摘要：【中文】一种基于机载激光雷达点云数据的高精DEM构建方法，首先进行点云数据预处理，为点云数据质量检查和点云数据大地定向；然后采用改进的三角网迭代加密滤波法，将每个待判定点赋予一个与距离相关的优先级，保证优先级最高的激光点优先判断，用于地面点云与非地面点云的分离；再使用已知高程点，构建邻近三角形并进行线性内插处理；最后采用逻辑分析法与高程误差分析法对DEM精度进行评价，实现基于机载激光雷达点云数据的高精DEM构建。本发明采用改进三角网迭代加密滤波法获得地面点云，再基于随机森林回归法内插DEM提供精密度，最后通过逻辑分析法与高程误差分析法，对LiDAR点云数据生成的DEM数据进行了精度评价。 【EN】A high-precision DEM construction method based on airborne laser radar point cloud data comprises the steps of firstly, preprocessing the point cloud data, and checking the quality of the point cloud data and orienting the point cloud data in the ground; then, an improved triangulation iterative encryption filtering method is adopted, each point to be judged is endowed with a priority level related to the distance, and the laser point with the highest priority level is ensured to be judged preferentially and used for separating ground point cloud from non-ground point cloud; then using the known elevation points to construct adjacent triangles and performing linear interpolation processing; and finally, evaluating the DEM precision by adopting a logic analysis method and an elevation error analysis method, and realizing the high-precision DEM construction based on airborne laser radar point cloud data. According to the method, ground point cloud is obtained by adopting an improved triangulation network iterative encryption filtering method, DEM is interpolated based on a random forest regression method to provide precision, and finally, the DEM data generated by the LiDAR point cloud data is subjected to precision evaluation through a logic analysis method and an elevation error analysis method.