Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets
Description
The Marine_PULSE dataset is a dataset focused on the automatic recognition of side-scan sonar images of marine engineering structures. The attached data has already been divided into training and test sets. Data were obtained with various side-scan sonar instruments, including EdgeTech4200FS, Benthos SIS-1624, Edgetech4200MP, Klein-2000, and Klein-3000. The Marine-PULSE dataset comprises 323 pipeline or cable (POC), 134 underwater residual mound (URM), 88 seabed surface (SS), and 82 engineering platform (EP) images. PULSE signifies both the dataset's image types and the ocean's effective information detectable by side-scan sonar. All images were processed using PostSurvey, a free data processing program by KNUDSEN, with raw target object images captured without post-processing.
Notes
Other
This dataset was funded by the Foundation item:
The National Natural Science Foundation of China under contract NO. 42102326; the Basic Scientific Fund for National Public Research Institutes of China under contract NO. 2022Q05; and The Shandong Provincial Natural Science Foundation, China under contract NO. ZR2020QD073.
Files
Marine_PULSE.zip
Files
(64.0 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:eaae3db8445b757eab6edf79f4ba44d8
|
64.0 MB | Preview Download |
Additional details
References
- DU X, SUN Y, SONG Y, et. al. A Comparative Study of Different CNN Models and Transfer Learning Effect for Underwater Object Classification in Side-Scan Sonar Images[J]. Remote Sensing, 2023, 15(3): 593. DOI:10.3390/rs15030593.
- DU X, SUN Y, SONG Y, et. al. Revealing the Potential of Deep Learning for Detecting Submarine Pipelines in Side-Scan Sonar Images: An Investigation of Pre-Training Datasets[J]. Remote Sensing, 2023, 15(19): 4873. DOI:10.3390/rs15194873.