Editorial: Three-Dimensional Mapping, Part 3

摘要

In many fielded applications of mobile robots, maps of the environment play an important role as they support various tasks including perception, planning, and navigation. Whereas the majority of work in the area of map learning has been devoted to the acquisition of two-dimensional representations, recently three-dimensional maps have gained substantial interest in the robotics community. Especially in the context of applications, three-dimensional maps provide better support for a wide variety of tasks. For example, robots that know about the three-dimensional structure of the environment can better avoid obstacles, can more reliably localize themselves, and can more robustly detect objects. Accordingly, three-dimensional representations provide benefits in all applications in which robots are deployed in real-world scenarios. Additionally, three-dimensional models of environments are envisioned to be useful in a wide area of applications, which goes far beyond robotics, such as architecture, emergency planning, interaction, and visualization. In all of these application domains, there is a need for methods that can automatically construct three-dimensional models.