Method for creating large-terrain simulation scene of unmanned aerial vehicle

Abstract

The invention discloses a method for creating a large-terrain simulation scene of an unmanned aerial vehicle. The method comprises the following steps: preparing elevation data, image data, cultural feature data, three-dimensional model data and texture data required for simulating a flight scene; importing the prepared data into vector data, and converting the vector data into vector data with aspecial vector file format; giving specific information to the vector data, assigning values according to purposes, and constructing a ground object model; determining the number and distance of LOD,the grid sampling interval and the size of virtual texture according to the size of the area, the precision of the texture and the flight height during flight simulation of the unmanned aerial vehicle; and generating a terrain model suitable for hardware rendering through the MetaFlight data model. The unmanned aerial vehicle simulation scene created by the method is basically consistent with an actual flight scene, any cultural characteristic data can be added, an infinite simulation flight scene can be theoretically manufactured, and the average loading and displaying time is within 30 seconds.

Description

Technical field[0001]The invention relates to the field of drone simulation technology, in particular to a method for creating a large terrain simulation scene of the drone.Background technique[0002]The traditional terrain generation method generally uses some commonly used parametric surfaces, such as Bezier surfaces and spline surfaces to generate the required three-dimensional terrain through interpolation and surface fitting methods. This method uses computer geometric modeling.[1], Using European geometry to describe the smooth surface and regular shape of the object. Due to the complexity of its mathematical modeling, the amount of calculation is large and a large number of surface fittings are required[2], The simulated elevations and images are randomly generated, so you need to configure higher computer hardware, the loading time is too long, and the 3D terrain created in this way is too different from the actual terrain[3], There is no immersive experience in the drone fli...

AI Technical Summary

Problems solved by technology

[0002] Traditional terrain generation methods generally use some commonly used parametric surfaces, such as Bezier surfaces and spline surfaces, to generate the required three-dimensional terrain through interpolation and surface fitting methods. This method uses computer geometry modeling [1] , using Euclidean geometry to describe the smooth surface and regular shape of the object, due to the complexity of its mathematical modeling, the amount of calculation is large and a large number of surface fittings are required [2] , the simulated elevation and images are randomly generated, which requires high-end computer hardware, and the loading time is too long, and the 3D terrain created in this way is too different from the actual terrain. [3] , there is no immersive feeling in the UAV simulation flight

Images