Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets
This study introduces a novel approach to the critical task of submarine pipeline or cable (POC) detection by employing GoogleNet for the automatic recognition of side-scan sonar (SSS) images. The traditional interpretation methods, heavily reliant on human interpretation, are replaced with a more r...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/19/4873 |
| _version_ | 1797575215526445056 |
|---|---|
| author | Xing Du Yongfu Sun Yupeng Song Lifeng Dong Xiaolong Zhao |
| author_facet | Xing Du Yongfu Sun Yupeng Song Lifeng Dong Xiaolong Zhao |
| author_sort | Xing Du |
| collection | DOAJ |
| description | This study introduces a novel approach to the critical task of submarine pipeline or cable (POC) detection by employing GoogleNet for the automatic recognition of side-scan sonar (SSS) images. The traditional interpretation methods, heavily reliant on human interpretation, are replaced with a more reliable deep-learning-based methodology. We explored the enhancement of model accuracy via transfer learning and scrutinized the influence of three distinct pre-training datasets on the model’s performance. The results indicate that GoogleNet facilitated effective identification, with accuracy and precision rates exceeding 90%. Furthermore, pre-training with the ImageNet dataset increased prediction accuracy by about 10% compared to the model without pre-training. The model’s prediction ability was best promoted by pre-training datasets in the following order: Marine-PULSE ≥ ImageNet > SeabedObjects-KLSG. Our study shows that pre-training dataset categories, dataset volume, and data consistency with predicted data are crucial factors affecting pre-training outcomes. These findings set the stage for future research on automatic pipeline detection using deep learning techniques and emphasize the significance of suitable pre-training dataset selection for CNN models. |
| first_indexed | 2024-03-10T21:35:30Z |
| format | Article |
| id | doaj.art-3725710fdb044cae9b008fedffac2707 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T21:35:30Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-3725710fdb044cae9b008fedffac27072023-11-19T15:01:12ZengMDPI AGRemote Sensing2072-42922023-10-011519487310.3390/rs15194873Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training DatasetsXing Du0Yongfu Sun1Yupeng Song2Lifeng Dong3Xiaolong Zhao4First Institute of Oceanography, Ministry of Natural Resources of the People’s Republic of China, Qingdao 266061, ChinaNational Deep Sea Center, Qingdao 266237, ChinaFirst Institute of Oceanography, Ministry of Natural Resources of the People’s Republic of China, Qingdao 266061, ChinaFirst Institute of Oceanography, Ministry of Natural Resources of the People’s Republic of China, Qingdao 266061, ChinaFirst Institute of Oceanography, Ministry of Natural Resources of the People’s Republic of China, Qingdao 266061, ChinaThis study introduces a novel approach to the critical task of submarine pipeline or cable (POC) detection by employing GoogleNet for the automatic recognition of side-scan sonar (SSS) images. The traditional interpretation methods, heavily reliant on human interpretation, are replaced with a more reliable deep-learning-based methodology. We explored the enhancement of model accuracy via transfer learning and scrutinized the influence of three distinct pre-training datasets on the model’s performance. The results indicate that GoogleNet facilitated effective identification, with accuracy and precision rates exceeding 90%. Furthermore, pre-training with the ImageNet dataset increased prediction accuracy by about 10% compared to the model without pre-training. The model’s prediction ability was best promoted by pre-training datasets in the following order: Marine-PULSE ≥ ImageNet > SeabedObjects-KLSG. Our study shows that pre-training dataset categories, dataset volume, and data consistency with predicted data are crucial factors affecting pre-training outcomes. These findings set the stage for future research on automatic pipeline detection using deep learning techniques and emphasize the significance of suitable pre-training dataset selection for CNN models.https://www.mdpi.com/2072-4292/15/19/4873side-scan sonarconvolutional neural networkstransfer learninggeological surveyGoogleNetYellow River Estuary |
| spellingShingle | Xing Du Yongfu Sun Yupeng Song Lifeng Dong Xiaolong Zhao Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets Remote Sensing side-scan sonar convolutional neural networks transfer learning geological survey GoogleNet Yellow River Estuary |
| title | Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets |
| title_full | Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets |
| title_fullStr | Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets |
| title_full_unstemmed | Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets |
| title_short | Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets |
| title_sort | revealing the potential of deep learning for detecting submarine pipelines in side scan sonar images an investigation of pre training datasets |
| topic | side-scan sonar convolutional neural networks transfer learning geological survey GoogleNet Yellow River Estuary |
| url | https://www.mdpi.com/2072-4292/15/19/4873 |
| work_keys_str_mv | AT xingdu revealingthepotentialofdeeplearningfordetectingsubmarinepipelinesinsidescansonarimagesaninvestigationofpretrainingdatasets AT yongfusun revealingthepotentialofdeeplearningfordetectingsubmarinepipelinesinsidescansonarimagesaninvestigationofpretrainingdatasets AT yupengsong revealingthepotentialofdeeplearningfordetectingsubmarinepipelinesinsidescansonarimagesaninvestigationofpretrainingdatasets AT lifengdong revealingthepotentialofdeeplearningfordetectingsubmarinepipelinesinsidescansonarimagesaninvestigationofpretrainingdatasets AT xiaolongzhao revealingthepotentialofdeeplearningfordetectingsubmarinepipelinesinsidescansonarimagesaninvestigationofpretrainingdatasets |