地物纹理的复杂性,使地物分类一直是遥感领域研究的热点和难点。机载激光技术可以直接获取地物的三维点云信息,影像数据可以提供丰富的波谱信息,因此,结合机载激光点云和航空影像主被动遥感数据对地物进行分类。首先,将三维点云数据转换成网格数据,并利用顶帽变换算子去除噪声点;然后,对网格数据进行区域分割,引入植被区域限制的分水岭算法生成聚类区域;最后,利用模糊证据理论对不同聚类区域进行地物类型识别。用ISPRS数据集中的机载激光点云和航空影像数据作为实验数据,对本方法进行精度评价。结果表明,与基于单点的证据理论分类方法相比,本方法的分类效果较好。

Due to the complexity of ground object texture, the classification of ground object has always been a hotspot and difficulty in the study of remote sensing. Airborne laser technology can directly obtain three-dimensional point cloud information of ground objects, and the image data can provide rich spectral information. Therefore, in this paper, the features are combined with the airborne laser point cloud and the active and passive remote sensing data of aerial images to classify the features. First, the three-dimensional point cloud data needs to be converted into grid data, and the top hat transform morphology method is used to remove the noise points. Then, the grid data is segmented and the watershed algorithm with vegetation area restriction is introduced to generate the clustering region. Finally, fuzzy evidence theory is used to identify the features of different clustering areas. Using the airborne laser point cloud and aerial image data on the ISPRS dataset as the experimental data, the accuracy of the proposed method is evaluated. The results show that the classification effect of this method is better than that of the single point based evidence classification method.