基于机载LiDAR粗糙度指数和回波强度的道路提取

Road Extraction Based on Roughness Index and Echo Intensity of Airborne LiDAR

机载LiDAR点云数据能提供回波强度和地面粗糙度指数,为遥感信息自动提取增添了有价值的约束信息。将这些信息引入到道路自动提取之中,结合光谱特征和道路描述因子建立了一种面向对象的道路联合提取方法。首先,由点云数据衍生归一化数字表面模型(nDSM)和粗糙度指数;然后对配准后的多源数据进行多分辨率分割,进而使用粗糙度指数和回波强度、道路描述因子等特征进行分类;最后,去除道路噪声,并获取准确的道路骨架网。实验结果表明,该方法对道路提取的准确度达95%以上。

Airborne LiDAR point cloud data includes echo intensity and roughness index. Introducing these in- formations into the automatic road detection process and combining with the spectral feature and road descrip- tion factors, an associated object-oriented detection method was established. Firstly, the nDSM （normalized Digital Surface Model） and roughness index are derived from the point cloud data. Secondly, the classified objects are produced through the multi-resolution segmentation on the registration multi-resource data. Thirdly, the classification is processed using roughness index, echo intensity and road description factors. Finally, the road noise is filtered and the road skeleton is obtained. The experimental results showed that the accuracy for road detection was more than 95 %.