【摘要】 提出了一种在高分辨率遥感影像上自动提取交通枢纽,即重要道路交汇口的有效方法。方法的基本过程如下:利用高斯马尔可夫随机场纹理模型与支撑向量机进行图像分类得到道路块,通过细化得到道路轴线并登记宽度信息,然后通过启发式规则进行轴线连接得到道路网。而后在路网基础上,自动提取高宽度道路轴线之间的交点作为交通枢纽。通过实验证明了方法的有效性。研究可广泛应用于城乡交通基础设施规划、物流规划、军事作战分析等领域。更多还原

【Abstract】 An effective method to extract transportation centers from high spatial resolution imagery is proposed.Firstly,road patches are obtained with combining Gaussian Markov random field texture model and support vector machine,and the road axes are obtained and their widths are registered with the thinning operator,and then the road network is connected with the heuristic strategy.Based on the extracted road network,the centers can be extracted by detecting the nodes of the high class road axes.The method is validated with the experiments on IKONOS imagery,which can be widely applied into the areas of transportation infrastructure and logistics planning,military decision,etc.更多还原