Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning
Numerous sensors can obtain images or point cloud data on land, however, the rapid attenuation of electromagnetic signals and the lack of light in water have been observed to restrict sensing functions. This study expands the utilization of two- and three-dimensional detection technologies in underw...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/3/884 |
| _version_ | 1827597097936879616 |
|---|---|
| author | Chia-Ming Tsai Yi-Horng Lai Yung-Da Sun Yu-Jen Chung Jau-Woei Perng |
| author_facet | Chia-Ming Tsai Yi-Horng Lai Yung-Da Sun Yu-Jen Chung Jau-Woei Perng |
| author_sort | Chia-Ming Tsai |
| collection | DOAJ |
| description | Numerous sensors can obtain images or point cloud data on land, however, the rapid attenuation of electromagnetic signals and the lack of light in water have been observed to restrict sensing functions. This study expands the utilization of two- and three-dimensional detection technologies in underwater applications to detect abandoned tires. A three-dimensional acoustic sensor, the BV5000, is used in this study to collect underwater point cloud data. Some pre-processing steps are proposed to remove noise and the seabed from raw data. Point clouds are then processed to obtain two data types: a 2D image and a 3D point cloud. Deep learning methods with different dimensions are used to train the models. In the two-dimensional method, the point cloud is transferred into a bird’s eye view image. The Faster R-CNN and YOLOv3 network architectures are used to detect tires. Meanwhile, in the three-dimensional method, the point cloud associated with a tire is cut out from the raw data and is used as training data. The PointNet and PointConv network architectures are then used for tire classification. The results show that both approaches provide good accuracy. |
| first_indexed | 2024-03-09T03:24:26Z |
| format | Article |
| id | doaj.art-6393b177e3be4e9c80154d8c33ee776c |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T03:24:26Z |
| publishDate | 2021-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-6393b177e3be4e9c80154d8c33ee776c2023-12-03T15:04:38ZengMDPI AGSensors1424-82202021-01-0121388410.3390/s21030884Multi-Dimensional Underwater Point Cloud Detection Based on Deep LearningChia-Ming Tsai0Yi-Horng Lai1Yung-Da Sun2Yu-Jen Chung3Jau-Woei Perng4Department of Mechanical and Electro-Mechanical Engineering, National Sun Yat-sen University, Kaohsiung 804, TaiwanDepartment of Mechanical and Electro-Mechanical Engineering, National Sun Yat-sen University, Kaohsiung 804, TaiwanNaval Meteorological and Oceanographic Office R.O.C., Kaohsiung 804, TaiwanNaval Academy R.O.C., Kaohsiung 804, TaiwanDepartment of Mechanical and Electro-Mechanical Engineering, National Sun Yat-sen University, Kaohsiung 804, TaiwanNumerous sensors can obtain images or point cloud data on land, however, the rapid attenuation of electromagnetic signals and the lack of light in water have been observed to restrict sensing functions. This study expands the utilization of two- and three-dimensional detection technologies in underwater applications to detect abandoned tires. A three-dimensional acoustic sensor, the BV5000, is used in this study to collect underwater point cloud data. Some pre-processing steps are proposed to remove noise and the seabed from raw data. Point clouds are then processed to obtain two data types: a 2D image and a 3D point cloud. Deep learning methods with different dimensions are used to train the models. In the two-dimensional method, the point cloud is transferred into a bird’s eye view image. The Faster R-CNN and YOLOv3 network architectures are used to detect tires. Meanwhile, in the three-dimensional method, the point cloud associated with a tire is cut out from the raw data and is used as training data. The PointNet and PointConv network architectures are then used for tire classification. The results show that both approaches provide good accuracy.https://www.mdpi.com/1424-8220/21/3/884BV5000deep learningunderwater point cloudunderwater object detection |
| spellingShingle | Chia-Ming Tsai Yi-Horng Lai Yung-Da Sun Yu-Jen Chung Jau-Woei Perng Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning Sensors BV5000 deep learning underwater point cloud underwater object detection |
| title | Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning |
| title_full | Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning |
| title_fullStr | Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning |
| title_full_unstemmed | Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning |
| title_short | Multi-Dimensional Underwater Point Cloud Detection Based on Deep Learning |
| title_sort | multi dimensional underwater point cloud detection based on deep learning |
| topic | BV5000 deep learning underwater point cloud underwater object detection |
| url | https://www.mdpi.com/1424-8220/21/3/884 |
| work_keys_str_mv | AT chiamingtsai multidimensionalunderwaterpointclouddetectionbasedondeeplearning AT yihornglai multidimensionalunderwaterpointclouddetectionbasedondeeplearning AT yungdasun multidimensionalunderwaterpointclouddetectionbasedondeeplearning AT yujenchung multidimensionalunderwaterpointclouddetectionbasedondeeplearning AT jauwoeiperng multidimensionalunderwaterpointclouddetectionbasedondeeplearning |