高分辨率遥感影像道路信息智能提取  

Intelligent Extraction of Road Information of High-resolution Remote Sensing Images

作　　者：朱杰民 刘陶胜[1] 娄洪武 万轶伟 黄月梅 ZHU Jiemin;LIU Taosheng;LOU Hongwu;WAN Yiwei;HUANG Yuemei(school of Civil and Ssurveying Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,china;Fuzhou Survey Institute Co.,Ltd.,350000,Fuzhou,Fujian)

机构地区：[1]江西理工大学土木与测绘工程学院,江西赣州341000 [2]福州市勘测院有限公司,福建福州350000

出　　处：《长江信息通信》2024年第9期10-12,16,共4页Changjiang Information & Communications

基　　金：大学生创新创业训练计划项目(202210407034)。

摘　　要：道路信息对城市规划、交通管理和基础设施建设至关重要。针对高分辨率遥感影像道路信息提取效率低、边界质量差和可视化程度低等问题,提出了一种基于改进ResUnet模型的遥感影像道路提取方法,并结合pyqt5框架实现了高分辨率遥感影像道路信息智能提取系统开发。文章以马萨诸塞州道路数据集为数据源,通过改进的ResUnet网络模型,实现了对高分辨率影像中道路信息的准确提取。结合高分辨率遥感影像道路信息智能提取系统能够快速、准确地完成道路提取任务,为城市交通管理和规划提供了重要支持。Road information is crucial to urban planning,traffic management and infrastructure construction.Aiming at the problems of low extraction efficiency,poor boundary quality and low visualization of high-resolution remote sensing image,a remote sensing image road extraction method based on improved ResUnet model is proposed,and the intelligent extraction system of high-resolution remote sensing image road information is developed combined with pyqt5 framework.Using the Massachusetts road data set as the data source,this paper achieves an accurate extraction of road information in high-resolution images through an improved ResUnet network model.Combined with the high-resolution remote sensing image road information intelligent extraction system can complete the road extraction task quickly and accurately,which provides important support for urban traffic management and planning.

关 键 词：高分辨率遥感影像 道路信息提取 ResUnet模型 pyqt5 智能提取系统 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]