| An Object-Based Classification Framework for ALS Point Cloud in Urban Area |
| Paper ID : 1130-GEOSPATIAL (R4) |
| Authors: |
|
Erfan Hasanpour Zaryabi *1, Mohammad Saadat Seresht2, Ebadat Ghanbari Parmehr3 1School of Surveying and Geospatial Engineering, University of Tehran, Tehran, Iran 2Dept. of Photogrammetry and Remote Sensing 3Babol Noshirvani University of Technology, dept. of Geomatics, Faculty of Civil Engineering, Babol, Iran |
| Abstract: |
| This article presents an automated and effective framework for segmentation and classification of airborne laser scanning (ALS) point cloud obtained from LiDAR-UAV sensors in urban areas. Segmentation and classification are among the main processes of the point cloud. They are used to transform 3D point measurements into a high-level, semantically rich representation. The proposed framework has three main parts, including the development of a supervoxel data structure, point cloud segmentation based on local graphs, and using three methods for object-based classification. The results of the point cloud segmentation with an average segmentation error of 0.15 show that the supervoxel structure with an optimal parameter for the number of neighbors can reduce the computational cost and the segmentation error. Moreover, weighted local graphs that connect neighboring supervoxels and examine their similarities play a significant role in improving and optimizing the segmentation process. Finally, three classification methods including Random Forest, Gradient Boosted Trees, and Bagging Decision Trees were evaluated. As a result, the extracted segments were classified with an average precision of more than 83%. |
| Keywords: |
| Point Cloud, Segmentation, Classification, Supervoxel, Voxel, Local Graph |
| Status : Paper Accepted (Poster Presentation) |
