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.

A lightweight dual dynamic ship detection network with complex background of inland river

Research on Improved Lightweight YOLOv5s for Multi-Scale Ship Target Detection

Improved YOLOv8n for Lightweight Ship Detection

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

Multi-Scale Ship Detection from SAR and Optical Imagery Via A More Accurate YOLOv3

A Ship Detection Model Based on Dynamic Convolution and an Adaptive Fusion Network for Complex Maritime Conditions

A Lightweight Algorithm for Ship Object Detection in Complex Marine Environments

A Ship Detection Model Based on YOLOX with Lightweight Adaptive Channel Feature Fusion and Sparse Data Augmentation

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

YOLO-Ships: Lightweight ship object detection based on feature enhancement

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

Lightweight algorithm for multi-scale ship detection based on high-resolution SAR images

An Efficient Lightweight SAR Ship Target Detection Network with Improved Regression Loss Function and Enhanced Feature Information Expression

A lightweight ship target detection model based on improved YOLOv5s algorithm

Lightweight SAR ship detection algorithm based on attention mechanism

YOLO-Lite: An Efficient Lightweight Network for SAR Ship Detection

Detection Technique Tailored for Small Targets on Water Surfaces in Unmanned Vessel Scenarios

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

YOLOSeaShip: a lightweight model for real-time ship detection

A Multiscale Ship Detection Algorithm Based on Optical Remote Sensing Image

Ship Detection Algorithm Based on YOLOv5 Network Improved with Lightweight Convolution and Attention Mechanism