Geração de caminhos em cenários 3D a partir de mapas de custo utilizando Mean shift

Abstract

In game development area, most of the time is spent on content creation (terrain, vegetation, paths, cities). In this way, many methods for automatic generation of content are designed to facilitate and accelerate the process of creating a game. The present paper proposes an automatic method of path generation in 3D scenario. Given a set of input images, which represents the scenario’s features, clusters are determined to build a final cost map, which is used to build the lowest cost path connecting two points in the scenario. This paper presents a system capable of generating cost maps using Perlin noise, classify the input images using Mean shift and find the lowest cost path using A*. The path generated can be exported and can be uses to enrich scenarios or to move characters.