影像与激光雷达数据的城市地物信息提取

Extracting urban ground object information from images and LiDAR data

针对城市地物信息提取中地物边界难以确定、分类精度不高的问题,该文提出一套综合利用影像及激光雷达点云高程信息的面向对象分类方法。在分割中,各类地物的最佳分割尺度由监督法分割精度评价确定,最终分割结果利用粒度理论下的分割尺度综合方法进行合成,能兼顾不同地物最优分割尺度,获得准确地物边界;在分类中,采用ReliefF特征选择算法度量从影像及点云数据提取的对象特征重要度,选择最佳特征组合,并采用多分类器组合方法进行分类,以消除Hughes现象,提高分类精度。选择德国斯图加特市两块实验区进行分类实验,结果表明:该方法有利于提高大范围城市地物精细信息提取的精度和效率,具有较高的应用价值。

The paper proposed an object-oriented classification method using aerial image and LiDAR data to extract urban ground objects.Firstly,the optimal segmentation scales of different ground objects were selected and synthesized to get accurate object boundaries.Then,ReliefF algorithm was used to select the optimal feature combination and eliminate the Hughes phenomenon.Finally,using the multiple classifier combination method to get the classification result.The classification results of two study regions in Stuttgart showed that this method improved the accuracy and efficiency of urban ground object extraction.