Urban geometry estimation from laser measurements

1. A method of generating a three-dimensional scene geometry, comprising:

obtaining, by one or more computing devices, a first plurality of images and corresponding distance measurements for a first vehicle trajectory;

obtaining, by one or more computing devices, a second plurality of images and corresponding distance measurements for a second vehicle trajectory, the second vehicle trajectory intersecting the first vehicle trajectory;

registering, by one or more computing devices, a relative vehicle position and orientation for one or more segments of each of a first vehicle trajectory and a second vehicle trajectory, wherein registering comprises snapping vehicle trajectories to read segments;

generating, by one or more computing devices, a three-dimensional scene geometry for each vehicle trajectory;

mapping, by one or more computing devices, the three-dimensional scene geometry for each vehicle trajectory onto a common reference system based on the registering; and

merging, by one or more computing devices, the individual scene geometries to generate a complete three-dimensional scene geometry, wherein merging comprises using a route repository to determine which roads intersect and therefore which vehicle trajectories to merge.

2. The method of claim 1 , wherein merging the 3D scene geometries comprises filtering and trimming the mapped 3D scene geometries to fit the one or more plausible building structures.

3. The method of claim 1 , wherein the distance measurements are laser rangefinder distance measurements.

4. The method of claim 1 , wherein registering uses the corresponding distance measurements for the first and second vehicle trajectories.

5. The method of claim 1 , wherein the distance measurements corresponding to each vehicle trajectory include measurements to objects parallel to the vehicle trajectory and measurements to objects orthogonal to the vehicle trajectory.

6. The method of claim 1 , wherein the objects parallel to the vehicle trajectory are located to the side of a vehicle and the objects orthogonal to the vehicle trajectory are located in front of or behind the vehicle.

7. The method of claim 1 , wherein the first and second vehicle trajectories are trajectories of the same vehicle.

8. The method of claim 1 , wherein the first vehicle trajectory is a trajectory of a first vehicle and the second vehicle trajectory is a trajectory of a second vehicle.

9. A non-transitory computer readable storage medium having instructions stored therein that, when executed by a processing device, cause the processing device to execute a method for generating a three-dimensional scene geometry, the method comprising:

obtaining a first plurality of images and corresponding distance measurements for a first vehicle trajectory;

obtaining a second plurality of images and corresponding distance measurements for a second vehicle trajectory, the second vehicle trajectory intersecting the first vehicle trajectory;

registering a relative vehicle position and orientation for one or more segments of each of a first vehicle trajectory and a second vehicle trajectory, wherein registering comprises snapping vehicle trajectories to road segments;

generating a three-dimensional scene geometry for each vehicle trajectory;

mapping the three-dimensional scene geometry for each vehicle trajectory onto a common reference system based on the registering; and

merging the individual scene geometries to generate a complete three-dimensional scene geometry, wherein merging comprises using a route repository to determine which roads intersect and therefore which vehicle trajectories to merge.

10. The computer readable storage medium of claim 9 , wherein the instructions for merging the 3D scene geometries further comprise instructions for filtering and trimming the mapped 3D scene geometries to fit the one or more plausible building structures.