Improved octree filtering algorithm of airborne LiDAR data in forest environment

ObjectiveBy airborne LiDAR point cloud data, the 3D coordinates of ground points can be obtained accurately. In order to enhance the accuracy of point cloud filtering, a filter method of point cloud data was proposed by the filtering analysis using airborne LiDAR point cloud data in Dayekou forest area of Gansu Province, western China.

MethodBased on the octree filtering algorithm, the complex terrain in forest was decomposed into a large number of slopes by the improved model. The accuracy of data segmentation was enhanced by reducing the nodes'size, with retaining the original information of the data. Additionally, for the uneven or tough terrain of forest, the slope analysis was introduced into the filter algorithm, to prevent the undergrowth points from being classified into ground points.

ResultThe experiment results showed that the method could improve the segmentation accuracy of point cloud data to some extent, and the total errors of filtering in three different terrains were 4.57%, 4.75%, 5.83%, respectively. The proved filtering model was practical to filter point cloud under different terrains in forest area.

ConclusionThus, the improved octree filtering algorithm proposed in this paper can make full use of the data structure features to achieve fast and high-precision filtering. Meanwhile, it saves time and cost, and lays the foundation for the subsequent extraction of forest parameters.