The application of current aerophotos for creation of digital terrain models

Abstract

Maps, plans, schemes are used to represent the surface of the earth. But also we can represent surface as a Digital Terrain Models (DTM). Digital Terrain Model is – three-dimensional topographic presentation of some part of Earth's surface. The excelling is representing as contour line on maps, plans and schemes but on DTM, the excelling is represented as different spectral brightness levels. DTM’s are permitting to simplify receiving of geographical information about objects with elevation using computers and appropriate software. We can receive DTM using different methods: radiometric survey, analyzing stereo photo and mechanical postponement contour lines with GIS-technologies. NASA launched a radiometric satellite Shuttle and had received radiometric photos with relief – CGIAR. We used these materials as the additional data. As a base for creation of digital terrain model we used twelve aerial photographs in three routes.

The work program was intended to prove possibility of using aerial photographs for creation DTM in ERDAS Imagine (Leica Photogrammetry Suite) software. The goal of work was creation of electronic version of Digital Terrain Model with highest possible resolution. Aero photos were created by camera RC20, with objective 15/4 UAGA-F and covering the territory of part of Zakarpattja region. The descriptions of survey system are focus – 152.78 mm, principal points of symmetry – x=-0.007 mm and y=0.000 mm, average height – 3820 m. The parameters of interior and exterior were received from Calibration Certificate, which was provided with aerial photographs. As experience shows, it is better to process only two shots simultaneously, which have longitudinal overlap, because while processing more photos occurs mean-square error. First of all, we need to upload pictures to LPS and need to build a pyramid of channels (Pyr) (for color image representation in RGB-format), set the elements of the internal (Int.) and external orientation (Ext.) of images. Interior orientation for data is the aerial photographs of 8 marginal points, marked with special labels to determine the position the center of the photo. The RMSE (Root Mean Square Error) must be less than 0.33 pixels. The elements of external orientation of images are the coordinates of centers of photos, elevation of photographing, as indicated on each image and parameters of tilt and rotation of images (Omega, Phi and Kappa).

After setting of these parameters, we can start setting of Control, Check and Tie Points (synonym support points). For additional geocoding of images we used topographic map (M 1 : 50 000) and CGIAR. We were working with UTM WGS 84 coordinate system. We used topographical map and CGIAR to set the elevation of Check points. Additional, we may use data, which we can receive from highly accurate GPS. Ground Control Points (GCP) – these are points, which have full collection of ground coordinates (X, Y and Z) in different coordinate systems. For additional control of Z-coordinate, we used topographical map with contour lines. The minimum number of GCP in affine model is three. Gathering of GCP is very hard and time-consuming process, accuracy of its implementation affects on the accuracy of triangulation and orthorectification processes. GCPs were gathered across the roads, at the edges of landmark buildings or land. We can’t set GCPs on the bonds of rivers, lakes etc, as these points are dynamic. We have to set more GCPs and to control changing of Root Mean Square Error for correct triangulation and orthorectification. Therefore to control RMSE we can use the report of triangulation process, which shows the errors of each GCP so we may remove or edit the ones which have the greatest deviation. When we receive satisfactory accuracy, we can start creation of DTM and ortorectification process. In additional, we must set other parameter – spatial resolution of DTM. This resolution is limited by resolution of aerial photographs. The accuracy obtained by the digital terrain model can be estimated on the basis of large-scale topographic maps, aerial photographs or other more detailed digital model. At small areas we can verify quality of DTM using high-precision GPS receivers or using way of tacheometric removal of known binding. Based on digital terrain models created for each pair of images were executed orthorectification. Such images can be used for precise measurements of position, the creation of various GIS systems, since they account for the relief of the territory and have a spatial geographic location. In the specialized GIS software it is easy to perform modeling and forecasting of various phenomena basing on orthorectificated images, significantly increases the accuracy of various objects spaces measuring after decryption.