TOWARD CONTINUOUS AND UPDATED 3D GEOSPATIAL TERRAIN MODELLING BASED ON DEM AND LiDAR DATA

The constant need for updated and accurate representation of our natural environment, which can be produced from various mapping and observation tools, is one of the major missions the geoinformation community has to provide a solution for. Digital Terrain Model (DTM) datasets that exist for the last few decades are amongst the main geospatial data widely used and applicable. Light Detection and Ranging (LiDAR), on the other hand, is relatively a newer geospatial measurement tool, which enables the quick production of the scanned surface and its coverage representation. Consequently, utilizing up-to-date and accurate datasets produced from LiDAR measurements for GIS related tasks, such as updating existing Digital Elevation Model (DEM) datasets, should be considered. Nevertheless, using simultaneously these geospatial datasets produced from various sources, which consists of different data class, requires addressing the issue of integrating data produced on different epochs. A simple 'insertion' of scattered LiDAR patches based solely on the reference coordinates systems may result in an ambiguous modelling and evident discontinuities in the produced updated terrain representation. Moreover, raw LiDAR datasets present surface features that need to be filtered out prior to the integration process. This paper presents a novel LiDAR data filtering algorithm accompanied by a two-stage hierarchical integration process. Implementing these enabled a precise global and local monitoring analysis of the inherent inconsistencies in the different datasets, which yields an accurate and continuous modelling and updating of LiDAR data within a wide DEM dataset.