The classification of airborne laser scanning (ALS) point clouds is a necessary step in the process of digital elevation model (DEM) construction. Motivated by the surface interpolation filtering scheme, a multi-resolution hierarchical classification (MHC) method was developed to perform
classification in our previous research. However, similar to almost all filtering methods, MHC suffers from the problem of misclassifying off-terrain and ground points in complex landscapes with steep slopes and large discontinuities. To improve the classification accuracy in the above landscapes,
the original MHC is improved in this article by adopting a region growing-based segmentation method to maximally capture the initial ground seeds of MHC. Furthermore, considering the high sensitivity of principle component analysis (PCA) to outliers, a robust PCA approach based on least trimmed
squares (LTS) is proposed to estimate point normals. Fifteen groups of data sets provided by the International Society for Photogrammetry and Remote Sensing (ISPRS) Working Group III/3 are utilized to compare the performances of the improved and original MHCs. Results indicate that the improved
MHC is more accurate than the original MHC, irrespective of accuracy measures. In addition, we also compare the results of our method with three new filtering methods developed in the years 2013 and 2014. It is found that regardless of landscapes (i.e. urban and rural samples), the improved
MHC is averagely more accurate than the three filtering methods in terms of type II error, and irrespective of accuracy measures (i.e. types I, II and total errors), our method outperforms the three filtering methods for the rural samples. In short, the robust segmentation-based MHC demonstrates
great potential to ALS-derived DEM construction in rural areas.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
State Key Laboratory of Mining Disaster Prevention and Control Co-founded by Shandong Province and the Ministry of Science and Technology, Shandong University of Science and Technology, Qingdao, China
Shool of Geodesy and Geomatics, Wuhan University, Wuhan, China
Department of Information Engineering, Shandong University of Science and Technology, Taian, China
Shandong Provincial Key Laboratory of Geomatics and Digital Technology of Shandong Province, Shandong University of Science and Technology, Qingdao, China
Publication date: February 16, 2016
More about this publication?