Systems and methods for determining real-time lane level snow accumulation

1. A method comprising:

inputting an image of a road into a scene se mentation neural network, the scene segmentation neural network is trained to output an edge map of the road based on the image of the road, the edge map indicates edges of objects and boundaries between different objects in the input image;

inputting the image, a map of the road comprising a road geometry of the road, and the edge map into a trained regressor neural network;

determining an estimated snow depth for each of one or more lanes of the road based on an output of the regressor neural network;

determining an optimal lane of the road for a vehicle to drive in based on the estimated snow depth for the one or more lanes of the road; and

operating the vehicle autonomously to follow the determined optimal lane.

2. The method of claim 1 , wherein the regressor neural network is trained using supervised learning techniques with training data having ground truth labels comprising measurements of snow depths.

3. The method of claim 1 , further comprising:

receiving an aggregate snow depth for the one or more lanes of the road from an edge computing device.

4. The method of claim 3 , further comprising:

determining a difference between the estimated snow depth for the one or more lanes of the road and the aggregate snow depth for the one or more lanes of the road;

determining whether the difference is greater than a threshold value; and

upon determination that the difference is greater than the threshold value, transmitting the estimated snow depth to the edge computing device.

5. The method of claim 3 , further comprising:

determining a difference between the estimated snow depth for the one or more lanes of the road and the aggregate snow depth for the one or more lanes of the road;

determining whether the difference is greater than a threshold value; and

upon determination that the difference is not greater than the threshold value, withholding transmitting the estimated snow depth to the edge computing device.

6. The method of claim 3 , further comprising:

displaying the aggregate snow depth for the one or more lanes of the road such that a simulated level of the aggregate snow depth for the one or more lanes is visible to one or more occupants of the vehicle.

7. The method of claim 3 , further comprising:

displaying an indication of the optimal lane such that the indication is visible to one or more occupants of the vehicle.

8. The method of claim 1 , further comprising:

determining whether the estimated snow depth for at least one lane of the one or more lanes of the road is above a threshold value; and

upon determination that the estimated snow depth for at least one lane of the one or more lanes of the road is above the threshold value, broadcasting the estimated snow depth to one or more other vehicles within communication range.

9. A vehicle system for a vehicle comprising:

one or more processors;

one or more memory modules;

one or more vehicle sensors; and

machine readable instructions stored in the one or more memory modules that, when executed by the one or more processors, cause the vehicle system to:

input an image of a road into a scene segmentation neural network, the scene segmentation neural network is trained to output an edge map of the road based on the image of the road, the edge map indicates edges of objects and boundaries between different objects in the input image;

input the image, a map of the road comprising a road geometry of the road, and the edge map into a trained regressor neural network;

determine an estimated snow depth for each of one or more lanes of the road based on an output of the regressor neural network;

determine an optimal lane of the road for the vehicle to drive in based on the estimated snow depth for the one or more lanes of the road; and

operate the vehicle autonomously to follow the determined optimal lane.

10. The vehicle system of claim 9 , wherein the machine readable instructions, when executed, further cause the vehicle system to:

receive an aggregate snow depth for the one or more lanes of the road from an edge computing device;

determine a difference between the estimated snow depth for the one or more lanes of the road and the aggregate snow depth for the one or more lanes of the road;

determine whether the difference is greater than a threshold value; and

upon determination that the difference is greater than the threshold value, transmit the estimated snow depth to the edge computing device.

11. The vehicle system of claim 10 , wherein the machine readable instructions, when executed, further cause the vehicle system to:

display the aggregate snow depth for the one or more lanes of the road such that a simulated level of the aggregate snow depth for the one or more lanes is visible to one or more occupants of the vehicle.

12. The vehicle system of claim 10 , wherein the machine readable instructions, when executed, further cause the vehicle system to:

display an indication of the optimal lane such that the indication is visible to one or more occupants of the vehicle.