Systems and methods for capturing images and annotating the captured images with information

1. A method for training a classifier of a mobile robot, the method comprising:

obtaining a plurality of image frames along a drive direction of the mobile robot,

the plurality of image frames comprising a base image frame corresponding to an initial pose of the mobile robot and subsequent image frames obtained at intervals during forward travel of the mobile robot,

the mobile robot having a forward facing camera mounted thereon for obtaining the image frames,

the camera having a field of view including a floor in front of the mobile robot, and

the mobile robot having a memory device configured to store a learned data set of a plurality of descriptors determined by mobile robot events;

assuming that a location is traversable floor, wherein the mobile robot is configured to detect traversable floor and non-traversable non-floor with one or more sensors mounted on the mobile robot;

determining that the location is non-traversable non-floor based on a robot sensor event at the location, the robot sensor event comprising detection of a collision;

retrieving from a frame buffer an image frame obtained immediately prior to the robot sensor event;

generating a floor descriptor corresponding to characteristics of the floor at a bottom of the image frame captured by the camera immediately prior to the robot sensor event;

generating a non-floor descriptor corresponding to characteristics of the non-traversable non-floor at a top of the image frame captured by the camera immediately prior to the robot sensor event; and

storing the floor descriptor and the non-floor descriptor in the learned data set.

2. The method of claim 1 , wherein the detection of the collision comprises determining that the mobile robot has collided with an obstacle, and further includes:

receiving a bumper signal indicating the collision from a bumper sensor of the mobile robot; and

verifying that the mobile robot was traveling straight prior to the collision for at least one second.

3. The method of claim 1 , wherein the location corresponds to a patch of pixels and one or more characteristics of the patch of pixels includes color and/or texture.

4. The method of claim 1 , wherein the learned dataset is continuously updated and oldest descriptors are replaced by new descriptors.

5. The method of claim 1 , wherein the learned dataset is stored in memory and remains accessible by the mobile robot between runs.

6. The method of claim 1 , wherein the learned dataset is unpopulated at a start of each new run and the classifier trains the learned dataset with descriptors over the run of the mobile robot.

7. An autonomous mobile robot comprising:

a drive configured to maneuver the autonomous mobile robot over a floor within an operating environment;

a camera mounted on the autonomous mobile robot having a field of view including the floor adjacent the autonomous mobile robot in a drive direction of the autonomous mobile robot;

a frame buffer that stores image frames obtained by the camera while the autonomous mobile robot is driving;

a memory device configured to store a learned data set of a plurality of descriptors corresponding to pixel patches in ones of the image frames corresponding to portions of the operating environment and determined by mobile robot sensor events; and

one or more processors configured to execute a training process for a classifier of the learned data set, the training process comprising:

determining based on one or more mobile robot sensor events that the autonomous mobile robot collided with an obstacle;

retrieving a pre-collision frame from the frame buffer; and

generating a descriptor corresponding to at least part of the obstacle observed in the pre-collision frame.

8. The autonomous mobile robot of claim 7 ,

identifying a lower portion of the pre-collision frame and an upper portion of the pre-collision frame;

generating a first descriptor corresponding to the floor that the autonomous mobile robot is traveling on based on the lower portion;

generate a second descriptor corresponding to at least part of the obstacle observed in the pre-collision frame based on the upper portion; and

storing the first descriptor and the second descriptor in the learned data set.

9. An autonomous mobile robot comprising:

a drive configured to maneuver the autonomous mobile robot over a floor within an operating environment;

a camera mounted on the autonomous mobile robot having a field of view including the floor adjacent the autonomous mobile robot in a drive direction of the autonomous mobile robot;

a frame buffer that stores image frames obtained by the camera while the autonomous mobile robot is driving; and

a memory device configured to store a learned data set of a plurality of descriptors corresponding to pixel patches in ones of the image frames corresponding to portions of the operating environment and determined by sensor events;

one or more processors executing a training process for a classifier of the learned data set, the training process comprising:

assuming that a portion of a base image frame is traversable floor, wherein the autonomous mobile robot is configured to detect traversable floor and non-traversable non-floor with one or more sensors mounted on the autonomous mobile robot;

determining whether the autonomous mobile robot has traversed a threshold distance in a same direction since obtaining the base image frame;

identifying an upper portion and a lower portion of the base image frame;

identifying a section of the lower portion of the base image frame corresponding to a current pose of the autonomous mobile robot, the section being an area of the floor depicted in the lower portion of the base image frame at a depth corresponding to a drive distance traversed by the mobile robot from an initial pose to the current pose;

generating a floor descriptor of the section;

generating a non-floor descriptor corresponding to characteristics of the non-traversable non-floor within the upper portion of the base image frame responsive to detecting a sensor event comprising a collision; and

storing the floor descriptor and the non-floor descriptor in the learned data set.

10. The autonomous mobile robot of claim 9 , wherein the training process further comprises not detecting a sensor event of a collision with an obstacle while traversing the threshold distance.

11. The autonomous mobile robot of claim 9 , wherein the training process further comprises:

detecting a new heading of the autonomous mobile robot; and

obtaining a plurality of image frames along the new heading of the autonomous mobile robot, the plurality of image frames comprising a new base image frame corresponding to a new initial pose of the autonomous mobile robot at the new heading.

12. The autonomous mobile robot of claim 9 , wherein determining whether the autonomous mobile robot has traversed the threshold distance in the same direction since obtaining the base image frame without detecting a sensor event comprises determining a distance between the initial pose of the mobile robot and the current pose of the mobile robot.

13. The autonomous mobile robot of claim 9 , wherein identifying the upper portion and the lower portion of the base image frame comprises identifying a horizon within the base image frame.

14. A method for training a classifier of a mobile robot, the method comprising:

obtaining a plurality of image frames along a drive direction of the mobile robot,

the plurality of image frames comprising a base image frame corresponding to an initial pose of the mobile robot and subsequent image frames obtained at intervals during forward travel of the mobile robot,

the mobile robot having a forward facing camera mounted thereon for obtaining the image frames,

the camera having a field of view including a floor in front of the mobile robot, and

the mobile robot having a memory device configured to store a learned data set of a plurality of descriptors determined by robot sensor events;

tracking a location as non-traversable non-floor based on the plurality of descriptors, wherein the mobile robot is configured to detect traversable floor and non-traversable non-floor with one or more sensors mounted on the mobile robot;

determining that the location is traversable floor after traveling a distance to the location and not detecting a robot sensor event at the location;

generating a floor descriptor corresponding to characteristics of the traversable floor at the location within one of the image frames captured by the camera;

determining that a new location is non-traversable non-floor based on detecting a robot sensor event comprising a collision at the new location;

generating a non-floor descriptor corresponding to characteristics of the non-traversable non-floor at the new location within one of the image frames captured by the camera; and

updating the descriptors in the learned data set.

15. The method of claim 14 , wherein the distance traveled is based on an expected change in a wheel encoder.

16. The method of claim 14 , further comprising detecting a heading change and resetting a tracking of the location to the new location that corresponds to the non-traversable non-floor.

17. The method of claim 14 , wherein the location corresponds to a patch of pixels and one or more characteristics of the patch of pixels includes at least one characteristic selected from the group consisting of: color; and texture.

18. The method of claim 14 , wherein not detecting the robot sensor event comprises determining that the mobile robot has not collided with an obstacle.

19. An autonomous mobile robot comprising:

a drive configured to maneuver the autonomous mobile robot over a floor within an operating environment;

a camera mounted on the mobile robot having a field of view including the floor adjacent the autonomous mobile robot in a drive direction of the autonomous mobile robot;

a frame buffer that stores image frames obtained by the camera while the autonomous mobile robot is driving; and

a memory device configured to store a learned data set of a plurality of descriptors corresponding to pixel patches in ones of the image frames corresponding to portions of the operating environment and determined by robot sensor events;

one or more processors executing a training process for a classifier of the learned data set, the training process comprising:

tracking a location as non-traversable non-floor based on the plurality of descriptors, wherein the autonomous mobile robot is configured to detect traversable floor and non-traversable non-floor with one or more sensors mounted on the mobile robot;

determining that the location is traversable floor after traveling a distance to the location and not detecting a robot sensor event at the location;

generating a floor descriptor corresponding to characteristics of the traversable floor at the location within one of the image frames captured by the camera;

determining that a new location is non-traversable non-floor based on detecting a robot sensor event comprising a collision at the new location;

generating a non-floor descriptor corresponding to characteristics of the non-traversable non-floor at the new location within one of the image frames captured by the camera; and

updating the descriptors in the learned data set.

20. The autonomous mobile robot of claim 19 , wherein the distance traveled is based on an expected change in a wheel encoder.

21. The autonomous mobile robot of claim 19 , wherein the training process further comprises detecting a heading change and resetting a tracking of the location to a new location that corresponds to non-traversable non-floor.

22. The autonomous mobile robot of claim 19 , wherein the location corresponds to a patch of pixels and one or more characteristics of the patch of pixels includes at least one characteristic selected from the group consisting of: color; and texture.

23. The autonomous mobile robot of claim 19 , wherein not detecting the sensor event comprises determining that the autonomous mobile robot has not collided with an obstacle.

24. An autonomous mobile robot comprising:

a drive configured to maneuver the autonomous mobile robot over a floor within an operating environment;

a camera mounted on the mobile robot having a field of view including the floor adjacent the autonomous mobile robot in a drive direction of the autonomous mobile robot;

a frame buffer that stores image frames obtained by the camera while the autonomous mobile robot is driving; and

a memory device configured to store a learned data set of a plurality of descriptors corresponding to pixel patches in one of the image frames corresponding to portions of the operating environment and determined by mobile robot sensor events;

one or more processors executing a training process for a classifier of the learned data set, the training process comprising:

detecting a plurality of obstacles located at a plurality of distances from the autonomous mobile robot;

tracking a first obstacle that is closest to the autonomous mobile robot and buffering descriptors of the first obstacle in the learned data set;

traveling a threshold distance and detecting a second obstacle that is closer to the autonomous mobile robot than the first obstacle;

resetting the autonomous mobile robot to not track the first obstacle; and

tracking the second obstacle and buffering descriptors of the second obstacle in the learned data set, wherein the descriptors of the first obstacle and/or the second obstacle are determined by at least one mobile robot sensor event comprising a collision.

25. The autonomous mobile robot of claim 24 , wherein an obstacle corresponds to a patch of pixels and one or more characteristics of the patch of pixels includes at least one characteristic selected from the group consisting of: color; and texture.

26. The autonomous mobile robot of claim 24 , wherein tracking the first obstacle and/or the second obstacle comprises tracking a movement of a patch of pixels through a plurality of the image frames based on movement of the autonomous mobile robot.

27. The autonomous mobile robot of claim 24 , wherein the training process further comprises traveling a threshold distance prior to capturing a new image.

28. The autonomous mobile robot of claim 24 , wherein the training process further comprises buffering descriptors of a closest obstacle and updating the learned dataset with the descriptors.