Abstract:Automatic ship detection is a crucial task within the domain of maritime transportation management. With the progressive success of convolutional neural networks (CNNs), a number of advanced CNN models have been presented in order to detect ships. Although these detection models have achieved marked performance, several undesired results may occur under complex maritime conditions, such as missed detections, false positives, and low detection accuracy. Moreover, the existing detection models endure large number of parameters and heavy computation cost. To deal with these problems, we suggest a lightweight ship model of detection called DSSM–LightNet based upon the improved YOLOv8n. First, we introduce a lightweight Dual Convolutional (DualConv) into the model to lower both the number of parameters and the computational complexity. The principle is that DualConv combines two types of convolution kernels, 3x3 and 1x1, and utilizes group convolution techniques to effectively reduce computational costs while processing the same input feature map channels. Second, we propose a Slim-neck structure in the neck network, which introduces GSConv and VoVGSCSP modules to construct an efficient feature-fusion layer. This fusion strategy helps the model better capture the features of targets of different sizes. Meanwhile, a spatially enhanced attention module (SEAM) is leveraged to integrate with a Feature Pyramid Network (FPN) and the Slim-neck to achieve simple yet effective feature extraction, minimizing information loss during feature fusion. CIoU may not accurately reflect the relative positional relationship between bounding boxes in some complex scenarios. In contrast, MPDIoU can provide more accurate positional information in bounding-box regression by directly minimizing point distance and considering comprehensive loss. Therefore, we utilize the minimum point distance IoU (MPDIoU) rather than the Complete Intersection over Union (CIoU) Loss to further enhance the detection precision of the suggested model. Comprehensive tests carried out on the publicly accessible SeaShips dataset have demonstrated that our model greatly exceeds other algorithms in relation to their detection accuracy and efficiency, while reserving its lightweight nature.

EfficientShip: A Hybrid Deep Learning Framework for Ship Detection in the River

Efficient Ship Detection in Complex Marine Environments Using Optical Remote Sensing Images

YOLOSeaShip: a lightweight model for real-time ship detection

Ship feature recognition methods for deep learning in complex marine environments

Intelligent Ship Detection in Remote Sensing Images Based on Multi-Layer Convolutional Feature Fusion

High-Efficiency and High-Precision Ship Detection Algorithm Based on Improved YOLOv8n

R-CNN-Based Ship Detection from High Resolution Remote Sensing Imagery

A New Efficient Ship Detection Method Based on Remote Sensing Images by Device–Cloud Collaboration

A Multiscale Ship Detection Algorithm Based on Optical Remote Sensing Image

Ship Detection Via Dilated Rate Search and Attention-Guided Feature Representation

A Lightweight Algorithm for Ship Object Detection in Complex Marine Environments

Ship Detection in High-Resolution Optical Remote Sensing Images Aided by Saliency Information

Improved YOLOv3 Based on Attention Mechanism for Fast and Accurate Ship Detection in Optical Remote Sensing Images

High-order Spatial Interactions Enhanced Lightweight Model for Optical Remote Sensing Image-based Small Ship Detection

Improved YOLOv8n for Lightweight Ship Detection

Ship detection of optical remote sensing image in multiple scenes

An efficient model for small object detection in the maritime environment

Lightweight Single-Stage Ship Object Detection Algorithm for Unmanned Surface Vessels Based on Improved YOLOv5

LMSD-Net: A Lightweight and High-Performance Ship Detection Network for Optical Remote Sensing Images

Ship detection in SAR images based on an improved faster R-CNN

A Prediction and Prior Information Guided SAR Ship Detection Method