Due to the global climate change, occurrence of landslide has become more frequent with a large magnitude and scale. Landslides-related disasters are considered one of the most catastrophic geohazards causing substantial human, economic and environmental losses. High resolution imaging and topographic information is essential for a better prediction of landslide occurrence. Whereas digital elevation model (DEM) with ~m accuracy are mostly acquired by aerial survey, DEM is not sufficiently accurate enough for mapping for landslide prediction and also expensive. Recent advances in Unmanned Aerial Vehicles-Light Detecting and Ranging (UAV-LiDAR) system provide high resolution topographic information with an abundant 3D point cloud data. In this study, UAV-LiDAR system has been constructed to acquire 3D point clouds and DEM at high resolution. 3D point clouds acquired from the mountainous terrain, having a densely wooded basin and a steep basin are represented and the topographic information was obtained by filtering the obstacles. The system we have built can be used not only for landslide analysis, but also for the maintenance managements of countermeasure facilities.