Comparison of two deep learning methods for ship target recognition with optical remotely sensed data
- PDF / 1,393,352 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 88 Downloads / 185 Views
(0123456789().,-volV)(0123456789(). ,- volV)
S.I. : HIGHER LEVEL ARTIFICIAL NEURAL NETWORK BASED INTELLIGENT SYSTEMS
Comparison of two deep learning methods for ship target recognition with optical remotely sensed data Dianjun Zhang1
•
Jie Zhan1 • Lifeng Tan2 • Yuhang Gao3 • Robert Zˇupan4
Received: 1 July 2020 / Accepted: 18 August 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract As an important part of modern marine monitoring systems, ship target identification has important significance in maintaining marine rights and monitoring maritime traffic. With the development of artificial intelligence technology, image detection and recognition based on deep learning methods have become the most popular and practical method. In this paper, two deep learning algorithms, the Mask R-CNN algorithm and the Faster R-CNN algorithm, are used to build ship target feature extraction and recognition models based on deep convolutional neural networks. The established models were compared and analyzed to verify the feasibility of target detection algorithms. In this study, 5748 remote sensing maps were selected as the dataset for experiments, and two algorithms were used to classify and extract warships and civilian ships. Experiments showed that for the accuracy of ship identification, Mask R-CNN and Faster R-CNN reached 95.21% and 92.76%, respectively. These results demonstrated that the Mask R-CNN algorithm achieves pixel-level segmentation. Compared with the Faster R-CNN algorithm, the obtained target detection effect is more accurate, and the performance in target detection and classification is better, which reflects the great advantage of pixel-level recognition. Keywords Full convolutional network Ship target recognition Pixel level Mask R-CNN Faster R-CNN Optical remote sensing images
1 Introduction As the main body of the ocean, accurate detection of ship targets is essential for marine applications, and it is also a key technology for maintaining the rights of the ocean and ensuring the safety of maritime navigation. In recent years, the construction of marine information infrastructure and
2
Dianjun Zhang [email protected]
Key Laboratory of Department of Culture and Tourism of Information Technology of Architectural Heritage Inheritance School of Architecture, Tianjin University, Tianjin 300072, China
3
Yuhang Gao [email protected]
College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China
4
Faculty of Geodesy, University of Zagreb, 10 000 Zagreb, Croatia
& Lifeng Tan [email protected]
Robert Zˇupan [email protected] 1
satellite remote sensing has also strengthened ocean monitoring, providing abundant video image data for marine security construction. However, at present, the navy, fishery administration, and fisheries departments still use traditional manual methods to monitor offshore waters. The detection indicates low intelligence, and the information cannot be processed in real time. This method not only consumes manp
Data Loading...
