高分辨率SAR影像道路交叉口自动提取方法

摘要

针对高分辨率SAR影像道路交叉口提取易受周围地物干扰的问题,设计了一种利用三角形检测模型提取道路交叉口的方法.该方法首先采取滤波与锐化消除道路交叉口区域噪声点并凸显道路边缘;接着根据灰度特征,通过形态学运算确定道路交叉口候选区域;最后在候选道路交叉口区域运用三角形检测模型,确认道路交叉口.使用3 m TanDEM-X数据进行实验,结果证明算法对各种常见二维道路交叉口模型均有较好识别效果.%In high resolution SAR images,there are various kinds of objects around road junctions,which interfere with the extraction of road junctions.Therefore,a new method using triangle detection model is proposed for detecting and identifying road junctions in this paper.Firstly,noise reduction and edge enhancement are done by filtering and sharpening.Secondly,based on the junctions gray feature,the center positions of the road junctions' candidate regions are obtained by using morphological transformation methods.Finally,road junctions are extracted using the triangular model.The experiment results indicate that this method is effective to various common road junctions extracting based on 3 m high-resolution TanDEM-X image.