ALGORITHMS AND ARCHITECTURE OF THE SOFTWARE SYSTEM OF AUTOMATED NATURAL AND ANTHROPOGENIC LANDSCAPE GENERATION

Abstract

Context. The problem of automation of the generation of natural and anthropogenic landscapes is considered. The subject of the research is methods of procedural generation of landscapes that quickly and realistically visualize natural and anthropogenic objects taking into account different levels of detail. 
 Objective. The goal of the work is to improve the rendering quality and efficiency of the procedural generation process of landscape surfaces at any level of detail based on the implementation of the developed method.
 Method. The proposed method of visualization involves the construction of a natural landscape using Bezier curves and surfaces and manual editing of individual segments; use of software agents that are responsible for individual steps of generating anthropogenic objects; adaptation of anthropogenic objects to the characteristics of natural landscapes; containerization of three-dimensional objects, which is used in various steps to organize the storage and loading of objects efficiently. A generated heightmap based on the Perlin noise algorithm is used to construct surfaces on individual segments of the natural landscape. Landscape processing software agents are used to unify the design of algorithms for creating and processing information about anthropogenic objects. Correct application operation and error resistance is guaranteed due to the inheritance of a specific interface by all implementations of agents. Containerization with two-level caching ensures the efficiency of display detailing. 
 Results. The developed method is implemented programmatically, and its efficiency is investigated for different variants of input data, which to the greatest extent determine the complexity of visualization objects. 
 Conclusions. The conducted experiments confirmed the efficiency of the proposed algorithmic software and its viability in practice in solving problems of automated landscape generation. Prospects for further research include improvement and expansion of the algorithms for procedural landscape generation, functionality complication of manual visualized object processing, and division of individual objects into separate hierarchies of containers.