Digital Surface Model Filtering for Building Object Extraction
Handoko Dwi Julian

Universitas Indo Global Mandiri

Abstract

Light Detection and Ranging (LiDAR) technology offers height data in the form of point cloud that can be used to extract building objects in urban areas.
Building object extraction is done using software which is carried out interactively and uses Digital Surface Model and Digital Terrain Model data which are produced from LiDAR point cloud data processing.
The value of Digital Surface Model and Digital Terrain Model is then taken to the difference value to obtain the value of Normalized Digital Surface Model (nDSM)
Image segmentation method is used to extract building objects using normalized Digital Surface Model (NDSM) data and using orthophoto image data as additional data
The results of the study include the number of buildings successfully extracted using NDSM data and orthophoto imagery in urban areas
Buildings that have not been extracted have the maximum results so they need to be further studied by adding other data such as LiDAR intensity data.

Keywords: Point Cloud, Segmentasi, LiDAR

Topic: Computer Science