A small underwater object detection model with enhanced feature extraction and fusion
Abstract In the underwater domain, small object detection plays a crucial role in the protection, management, and monitoring of the environment and marine life. Advancements in deep learning have led to the development of many efficient detection techniques. However, the complexity of the underwater...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-85961-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832594711832952832 |
|---|---|
| author | Tao Li Yijin Gang Sumin Li Yizi Shang |
| author_facet | Tao Li Yijin Gang Sumin Li Yizi Shang |
| author_sort | Tao Li |
| collection | DOAJ |
| description | Abstract In the underwater domain, small object detection plays a crucial role in the protection, management, and monitoring of the environment and marine life. Advancements in deep learning have led to the development of many efficient detection techniques. However, the complexity of the underwater environment, limited information available from small objects, and constrained computational resources make small object detection challenging. To tackle these challenges, this paper presents an efficient deep convolutional network model. First, a CSP for small object and lightweight (CSPSL) module is introduced to enhance feature retention and preserve essential details. Next, a variable kernel convolution (VKConv) is proposed to dynamically adjust the convolution kernel size, enabling better multi-scale feature extraction. Finally, a spatial pyramid pooling for multi-scale (SPPFMS) method is presented to preserve the features of small objects more effectively. Ablation experiments on the UDD dataset demonstrate the effectiveness of the proposed methods. Comparative experiments on the UDD and DUO datasets demonstrate that the proposed model delivers the best performance in terms of computational cost and detection accuracy, outperforming state-of-the-art methods in real-time underwater small object detection tasks. |
| format | Article |
| id | doaj-art-a3eba1f5eb9b4681881c85a2bdda605e |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-a3eba1f5eb9b4681881c85a2bdda605e2025-01-19T12:23:14ZengNature PortfolioScientific Reports2045-23222025-01-0115111310.1038/s41598-025-85961-9A small underwater object detection model with enhanced feature extraction and fusionTao Li0Yijin Gang1Sumin Li2Yizi Shang3School of Human Settlements and Civil Engineering, Xi’an Jiaotong UniversitySchool of Human Settlements and Civil Engineering, Xi’an Jiaotong UniversitySchool of Information Engineering, Minzu University of ChinaChina Institute of Water Resources and Hydropower ResearchAbstract In the underwater domain, small object detection plays a crucial role in the protection, management, and monitoring of the environment and marine life. Advancements in deep learning have led to the development of many efficient detection techniques. However, the complexity of the underwater environment, limited information available from small objects, and constrained computational resources make small object detection challenging. To tackle these challenges, this paper presents an efficient deep convolutional network model. First, a CSP for small object and lightweight (CSPSL) module is introduced to enhance feature retention and preserve essential details. Next, a variable kernel convolution (VKConv) is proposed to dynamically adjust the convolution kernel size, enabling better multi-scale feature extraction. Finally, a spatial pyramid pooling for multi-scale (SPPFMS) method is presented to preserve the features of small objects more effectively. Ablation experiments on the UDD dataset demonstrate the effectiveness of the proposed methods. Comparative experiments on the UDD and DUO datasets demonstrate that the proposed model delivers the best performance in terms of computational cost and detection accuracy, outperforming state-of-the-art methods in real-time underwater small object detection tasks.https://doi.org/10.1038/s41598-025-85961-9UnderwaterDeep learningSmall object detectionMultiscale fusion |
| spellingShingle | Tao Li Yijin Gang Sumin Li Yizi Shang A small underwater object detection model with enhanced feature extraction and fusion Scientific Reports Underwater Deep learning Small object detection Multiscale fusion |
| title | A small underwater object detection model with enhanced feature extraction and fusion |
| title_full | A small underwater object detection model with enhanced feature extraction and fusion |
| title_fullStr | A small underwater object detection model with enhanced feature extraction and fusion |
| title_full_unstemmed | A small underwater object detection model with enhanced feature extraction and fusion |
| title_short | A small underwater object detection model with enhanced feature extraction and fusion |
| title_sort | small underwater object detection model with enhanced feature extraction and fusion |
| topic | Underwater Deep learning Small object detection Multiscale fusion |
| url | https://doi.org/10.1038/s41598-025-85961-9 |
| work_keys_str_mv | AT taoli asmallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT yijingang asmallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT suminli asmallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT yizishang asmallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT taoli smallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT yijingang smallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT suminli smallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion AT yizishang smallunderwaterobjectdetectionmodelwithenhancedfeatureextractionandfusion |