Simultaneous Localization and Mapping Based on Semantic World Modelling

In mobile robotics the problem of simultaneous localization and mapping is quite complex. However, by using smart constraints, the problem can be reduced considerably. Instead of constraining the issue to a specific robotic system or its movement behavior, we show how semantic environment perception and modeling allows for another point of view and therefore a simple solution for the problem. We present a method for application independent localization and mapping based on semantic landmarks and the concept of visual odometry. Central starting point is a generic landmark definition, allowing for a reduction of the 3d localization problem to a more simple search for an affine transformation in 2d space. These semantic landmarks are simultaneously used to map the surrounding environment of the robot, resulting in a widely applicable world model.