基于改进YOLOX的水库水面漂浮物目标检测算法

Improved YOLOX-based object detection algorithm for water surface floating objects in reservoirs

谭文群 ¹曾祥君 ¹包学才 ¹梁义 ¹许小华²

扫码查看

作者信息

1. 南昌工程学院江西省水信息协同感知与智能处理重点实验室,江西南昌 330099;南昌工程学院信息工程学院,江西南昌 330099
2. 江西省水利科学院,江西南昌 330029
折叠

摘要

针对目前水库水面小目标漂浮物检测识别精度低的问题,提出基于改进YOLOX的水库水面漂浮物目标检测算法.此算法引入新型dark2 模块融入主干网络并拓展主干网络的分支输出结构,提升主干网络对图片的特征提取能力.在此基础上,提出改进特征融合模块(ZL-FPN),用于增强特征图信息融合,提高对水库水面小目标漂浮物的检测精度.结果表明:改进后算法的mAP值比YOLOv4 和原YOLOX算法分别提升了29.93%和12.11%,有效提升了水库水面漂浮物检测精度.研究成果可为提升水库智能化管理水平提供有效技术支撑.

Abstract

To address the issue of low accuracy in detecting small floating objects on a reservoir,we proposed a YOLOX-based de-tection framework for water surface floating object recognition.The proposed detector introduces a novel dark2 module,which was em-bedded into the backbone as a plug-and-play module,to develop the branch structure and enhance feature extraction and representa-tion for given images.Furthermore,we designed a modified feature aggregation module(ZL-FPN)to facilitate the fusion and interac-tion of multi-scale features,and the detection accuracy of small floating objects on a reservoir was improved.The results demonstrated that the proposed model obtained 29.93%and 12.11%performance gains compared with YOLOv4 and the original YOLOX.The re-search findings can provide effective technical support for improving the level of intelligent management of reservoirs.

关键词

水面小目标漂浮物/目标检测/YOLOX算法/水库智能化管理

Key words

small floating objects on water surface/object detection/YOLOX algorithm/intelligent management of reservoirs

引用本文复制引用

基金项目

国家自然科学基金项目(61961026)

江西省科技厅重大科技研发专项"揭榜挂帅"制项目(20213AAG01012)

出版年

2024

人民长江

水利部长江水利委员会

人民长江

CSTPCD北大核心

影响因子：0.451

ISSN：1001-4179

参考文献量20

段落导航