By: Itay Bar-On:
Abstract:
The research provides insights to a new photogrammetric approach that deals with the quality of Digital Surface Model (DSM) generation, spatially. With the use of Leica Photogrammetry Suite (LPS) software module, aerial triangulation was preformed twice. First, with two high-resolution aerial images (UltraCam images) and secondly with orthoimages produced from the first aerial triangulation. For the generated DSM’s, validation took place by means of Differential GPS (DGPS) point analysis and by generating an orthoDSM (DSM from orthoimages). Furthermore, the DSM generated from the best image layer is used to estimate and compare forest volume with conventional forestry measurements.
The main results indicate that an overestimation of heights takes place when generating a DSM and there are significant differences when using default or expert DSM generation settings. The main research conclusions are that the DSM validation should not be done by the internal validation procedure. The potential to estimate spatially DSM generation errors, with the orthoDSM generation and the significance of Ground Control Points (GCP’s) distribution in the DSM generation are shown.
Further research and improvements should take place to develop a fully functional and adaptive spatial DSM error estimation tool.