Determining travel path features based on retroreflectivity

1. A method for determining geographic features corresponding to a road network to develop a navigation database, the method comprising:

collecting data points via a light detection and ranging (LIDAR) device while driving on a road of the road network;

determining whether energy of the collected data points is substantially constant through a scan of the LIDAR device;

filtering the collected data points based on a retroreflectivity threshold and whether the energy of the collected data points is substantially constant;

identifying data points that correspond to a road feature based on the filtered data points, wherein the road feature is treated with a retroreflective substance;

calculating a normal vector relative to a plane including the road feature that fits a portion of the identified data points; and

calculating an angle between the normal vector and a driving direction.

2. The method of claim 1 , wherein the road feature comprises a road sign.

3. The method of claim 1 , wherein identifying the data points comprises:

eliminating outlier data points based on a number of neighboring data points within a specific radius.

4. The method of claim 1 , wherein identifying the data points comprises:

identifying a cluster of the filtered data points; and

applying a growing algorithm to identify additional data points corresponding to the road feature.

5. The method of claim 1 , further comprising:

generating a data model representing the road feature based on the identified data points; and

storing the data model in the navigation database.

6. The method of claim 5 , wherein generating the data model comprises:

fitting a geometric shape to a boundary associated with the identified data points.

7. The method of claim 5 , wherein generating the data model comprises:

matching the identified data points with collected video or camera data.

8. The method of claim 1 , wherein identifying the data points comprises:

applying a grid to a portion of the filtered data points to determine whether the portion of the filtered data points correspond to separate road features.

9. A method for determining features on a travel path to develop a map database, the method comprising:

collecting data points via a laser device while moving on the travel path;

extracting data points representing a path surface of the travel path from the collected data points;

determining whether energy of the data points is substantially constant through a scan of the laser device;

filtering the extracted data points based on a return intensity threshold and whether the energy of the data points is substantially constant, wherein values above the return intensity threshold indicate retroreflective substances;

calculating a normal vector relative to a plane of a road marking that fits a portion of the filtered data points;

calculating an angle between the normal vector and a driving direction; and

identifying data points that correspond to the road marking on the travel path based on the portion of the filtered data points.

10. The method of claim 9 , wherein the marking comprises a road lane marking.

11. The method of claim 9 , wherein extracting the data points representing the path surface comprises:

calculating an average ground elevation of the travel path.

12. The method of claim 11 , further comprising:

determining borders of the travel path based on the average ground elevation.

13. The method of claim 9 , further comprising:

generating a data model of the marking based on the identified data points; and

storing the data model in the map database.

14. A system for determining geographic features corresponding to a path to develop a navigation database, the system comprising:

a transportation device;

a light detection and ranging (LIDAR) device supported by the transportation device, the LIDAR device configured to collect data points corresponding to the path, wherein a path feature is identified by thresholding the collected data points as a function of a scan of the LIDAR device being substantially constant; and

a processor configured to identify a plane of data points among the collected data points and calculate an angle between a direction of the transportation device and a normal vector of the plane of data points of the path feature, wherein the angle indicates a type of the path feature.

15. The system of claim 14 , wherein the transportation device comprises a vehicle.

16. The system of claim 14 , wherein the transportation device comprises a device configured to be carried by a pedestrian.

17. The system of claim 14 , wherein the path feature comprises a road sign.

18. The system of claim 14 , wherein the path feature comprises a lane marking.

19. The method of claim 1 , wherein the road feature comprises a lane marking.

20. The method of claim 1 , wherein calculating the angle between the normal vector and the driving direction comprises calculating a dot product between the normal vector and the driving direction.

21. The method of claim 20 , further comprising:

comparing the dot product to a first threshold; and

identifying the road feature as a road sign if the dot product exceeds the first threshold.

22. The method of claim 21 , further comprising:

comparing the dot product to a second threshold; and

identifying the road feature as a lane marking if the dot product is less than the second threshold.