海洋测绘2025，Vol.45Issue(5)：26-30,5.DOI:10.3969/j.issn.1671-3044.2025.05.006
基于改进的YOLOv7沉船目标检测算法研究
Research on shipwreck target detection algorithm based on improved YOLOv7
摘要
Abstract
Aiming at the low recognition accuracy of traditional underwater target detection model,this paper proposes an improved target detection algorithm based on YOLOv7 using side-scan sonar wreck image as data set,and conducts detection on the measured strip image to verify the effectiveness of the improved model.Firstly,GAM attention mechanism is added to the neck network to improve the performance of the network by reducing information loss and amplifying global interaction.Then,some Concat splicing layers in the network are replaced by BiFPN structure to achieve efficient bidirectional cross-scale connection and weighted feature fusion.Finally,YOLOv7-P network is formed.The experimental results show that compared with the initial model,the improved model has an increase of 4.1 percentage points in F1 value,an increase of 4.4 percentage points in mAP50 accuracy,reaching 98.1％,and an increase of 4.1 percentage points in MAP50:95 accuracy,reaching 80.9％.Overall,the improved effect is significant.It provides an effective model for underwater target detection and recognition.关键词
海洋测绘/深度学习/目标检测/侧扫声纳/YOLOv7模型Key words
hydrographic surveying and charting/deep learning/object detection/side-scan sonar/YOLOv7 model分类
天文与地球科学引用本文复制引用
陈俊杰,王晓,朱邦彦,陈文广,刘家琪..基于改进的YOLOv7沉船目标检测算法研究[J].海洋测绘,2025,45(5):26-30,5.基金项目
国家自然科学青年基金(41806117) （41806117）
江苏省海洋科技创新项目(JSZRHYKJ202201). （JSZRHYKJ202201）
