Traffic prediction based on map images for autonomous driving
In one embodiment, in response to perception data perceiving a driving environment surrounding an ADV, a map image of a map covering a location associated with the driving environment is obtained. An image recognition is performed on the map image to recognize one or more objects from the map image. An object may represent a particular road, a building structure (e.g., a parking lot, an intersection, or a roundabout). One or more features are extracted from the recognized objects, where the features may indicate or describe the traffic condition of the driving environment. Behaviors of one or more traffic participants perceived from the perception data are predicted based on the extracted features. A trajectory for controlling the ADV to navigate through the driving environment is planned based on the predicted behaviors of the traffic participants. A traffic participant can be a vehicle, a cyclist, or a pedestrian.
1. A computer-implemented method for operating an autonomous driving vehicle, the method comprising:
in response to perception data perceiving a driving environment surrounding an autonomous driving vehicle (ADV), obtaining, by the ADV, a map image of a regular definition map covering a predetermined proximity with respect to a location associated with the driving environment, from a server or a persistent storage device of the ADV, wherein the map image includes a satellite image contained in the regular definition map;
performing, by the ADV, an image recognition of the map image to recognize one or more objects from the map image, wherein the one or more objects are recognized based on at least one of shapes or colors of the objects obtained from the map image;
extracting, by the ADV, one or more features from the recognized objects, wherein the one or more features indicate a traffic condition of the driving environment, wherein extracting one or more features comprises predicting possible a road sign based on a type of the recognized objects including a roundabout;
predicting, by the ADV, behaviors of one or more traffic participants perceived from the perception data based on the extracted features; and
planning, by the ADV, a trajectory for controlling the ADV to navigate through the driving environment based on the predicted behaviors of the one or more traffic participants.
2. The method of claim 1 , wherein the one or more objects represent an intersection of a plurality of roads.
3. The method of claim 1 , wherein the one or more objects represent a roundabout coupled to a plurality of roads.
4. The method of claim 1 , wherein the one or more objects represent a parking lot along a road.
5. The method of claim 1 , wherein extracting one or more features comprises estimating a curvature of a road based on the map image.
6. The method of claim 1 , wherein the map image does not contain the road sign.
7. A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations, the operations comprising:
in response to perception data perceiving a driving environment surrounding an autonomous driving vehicle (ADV), obtaining, by the ADV, a map image of a regular definition map covering a predetermined proximity with respect to a location associated with the driving environment, from a server or a persistent storage device of the ADV, wherein the map image includes a satellite image contained in the regular definition map;
performing, by the ADV, an image recognition of the map image to recognize one or more objects from the map image, wherein the one or more objects are recognized based on at least one of shapes or colors of the objects obtained from the map image;
extracting, by the ADV, one or more features from the recognized objects, wherein the one or more features indicate a traffic condition of the driving environment, wherein extracting one or more features comprises predicting possible a road sign based on a type of the recognized objects including a roundabout;
predicting, by the ADV, behaviors of one or more traffic participants perceived from the perception data based on the extracted features; and
planning, by the ADV, a trajectory for controlling the ADV to navigate through the driving environment based on the predicted behaviors of the one or more traffic participants.
8. The machine-readable medium of claim 7 , wherein the one or more objects represent an intersection of a plurality of roads.
9. The machine-readable medium of claim 7 , wherein the one or more objects represent a roundabout coupled to a plurality of roads.
10. The machine-readable medium of claim 7 , wherein the one or more objects represent a parking lot along a road.
11. The machine-readable medium of claim 7 , wherein extracting one or more features comprises estimating a curvature of a road based on the map image.
12. The machine-readable medium of claim 7 , wherein the map image does not contain the road sign.
13. A data processing system, comprising:
a processor; and
a memory coupled to the processor to store instructions, which when executed by the processor, cause the processor to perform operations, the operations including
in response to perception data perceiving a driving environment surrounding an autonomous driving vehicle (ADV), obtaining, by the ADV, a map image of a regular definition map covering a predetermined proximity with respect to a location associated with the driving environment, from a server or a persistent storage device of the ADV, wherein the map image includes a satellite image contained in the regular definition map,
performing, by the ADV, an image recognition of the map image to recognize one or more objects from the map image, wherein the one or more objects are recognized based on at least one of shapes or colors of the objects obtained from the map image,
extracting, by the ADV, one or more features from the recognized objects, wherein the one or more features indicate a traffic condition of the driving environment, wherein extracting one or more features comprises predicting possible a road sign based on a type of the recognized objects including a roundabout;
predicting, by the ADV, behaviors of one or more traffic participants perceived from the perception data based on the extracted features, and
planning, by the ADV, a trajectory for controlling the ADV to navigate through the driving environment based on the predicted behaviors of the one or more traffic participants.
14. The system of claim 13 , wherein the one or more objects represent an intersection of a plurality of roads.
15. The system of claim 13 , wherein the one or more objects represent a roundabout coupled to a plurality of roads.
16. The system of claim 13 , wherein the one or more objects represent a parking lot along a road.
17. The system of claim 13 , wherein extracting one or more features comprises estimating a curvature of a road based on the map image.
18. The system of claim 13 , wherein the map image does not contain the road sign.