Mapping a vehicle environment
A computer-implemented method and device for mapping a vehicle environment of a vehicle are disclosed. The method comprises determining an occupancy grid representing the vehicle environment, including occupancy probability information of a first set of object detections. The occupancy probability information is determined from first positioning information obtained from a first sensor system. Semantic information and second positioning information associated with the semantic information from one or more semantic information sources is obtained. The semantic information comprises object classification information of a second set of object detections and the second positioning system indicates one or more positions of the second set of object detections with respect to the vehicle. The object classification information of the second set of object detections is combined with the occupancy probability information of the occupancy grid to generate a classified occupancy grid.
1. A method comprising:
determining an occupancy grid representing a vehicle environment, the occupancy grid comprising occupancy probability information of a first set of object detections, the occupancy probability information determined from first positioning information obtained from a first sensor system, the first positioning information indicating one or more positions of the first set of object detections with respect to a vehicle;
obtaining semantic information and second positioning information associated with the semantic information from one or more semantic information sources, the semantic information comprising object classification information of a second set of object detections, the second positioning system indicating one or more positions of the second set of object detections with respect to the vehicle;
determining a semantic grid representing a vehicle environment using the semantic information and the second positioning information, the semantic grid comprising a grid of evidence values for one or more object classification types; and
combining the object classification information of the semantic grid with the occupancy probability information of the occupancy grid to generate a classified occupancy grid.
2. The method of claim 1 , wherein determining the semantic grid comprises:
projecting an object detection of the second set of object detections on to a grid representing the vehicle environment;
determining an object spatial area for the object detection of the second set of object detections based on an uncertainty value of the second positioning information of the object detection; and
assigning cells of one or more classification grids as being occupied by the object detection of the second set of object detections based on the object spatial area, each classification relating to a different classification type.
3. The method of claim 1 , further comprising:
calculating a confidence value for one or more cells of the semantic grid by applying a pignistic transformation to an evidence value of that cell; and
combining the confidence value of one more cells of the semantic grid with an occupancy probability value of one or more corresponding cells of the occupancy grid.
4. The method of claim 3 , wherein combining the confidence value of one or more cells of the semantic grid with an occupancy probability value of one or more corresponding cells of the occupancy grid comprises multiplying the confidence value of the one more cells of the semantic grid with an occupancy probability value of one or more corresponding cells of the occupancy grid.
5. The method of claim 1 , wherein combining the object classification information of the semantic grid with the occupancy probability information of the occupancy grid comprises:
selecting an object detection of the second set of object detections having an object classification;
identifying a provisional location of the object detection of the second set of object detections using the second positioning information;
identifying an occupied region of the occupancy grid corresponding to an object detection of the first set of object detections as being proximate to the provisional location of the object detection of the second set of object detections; and
assigning the object classification of the object detection of the second set of object detections to the occupied region of the occupancy grid corresponding to the object detection of the first set of object detections.
6. The method of claim 5 , wherein identifying an occupied region of the occupancy grid comprises:
comparing a distance between the provisional location of the object detection of the second set of object detections and the occupied region of the occupancy grid to a threshold distance; and
assigning the object classification to the cells of the occupied region of the occupancy grid if the distance is less than or equal to the threshold distance.
7. The method of claim 1 , wherein determining a semantic grid comprises accumulating semantic evidence values in each cell of the semantic grid using an algorithm using the Dempster-Shafer framework.
8. The method of claim 1 , wherein the one or more semantic information sources comprises a camera system configured to capture one or more images of the vehicle environment and to extract the semantic information and the second positioning information from the one or more captured images.
9. A non-transitory computer-readable storage medium storing computer-readable instructions that, when executed by at least one processor, configure the at least one processor to:
determine an occupancy grid representing a vehicle environment, the occupancy grid comprising occupancy probability information of a first set of object detections, the occupancy probability information determined from first positioning information obtained from a first sensor system, the first positioning information indicating one or more positions of the first set of object detections with respect to a vehicle;
obtain semantic information and second positioning information associated with the semantic information from one or more semantic information sources, the semantic information comprising object classification information of a second set of object detections, the second positioning system indicating one or more positions of the second set of object detections with respect to the vehicle;
determine a semantic grid representing a vehicle environment using the semantic information and the second positioning information, the semantic grid comprising a grid of evidence values for one or more object classification types; and
combine the object classification information of the second set of object detections with the occupancy probability information of the occupancy grid to generate a classified occupancy grid.
10. The non-transitory computer-readable storage medium of claim 9 , wherein, to determine the semantic grid, the at least one processor is configured to:
project an object detection of the second set of object detections on to a grid representing the vehicle environment;
determine an object spatial area for the object detection of the second set of object detections based on an uncertainty value of the second positioning information of the object detection; and
assign cells of one or more classification grids as being occupied by the object detection of the second set of object detections based on the object spatial area, each classification relating to a different classification type.
11. A computing device comprising:
a first sensor system;
at least one processor; and
a computer-readable storage medium storing computer-readable instructions that, when executed by the at least one processor, configure the at least one processor to:
determine an occupancy grid representing a vehicle environment, the occupancy grid comprising occupancy probability information of a first set of object detections, the occupancy probability information determined from first positioning information obtained from the first sensor system, the first positioning information indicating one or more positions of the first set of object detections with respect to a vehicle;
obtain semantic information and second positioning information associated with the semantic information from one or more semantic information sources, the semantic information comprising object classification information of a second set of object detections, the second positioning system indicating one or more positions of the second set of object detections with respect to the vehicle;
determine a semantic grid representing a vehicle environment using the semantic information and the second positioning information, the semantic grid comprising a grid of evidence values for one or more object classification types; and
combine the object classification information of the second set of object detections with the occupancy probability information of the occupancy grid to generate a classified occupancy grid.
12. The computing device of claim 11 , wherein, to determine the semantic grid, the processor is configured to:
project an object detection of the second set of object detections on to a grid representing the vehicle environment;
determine an object spatial area for the object detection of the second set of object detections based on an uncertainty value of the second positioning information of the object detection; and
assign cells of one or more classification grids as being occupied by the object detection of the second set of object detections based on the object spatial area, each classification relating to a different classification type.
13. The computing device of claim 11 , wherein the at least one processor is further configured to:
calculate a confidence value for one or more cells of the semantic grid by applying a pignistic transformation to an evidence value of that cell; and
combine the confidence value of one more cells of the semantic grid with an occupancy probability value of one or more corresponding cells of the occupancy grid.
14. The computing device of claim 13 , wherein, to combine the confidence value of the one or more cells of the semantic grid with the occupancy probability value of the one or more corresponding cells of the occupancy grid, the at least one processor is configured to multiply the confidence value of the one more cells of the semantic grid with an occupancy probability value of one or more corresponding cells of the occupancy grid.
15. The computing device of claim 11 , wherein the first sensor system comprises an active positioning sensor system and the one or more semantic information sources comprises a passive positioning sensor system.
16. The computing device of claim 11 , wherein the first sensor system is a radar system, a lidar system, and/or an ultrasound system.
17. The computing device of claim 11 , wherein the computing device comprises the vehicle.