|M.Sc Thesis||Department of Civil and Environmental Engineering|
|Supervisor:||Prof. Doytsher Yerach|
|Full Thesis text - in Hebrew|
The technological developments occurred in recent years in the field of digital photogrammetry yield the production of reliable digital Orthophoto and other topographic products, such as: Digital Terrain Model (DTM) and Digital Surface Model (DSM). These products are needed especially for various survey related and Geographic Information System (GIS) applications. The existing methods for generating digital Orthophoto require the interpretation of aerial photographs and DTMs. In spite of the fact that the existing methods utilize dense and accurate data, the final product is not always reliable. This is due to several limitations that demand further intervention, for example: double mapping, hidden areas and geometric distortions of objects (such as buildings).
In recent years, Light Detection and Ranging (LiDAR) technology, which is relatively a new surveying tool, is becoming an important source for producing up-to-date and accurate topographic data. This technology, which is based on Airborne Laser Scanner combined with GPS and INS systems, enables to acquire a dense and accurate 3D data of the surveyed area, i.e., Digital Surface Model (DSM). The main idea of this research is to use the advantages of LiDAR data, in particular its accuracy, density and reliability. Hence, utilizing LiDAR data could solve the problems and difficulties mentioned before.
This research presents a robust method for extracting high resolution and accurate DTM from LiDAR data. This is achieved by implementing a classification process of the LiDAR data into two main groups: terrain and off-terrain points. The robust method uses orthogonal polynomials to approximate the terrain relief. To improve the DTM's description accuracy, the implementation of morphological filter and segmentation process (achieved by region growing algorithm) is carried out. The results of these procedures yielded the classification of LiDAR data which is used to extract an accurate DTM.
In order to achieve a comprehensive solution for Orthophoto generation, the extraction of man-made objects, such as buildings, as well as solving the objects' geometric distortions are required. For these, a segmentation based on region growing algorithm is implemented, finalizing with a generalization that utilizes least square technique. These enable the reconstruction of buildings' roofs, and hence the extraction of hidden areas by using Z-buffer algorithm designated for precluding double mapping effect.
The complete processes and algorithms implemented in this research enables the generation of three rectified images: terrain relief, objects and hidden areas. By integrating these rectified images an accurate and reliable digital Orthophoto is being generated.