Ship Detection with Deep Learning in Optical Remote-Sensing Images: A Survey of Challenges and Advances

Tianqi Zhao,Yongcheng Wang,Zheng Li,Yunxiao Gao,Chi Chen,Hao Feng,Zhikang Zhao

DOI: https://doi.org/10.3390/rs16071145

IF: 5

2024-03-26

Remote Sensing

Abstract:Ship detection aims to automatically identify whether there are ships in the images, precisely classifies and localizes them. Regardless of whether utilizing early manually designed methods or deep learning technology, ship detection is dedicated to exploring the inherent characteristics of ships to enhance recall. Nowadays, high-precision ship detection plays a crucial role in civilian and military applications. In order to provide a comprehensive review of ship detection in optical remote-sensing images (SDORSIs), this paper summarizes the challenges as a guide. These challenges include complex marine environments, insufficient discriminative features, large scale variations, dense and rotated distributions, large aspect ratios, and imbalances between positive and negative samples. We meticulously review the improvement methods and conduct a detailed analysis of the strengths and weaknesses of these methods. We compile ship information from common optical remote sensing image datasets and compare algorithm performance. Simultaneously, we compare and analyze the feature extraction capabilities of backbones based on CNNs and Transformer, seeking new directions for the development in SDORSIs. Promising prospects are provided to facilitate further research in the future.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

### Problems Addressed by the Paper

This paper primarily focuses on the issue of ship detection in optical remote sensing images and discusses in detail the challenges faced in this field and the corresponding improvement methods. Specifically:

1. **Objectives of Ship Detection**:
- Automatically identify whether there are ships in the image and accurately classify and locate them.

2. **Importance of Ship Detection**:
- It has important applications in fields such as fishery management, maritime patrol, and maritime rescue, contributing to ship traffic management and maritime safety.

3. **Challenges Faced**:
- Complex marine environments, insufficient feature distinction, large scale variations, dense and rotational distribution, large aspect ratios, and imbalance between positive and negative samples.

4. **Limitations of Existing Methods**:
- Early methods relied on fixed patterns and manually designed features, resulting in large redundant computations and weak anti-interference capabilities.
- Although deep learning methods can extract stronger semantic information, the results still need improvement under the influence of complex marine environments and ship characteristics. Balancing accuracy and speed is also a major challenge.

5. **Unique Perspective of the Literature Review**:
- Unlike other review papers, this paper focuses on identifying the challenges in ship detection and proposing corresponding solutions.
- It systematically reviews traditional methods, convolutional neural network (CNN)-based methods, and Transformer-based methods.
- Summarizes ship datasets and evaluation metrics, and for the first time, separates and aggregates ship information in comprehensive datasets.
- Analyzes the performance improvement effects of existing solutions and the feature extraction capabilities of different backbone networks.

In summary, this paper aims to provide guidance and reference for further research in this field by deeply analyzing the challenges in ship detection and their solutions.