论文部分内容阅读
通过激光雷达数据反演林木参数,特别是树高,将极大地推进激光雷达在林业上的应用。本文首先对LiDAR获取的高密度点云进行预处理,生成适合森林参数提取的规则网格数字表面模型,然后利用形态学滤波的方法逐步去掉非地形要素,形成数字高程模型,最后利用数字表面模型减去得到的数字高程模型,可得到正则化数字表面模型,并求出地物的相对高度信息,在林木上就是最终得到所需要的树木平均高度信息。结合真实LH System ALS40数据进行实验,验证了本文方法的可行性。
Inversion of tree parameters, especially tree height, by Lidar data will greatly facilitate the application of Lidar in forestry. In this paper, we first pre-process the high-density point cloud obtained by LiDAR to generate a regular grid digital surface model suitable for forest parameter extraction. Then we use morphological filtering to gradually remove non-terrain elements to form a digital elevation model. Finally, we use digital surface model Subtracting the obtained digital elevation model, the regularized digital surface model is obtained, and the relative height information of the ground object is obtained. On the forest tree, the average height information of the trees is finally obtained. Experiments with real LH System ALS40 data verify the feasibility of this method.