Support Vector Machine for Land Cover Classification using Lidar Data

M. I. Hariyono, Rokhmatuloh, M. P. Tambunan, R. S. Dewi

Research output: Contribution to journalConference articlepeer-review

1 Citation (Scopus)


The Lidar technology is widely used in various studies for mapping needs. In this study was to extract land cover using Lidar data by incorporating a support vector machine (SVM) approach. The study was located in the city of Lombok, Nusa Tenggara Barat. Image extraction was performed on single wavelength Lidar data to produce intensity and elevation (Digital Surface Model) features. Feature extraction of Lidar data was implemented by using a pixel-based approach. The extracted features used as an attribute for training data to generate the SVM prediction model. The prediction model to predict the types of land cover in the study area such as buildings, trees, roads, bare soil, and low vegetations. For accuracy assessment purposes, we used topographic map available in shapefile format as the reference map and estimated the accuracies of the resulted classifications. In this study, land cover classification used combination bands which improved the overall accuracy by approximately 20%. The use of the intensity data in this band combination was the reason for the increasing accuracy.

Original languageEnglish
Article number012095
JournalIOP Conference Series: Earth and Environmental Science
Issue number1
Publication statusPublished - 1 Nov 2021
Event3rd Southeast Asian Conference on Geophysics: Future Challenges and Opportunities in Geophysics, SEACG 2020 - Bandung, Virtual, Indonesia
Duration: 3 Nov 20205 Nov 2020


Dive into the research topics of 'Support Vector Machine for Land Cover Classification using Lidar Data'. Together they form a unique fingerprint.

Cite this