一种基于超像素分割的遥感图像道路提取方法

A Road Extraction Method for Remote Sensing Images Based on Super-Pixel Segmentation

遥感图像中道路提取在军事、交通等领域具有重要的应用价值。在相关研究分析的基础上,提出了一种基于超像素分割的遥感图像道路提取方法。首先采用SLIC算法对高分辨率遥感图像进行分割处理;其次在手工标注道路起止点的基础上,基于超像素的最短路径算法对样本集的道路像素点数据进行扩充和积累;最后采用XGBoost模型进行道路提取。试验结果表明,相对于SVM、GBDT等算法,XGBoost算法具有较好的准确性,能够清晰地提取出道路网络。

Road extraction in remote sensing images has important application value in military,transportation,and other fields.Based on the relevant research and analysis,a remote sensing image road extraction method based on super-pixel segmentation is proposed in the paper.Firstly,the high-resolution remote sensing image is divided by using SLIC algorithm.Secondly,on the basis of manually marking the start and end point of the road,the road pixel data for the sample set is expanded and accumulated based on the shortest path algorithm of super-pixel.Finally,the XGBoost model is used for road extraction.The experimental results show that XGBoost algorithm has good accuracy and can extract the road network clearly compared with SVM,GBDT and other algorithms.