish a correlation between the leakage signal and the magnitude of the defect to make accurate predictions through regression analysis. This research introduces EffiTriDimNet (ETDN), a multi-task convolutional neural network that combines one-dimensional pipeline defect leakage detection data into a unified feature map while simultaneously measuring the three-dimensional characteristics of the defects. This multi-task learning model utilizes a shared feature extraction layer and numerous branching networks to create individual predictions for each task: length, width, and depth. The network uses the EfficientNet structure for feature extraction, while the training and validation process utilizes the PyTorch Lightning framework. Experimental results demonstrate that when compared to other algorithms, ETDN outperforms them on the identical test set, exhibiting average absolute errors of 0.1920, 0.1809, and 0.2904 for length, width, and depth, respectively. |
| format | Article |
| id | doaj-art-46c0b7444c9d479d87ffae29823f1725 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-46c0b7444c9d479d87ffae29823f17252025-01-21T00:01:40ZengIEEEIEEE Access2169-35362025-01-01139089910110.1109/ACCESS.2025.352554010820322EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in PipelinesDi Wu0https://orcid.org/0009-0009-6566-7563Yong Hong1Jie Wang2Shaojun Wu3Zhihao Zhang4Yizhang Liu5School of Mechanical Engineering, Shanghai Dianji University, Shanghai, ChinaSchool of Mechanical Engineering, Shanghai Dianji University, Shanghai, ChinaWuhu Special Equipment Inspection Institute, Wuhu, Anhui, ChinaWuhu Special Equipment Inspection Institute, Wuhu, Anhui, ChinaSchool of Mechanical Engineering, Shanghai Dianji University, Shanghai, ChinaSchool of Mechanical Engineering, Shanghai Dianji University, Shanghai, ChinaQuantitative analysis of the magnetic leakage signal is crucial for evaluating the magnitude of pipeline damage after fault detection. To address these issues, it is necessary to establish a correlation between the leakage signal and the magnitude of the defect to make accurate predictions through regression analysis. This research introduces EffiTriDimNet (ETDN), a multi-task convolutional neural network that combines one-dimensional pipeline defect leakage detection data into a unified feature map while simultaneously measuring the three-dimensional characteristics of the defects. This multi-task learning model utilizes a shared feature extraction layer and numerous branching networks to create individual predictions for each task: length, width, and depth. The network uses the EfficientNet structure for feature extraction, while the training and validation process utilizes the PyTorch Lightning framework. Experimental results demonstrate that when compared to other algorithms, ETDN outperforms them on the identical test set, exhibiting average absolute errors of 0.1920, 0.1809, and 0.2904 for length, width, and depth, respectively.https://ieeexplore.ieee.org/document/10820322/Pipeline defect predictionnon-destructive testingEFFICIENTNET-b0feature extractionregression analysisPyTorch lightning framework |
| spellingShingle | Di Wu Yong Hong Jie Wang Shaojun Wu Zhihao Zhang Yizhang Liu EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines IEEE Access Pipeline defect prediction non-destructive testing EFFICIENTNET-b0 feature extraction regression analysis PyTorch lightning framework |
| title | EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines |
| title_full | EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines |
| title_fullStr | EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines |
| title_full_unstemmed | EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines |
| title_short | EfficientNet-b0-Based 3D Quantification Algorithm for Rectangular Defects in Pipelines |
| title_sort | efficientnet b0 based 3d quantification algorithm for rectangular defects in pipelines |
| topic | Pipeline defect prediction non-destructive testing EFFICIENTNET-b0 feature extraction regression analysis PyTorch lightning framework |
| url | https://ieeexplore.ieee.org/document/10820322/ |
| work_keys_str_mv | AT diwu efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines AT yonghong efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines AT jiewang efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines AT shaojunwu efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines AT zhihaozhang efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines AT yizhangliu efficientnetb0based3dquantificationalgorithmforrectangulardefectsinpipelines |