Method and system for automated vehicle charging

1. A method for facilitating automated vehicle charging, comprising:

at a camera of a vehicle charging system, capturing an image depicting a portion of a vehicle, the portion comprising a cover arranged between the camera and a charging connector of the vehicle;

based on the image, determining an edge map defining a set of components;

determining a set of feature vectors, comprising, for each component of the set, determining a respective feature vector based on the component;

based on the set of feature vectors and a set of expected feature vectors for a cover, selecting a candidate component;

based on the candidate component, determining a set of geometric features;

based on the set of geometric features, determining a relative orientation of the vehicle with respect to the camera; and

based on the relative orientation, controlling the vehicle charging system to reposition the camera such that the camera and vehicle substantially define a desired relative orientation.

2. The method of claim 1 , wherein determining the edge map comprises:

determining a subset of the image, wherein the subset depicts the cover; and

detecting a set of edges within the subset of the image, wherein the edge map is representative of the set of edges.

3. The method of claim 2 , wherein determining the subset of the image comprises detecting a depiction of the cover using a neural network.

4. The method of claim 2 , wherein the set of components consists of a component.

5. The method of claim 1 , wherein, for each component of the set, the respective feature vector comprises a respective circularity feature.

6. The method of claim 5 , wherein, for each component of the set, determining the respective circularity feature comprises:

determining a convex hull of the component; and

determining a circularity of the convex hull, wherein the respective circularity feature is equal to the circularity.

7. The method of claim 5 , wherein, for each component of the set, the respective feature vector further comprises at least one of a respective eccentricity feature and a respective rectangularity feature.

8. The method of claim 1 , wherein the set of geometric features comprises a candidate component aspect ratio.

9. The method of claim 1 , wherein the set of geometric features comprises a plurality of candidate component dimensions.

10. The method of claim 1 , wherein the set of geometric features further comprises a candidate component location within the image.

11. The method of claim 1 , wherein controlling the vehicle charging system to reposition the camera comprises:

based on the relative orientation, performing a first camera repositioning;

after performing the first camera repositioning, capturing a second image at the camera;

based on the second image, determining a revised relative orientation of the vehicle with respect to the camera; and

based on the revised relative orientation, performing a second camera repositioning.

12. The method of claim 1 , further comprising, after controlling the vehicle charging system to reposition the camera:

at the camera, capturing a second image depicting the cover; and

based on the second image, controlling an actuator of the vehicle charging system to actuate the cover.

13. The method of claim 12 , wherein the actuator is rigidly coupled to the camera.

14. The method of claim 12 , wherein controlling the actuator to actuate the cover comprises, based on the second image:

determining a location of a cover actuation point;

determining a distance between the actuator and the cover actuation point; and

controlling the actuator to contact the cover actuation point.

15. The method of claim 1 , further comprising, after controlling the vehicle charging system to reposition the camera:

opening the cover;

at the camera, capturing a second image depicting the charging connector;

based on the second image, determining a location of the charging connector relative to the camera; and

based on the location, connecting an electrical power supply to the charging connector.

16. A method for facilitating automated vehicle charging, comprising:

at a camera of a vehicle charging system, capturing an image depicting a portion of a vehicle, the portion comprising a cover arranged between the camera and a charging connector of the vehicle;

using a computer vision technique, determining a location of a depiction of the cover within the image;

based on the location, determining a subset of the image, wherein the subset depicts the cover;

determining an edge map of the subset of the image, the edge map defining a component;

based on the component, determining a set of geometrical features; and

based on the set of geometrical features and a set of expected geometrical features for the cover, determining that the component is representative of the cover.

17. The method of claim 16 , wherein using the computer vision technique comprises providing the image to a region-based convolutional neural network.

18. The method of claim 2 , wherein determining the edge map comprises:

using a first edge detection threshold, detecting a first set of edges within the subset of the image;

based on the first set of edges, determining a first edge pixel count;

determining that the first edge pixel count is not within an expected edge pixel count range;

based on the first edge pixel count, determining a second edge detection threshold different from the first edge detection threshold; and

in response to determining that the first edge pixel count is not within the expected edge pixel count range, using the second edge detection threshold, detecting a second set of edges within the subset of the image, wherein the edge map comprises the second set of edges.

19. The method of claim 18 , wherein determining the edge map further comprises, based on the second set of edges, determining a second edge pixel count, wherein the second edge pixel count is within the expected edge pixel count range.

20. The method of claim 16 , further comprising, after determining that the component is representative of the cover:

based on the image, determining a relative position and orientation of the cover with respect to the camera; and

based on the relative position and orientation, controlling the vehicle charging system to actuate the cover.