Abstract
Currently, information on the spatial description of objects is used in many areas of human activity. One of these types of information is the coordinates of objects. Such data are used in cartography, in the construction of digital maps and 3D models, for the operation of navigation aids, etc. In the automated creation of digital models of terrain relief, one of the main qualitative indicators is the accuracy of determining the height of objects. The main influence on this indicator is made by the parallax measurement error when processing stereo images. To obtain a formula for calculating the accuracy of measuring the height of objects, let’s use the expansion of the function in a Taylor series. Using the Cramer-Rao formula for the potential accuracy of measuring the coordinates of the image of the object in the image, the Fourier transform and Parseval's equality, the formula for the potential accuracy of combining stereo images (parallax measurements) is obtained. The analysis of the obtained formulas shows that the image alignment accuracy deteriorates with an increase in the noise power spectral density in the first and second images and a decrease in the similarity of one image to another, as well as with a decrease in the effective width of the mutual spatial spectrum of stereo images. As the value of the stereo recognition basis increases, the error in measuring the heights of objects first improves, and then worsens. This deterioration is due to the fact that stereo pair images are obtained from different spatial points and at the same time perspective distortions and distortions in relief appear on the images. Accordingly, with an increase in the basis of shooting, these distortions will increase. This approach can be used when planning the mode of stereo shooting and equipment for removing the earth's surface for mapping, obtaining 3D models, etc.
Highlights
At present, coordinate information about objects located on the earth’s surface is necessary for constructing digital terrainDETERMINATION OF THE ERROR OF MEASURING THE HEIGHTS OF THE OBJECTS DURING THE AUTOMATIC PROCESSING OFSTEREO PICTURES measurement error, that is, the difference in the coordinates of the same point in the first and second pictures of the stereo pair
Let’s take, as a first approximation, an additive model of the influence of errors in determining the variables included in this of objects [4]. It is this indica- fact that stereo pair images are obtained from different spatial formula
After expanding this tor that affects the quality of a points and at the same time perspective distortions and dis- function into a Taylor series in digital card and, as a result, the tortions in relief appear on the images
Summary
Coordinate information about objects located on the earth’s surface is necessary for constructing digital terrain. One of the main indicators affecting the quality of the obtained DRM is the accuracy of obtaining the coordinates of objects. There are many software products for building and analyzing DRM. These are such GIS packages and modules as Spatial Analyst, 3D Analyst, ArcGIS packages, MapInfo programs (MapInfo Corp.), Autodesk Map 3D AutoCAD systems (Autodesk Inc.) and others. Common to them is the automation of DRM construction, including the combination of stereo pair images, which is necessary for measuring the heights of objects [2, 3]. When creating a DRM of the earth’s surface, one of the main qualitative indicators is the accuracy of determining the height.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
