Skaitmeninio reljefo modelio patikslinimas mozaikos funkcija

Abstract

The introduction presents the importance of the problems, aims and methods of research, scientific novelty and practical importance of the digital terrain model. “Analysis of Methods” presents a review of the techniques of compiling digital terrain models (DTM) depending on statistical analyses of stochastic data. The data conditioned the geographic position and parameters of the algorithms of modeling. DTM are compiled by the kriging methods. DTM was made using a circular variogram. Analysis of the possibilities and accuracy of DTM modeling by means of the kriging method are presented. Initial data were obtained by stereophotogrammetric and analytic methods with a PLANICOMP P2 photogrammetric instrument, PCAP, P-ATM, BINGO (Germany), NLHBUNT (Norway) software and a PHODIS photogrammetry work station (Germany). The DTM was created by the kriging method with ESRI ArcGis software where the size of cells is 2 ´ 2 m. The accuracy of the DTM was evaluated using the Extraction package of ESRI ArcGIS software. “The impact of buffer zone size on the value of DTM standard deviation” presents the DTM applied to a buffer zone from 5 to 35 m. The accuracy of the results are calculated. DTM model accuracy directly depended on the size of the buffer zone. The chapter “Creation of DTM mosaic” describes the possibility of using old and upgraded data for DTM creation. Results were the most accurate using the DTM. General conclusions 1. The modeling terrain model is recommended for evaluating data accuracy. The mean-root-square of the high points was mDz = 0.14 m. 2. The accuracy of DTM is directly dependent on the size of the buffer zone. Standard deviation for the buffer zone 5 m is s = 0.70 m and for 35 m s = 5.97 m. 3. The best results are achieved when it is possible to renew the old DTM data base. During the experiment, the accuracy improved by 33% and standard deviation decreased from 5.09 to 3.39 m.