北京测绘2026，Vol.40Issue(9)：1225-1231,7.DOI:10.19580/j.cnki.1007-3000.2025110028
一种影像特征挖掘的无监督道路快速提取方法
An unsupervised fast road extraction method based on image feature mining
摘要
Abstract
Roads constitute a crucial component of fundamental geographic information data,and their automated extraction has long been a key research topic in the field of remote-sensing monitoring.Traditional manual recognition is inefficient and time-consuming.Therefore,this study explored road information detection in remote-sensing images based on cluster analy-sis technology.First,the image was segmented;then clustering was performed,combined with an edge detection algorithm and a road-search decision rule to rapidly extract road information.The results demonstrate that the proposed method achieves a comprehensive F1-score above 80%,and the processing time for a single image ranges from 2 to 6 s.It effectively and rapidly extracts road edges,making it feasible as an auxiliary tool in surveying and mapping production and capable of substantially improving work efficiency.关键词
聚类分析/道路提取/智能识别/边缘检测/测绘生产Key words
cluster analysis/road extraction/intelligent identification/edge detection/surveying and mapping production分类
天文与地球科学引用本文复制引用
马红,陈浩..一种影像特征挖掘的无监督道路快速提取方法[J].北京测绘,2026,40(9):1225-1231,7.基金项目
广东省省级科技计划(2021B1212100003) （2021B1212100003）
广东省自然资源厅科技项目(GDZRZYKJ2023001). （GDZRZYKJ2023001）
