Method and system for creating a semantic representation of the environment of a vehicle
The invention relates to a method for creating a semantic representation of the environment of a vehicle, having the following steps: acquiring (S 1 ) sensor measuring data by means of at least one first ( 2 a ) and/or second environment detection sensor ( 2 b ) of the vehicle, detecting (S 2 ) objects and free spaces in the environment of the vehicle, creating (S 3 ) a grid map having occupied and free grid cells based on the detected objects and free spaces; determining (S 4 ) height information by means of the at least one first ( 2 a ) and/or second environment detection sensor ( 2 b ); adding (S 5 ) the height information to each grid cell; allocating (S 6 ) semantic information to the grid cells based on the height information.
1. A method for creating a semantic representation of the environment of a vehicle, the method comprising:
acquiring sensor measuring data by at least one first and/or second environment detection sensor of the vehicle,
detecting objects and free spaces in the environment of the vehicle,
creating a grid map having occupied and free grid cells based on the detected objects and free spaces;
determining height information of the detected objects and free spaces by the at least one first and/or second environment detection sensor;
adding the height information of the detected objects and free spaces to each grid cell; and
allocating semantic information to the grid cells based on the height information of the detected objects and free spaces,
wherein the semantic information is selected from negative height, free space, a detected object that can be driven over conditionally, and/or the detected object that can be driven under, and
wherein the grid map is extended by corresponding layers based on the semantic information.
2. The method according to claim 1 , wherein a driving function is adapted based on the allocated semantic information.
3. The method according to claim 1 , wherein the allocated semantic information is used for self-localization.
4. The method according to claim 1 , wherein the semantic information further includes a directly measured free space.
5. A system for creating a semantic representation of the environment of a vehicle, the system comprising at least one first and/or second environment detection sensor for acquiring sensor measuring data and a data processing device, wherein the data processing device is configured to:
analyze the acquired sensor measuring data and to detect objects and free spaces,
create a grid map having grid cells based on the detected objects and free spaces,
determine height information of the detected objects and free spaces from the sensor data of the first and/or second environment detection sensor,
add the height information of the detected objects and free spaces to the grid cells, and
allocate semantic information to the grid cells based on the height information of the detected objects and free spaces,
wherein the semantic information is selected from negative height, free space, a detected object that can be driven over conditionally, and/or the detected object that can be driven under, and
wherein the grid map is extended by corresponding layers based on the semantic information.