System and method for drift-free global trajectory estimation of a mobile platform

1. A system for trajectory estimation of a mobile platform, the system comprising:

one or more processors and a memory, the memory being a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a straight line, the corrected position for each point is determined based on a geometric distance, the geometric distance being a distance from each point to a starting point of the latest trajectory segment; and

connecting the corrected positions to form the corrected trajectory.

2. The system as set forth in claim 1 , further comprising an operation of causing a mobile platform to traverse a path that coincides with the drift-free trajectory.

3. The system as set forth in claim 1 , wherein if the location is determined to be a new location, storing coordinates for the new location in a trajectory buffer and designating the initial trajectory estimate as the drift-free trajectory.

4. The system as set forth in claim 1 , wherein if trajectory correction is not required, then designating the initial trajectory estimate as the drift-free trajectory.

5. The system as set forth in claim 1 , wherein generating the initial trajectory estimate for the mobile platform is performed using a simultaneous localization and mapping (SLAM) module.

6. The system as set forth in claim 1 , wherein the sensor data is generated using a sensor system selected from a group consisting of only a monocular camera, only a stereo camera, or a combination of a monocular camera with a depth image sensor.

7. The system as set forth in claim 1 , wherein in determining if trajectory correction is required, trajectory correction is required if a current position of the mobile platform at the location is greater than a predetermined distance threshold from a previously recorded position at the location.

8. A system for trajectory estimation of a mobile platform, the system comprising:

one or more processors and a memory, the memory being a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a curved line, the corrected position for each point is determined based on a curve distance, the curve distance being a distance from each point to a starting point of the latest trajectory segment along the curved line; and

connecting the corrected positions to form the corrected trajectory.

9. A computer program product for trajectory estimation of a mobile platform, the computer program product comprising:

a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a straight line, the corrected position for each point is determined based on a geometric distance, the geometric distance being a distance from each point to a starting point of the latest trajectory segment; and

connecting the corrected positions to form the corrected trajectory.

10. The computer program product as set forth in claim 9 , further comprising instructions for causing a mobile platform to traverse a path that coincides with the drift-free trajectory.

11. The computer program product as set forth in claim 9 , wherein if the location is determined to be a new location, storing coordinates for the new location in a trajectory buffer and designating the initial trajectory estimate as the drift-free trajectory.

12. The computer program product as set forth in claim 9 , wherein if trajectory correction is not required, then designating the initial trajectory estimate as the drift-free trajectory.

13. The computer program product as set forth in claim 9 , wherein generating the initial trajectory estimate for the mobile platform is performed using a simultaneous localization and mapping (SLAM) module.

14. The computer program product as set forth in claim 9 , wherein the sensor data is generated using a sensor system selected from a group consisting of only a monocular camera, only a stereo camera, or a combination of a monocular camera with a depth image sensor.

15. The computer program product as set forth in claim 9 , wherein in determining if trajectory correction is required, trajectory correction is required if a current position of the mobile platform at the location is greater than a predetermined distance threshold from a previously recorded position at the location.

16. A computer program product for trajectory estimation of a mobile platform, the computer program product comprising:

a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a curved line, the corrected position for each point is determined based on a curve distance, the curve distance being a distance from each point to a starting point of the latest trajectory segment along the curved line; and

connecting the corrected positions to form the corrected trajectory.

17. A computer implemented method for trajectory estimation of a mobile platform, the method comprising an act of:

executing instructions encoded on a non-transitory computer-readable medium, such that upon execution, one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a straight line, the corrected position for each point is determined based on a geometric distance, the geometric distance being a distance from each point to a starting point of the latest trajectory segment; and

connecting the corrected positions to form the corrected trajectory.

18. The method as set forth in claim 17 , further comprising an operation of causing a mobile platform to traverse a path that coincides with the drift-free trajectory.

19. The method as set forth in claim 17 , wherein if the location is determined to be a new location, storing coordinates for the new location in a trajectory buffer and designating the initial trajectory estimate as the drift-free trajectory.

20. The method as set forth in claim 17 , wherein if trajectory correction is not required, then designating the initial trajectory estimate as the drift-free trajectory.

21. The method as set forth in claim 17 , wherein generating the initial trajectory estimate for the mobile platform is performed using a simultaneous localization and mapping (SLAM) module.

22. The method as set forth in claim 17 , wherein the sensor data is generated using a sensor system selected from a group consisting of only a monocular camera, only a stereo camera, or a combination of a monocular camera with a depth image sensor.

23. The method as set forth in claim 17 , wherein in determining if trajectory correction is required, trajectory correction is required if a current position of the mobile platform at the location is greater than a predetermined distance threshold from a previously recorded position at the location.

24. A computer implemented method for trajectory estimation of a mobile platform, the method comprising an act of:

executing instructions encoded on a non-transitory computer-readable medium, such that upon execution, one or more processors perform operations of:

generating an initial trajectory estimate for the mobile platform based on sensor data regarding a scene;

determining if a location is a new location or a revisited location using real-time feature point detection;

if the location is determined to be a revisited location, determining if trajectory correction is required; and

if trajectory correction is required, generating a corrected trajectory as a drift-free trajectory, the corrected trajectory being generated based on a previously recorded position at the location;

wherein in generating the corrected trajectory, the corrected trajectory is generated by performing operations of:

sampling equally-spaced points from a latest trajectory segment in a buffered trajectory;

determining corrected positions for each point in the latest trajectory segment, such that when the latest trajectory segment is a curved line, the corrected position for each point is determined based on a curve distance, the curve distance being a distance from each point to a starting point of the latest trajectory segment along the curved line; and

connecting the corrected positions to form the corrected trajectory.