In robotic terminology mapping is the process by which a robot moves around its environment, scans it with its sensors and makes representations of the environment. These representations of the environment made by the robot is called the map.
What are the forms of representing a map?
With range sensors the following representations are possible:
- Raw map of sensor readings: This is a map obtained by simply plotting the range measurements obtained by the sensor with reference to a reference frame. Such a map looks like this
- A grid based map: In a grid based representation each cell in the grid is marked as occupied or unoccupied by an obstacle. Such a representation looks like this
- A feature based map: In this representation the map is shown in terms of geometric primitives such as line segments and corners.
- Topological map: A topological map is in general a higher level representation. It shows a network of topological features such as rooms, corridors and intersections. A typical topological map looks as follows.
Why is mapping important?
Mapping is pivotal for autonomous navigation as well as for robot localization. Through the representations of the environment a robot can plan collision free paths or it could make use of the map to determine its location in the environment.