Review article

Influence of the Choice of Pit Removal Method on Digital Elevation Model Obtained by Image Correlation

Sanja Šamanović orcid.org/0000-0003-1212-5583 ; Faculty of Geodesy, University of Zagreb, Zagreb, Croatia
Damir Medak ; Faculty of Geodesy, University of Zagreb, Zagreb, Croatia
Duška Kunštek ; Faculty of Civil Engineering, University of Zagreb, Zagreb, Croatia

Abstract

Geoinformatic analysis on large areas imposes the need for increasing automatization not only in processing, but also in data collecting. Research requisite was the method of image correlation, which automates and accelerates the process of collecting the raster elevation dataset. Data was collected by the combination of two basic methodologies, area based and feature based matching methods. In the resulted Digital Elevation Model (DEM) the depressions were removed. Depressions in the DEM prevent the simulation of the surface runoff and can be either the actual terrain features or the result caused by sampling error. Depression removal was performed by two basic methods implemented in GIS software, the carving or branching method and the filling method. Both methods were used on three geomorphometrically different areas of Croatia: the mountain range Biokovo, the mountain Papuk and lowlands in the municipality Nijemci. Image correlation was done in two resolutions, 5x5 and 25x25 meters. Data analysis included the geomorphometrical and the quantitative analysis, and for each area, the method for depression removal was recommended as the method that changes the least the geomorphometry of the DEM.

Keywords

DEM; image correlation; area based correlation; feature based matching correlation; depression removal; carving or branching method; filling method

Hrčak ID:

178884

URI

https://hrcak.srce.hr/178884

Publication date:

31.12.2016.

Article data in other languages: croatian

Visits: 1.740 *