Lane feature detection in aerial images based on road geometry

1. A method for lane feature detection from an image according to predetermined path geometry, the method comprising:

receiving an image including at least one path;

selecting map data corresponding to the at least one path, the map data defining the predetermined path geometry, wherein the predetermined path geometry is defined according to lane boundaries that are spaced by a calculated width from a centerline;

modifying the image according to the selected map data including the predetermined path geometry;

generating, using a processor, a lane feature prediction model based on the modified image; and

providing a subsequent image to the lane feature prediction model for a prediction of at least one lane feature.

2. The method of claim 1 , further comprising:

generating a ground truth mask including the predetermined path geometry, wherein the image is modified according to the ground truth mask.

3. The method of claim 2 , wherein the ground truth mask includes pixels of a first value for the predetermined path geometry and pixels of a second value for modified portions of the image.

4. The method of claim 1 , further comprising:

accessing a width value for the at least one path from the map data, the predetermined path geometry having a dimension corresponding to the width value.

5. The method of claim 4 , further comprising:

receiving probe data for the at least one path; and

analyzing the probe data to derive the width value.

6. The method of claim 4 , further comprising:

performing an image processing algorithm on the image including the at least one path, wherein the image processing algorithm includes segmentation of a road surface or filtering of pixel values for a centerline for the at least one path,

wherein the width value is derived from an output of the image processing algorithm.

7. The method of claim 1 , further comprising:

accessing the centerline for the at least one path from the map data.

8. The method of claim 7 , wherein the calculated width is based on a number of lanes included in the at least one path or a functional classification of the at least one path.

9. The method of claim 1 , further comprising:

dividing the modified image into a plurality of training images according to a patch size, wherein each of the plurality of training images includes at least a portion of the at least one path.

10. The method of claim 9 , further comprising:

rotating at least one of the plurality of training images to a predetermined angle.

11. The method of claim 9 , wherein the plurality of training images are spaced apart in a direction of the at least one path.

12. The method of claim 1 , further comprising:

identifying a scaling factor for a relationship between pixel size in the image and geographic distance in the map data.

13. The method of claim 1 , further comprising:

receiving a navigation request; and

providing a navigation message including the prediction of at least one lane feature in response to the navigation request.

14. An apparatus for lane feature detection from an image according to predetermined path geometry, the apparatus comprising:

a controller configured to identify map data corresponding to at least one path, the map data defining the predetermined path geometry, wherein the predetermined path geometry is defined according to lane boundaries that are spaced by a calculated width from a centerline;

an aerial image editor configured to modify the image according to the map data including the predetermined path geometry; and

a lane feature model trained according to the modified image and configured to identify at least one lane feature from a subsequent image.

15. The apparatus of claim 14 , further comprising:

an image trainer configured to identify a plurality of training patch images from the modified image.

16. The apparatus of claim 15 , wherein the image trainer is configured rotate at least one of the plurality of training patch image to a predetermined angle.

17. The apparatus of claim 14 , wherein the controller identifies a scaling factor to relate the map data and the image, wherein the predetermined path geometry is sized according to the scaling factor.

18. The apparatus of claim 14 , wherein the aerial image editor modifies the image according to a ground truth mask.

19. The apparatus of claim 18 , wherein the ground truth mask includes a first mask value for portions inside the predetermined path geometry and a second mask for portions outside the predetermined path geometry.

20. A non-transitory computer readable medium including instructions that when executed by a process are configured to perform:

receiving an aerial image for a geographic area;

receiving road network data for the geographic area, the road network data defining a predetermined path geometry, wherein the predetermined path geometry is defined according to lane boundaries that are spaced by a calculated width from a centerline;

modifying the aerial image according to the road network data; and

providing the modified aerial image to a lane feature detection model.