METHOD FOR NAVIGATING A MOVABLE DEVICE ALONG AN INCLINED SURFACE

Abstract

A method for navigating a movable device is disclosed. The movable device comprises a moving system arranged to move the movable device along a surface, a control system arranged to control movement of the movable device by means of the moving system, and a vision-based detector. At least one fiducial mounted at or near a position of the movable device, is captured by means of the vision-based detector and recognised. A position and/or a pose of the movable device is determined, based on the recognised at least one fiducial. An inclination of the movable device is detected, and the determined position and/or pose of the movable device is corrected, based on the detected inclination of the movable device. Finally, the movable device is navigated based on the corrected position and/or pose of the movable device.

Claims

1. A method for navigating a movable device, the movable device comprising a moving system arranged to move the movable device along a surface, a control system arranged to control movement of the movable device by means of the moving system, and a vision-based detector, the method comprising the steps of: capturing at least one visually recognisable fiducial mounted at or near a position of the movable device, by means of the vision-based detector, and recognising the at least one fiducial, determining a position and/or a pose of the movable device, based on the recognised at least one fiducial and based on a known position and/or orientation of the recognised at least one fiducial, detecting an inclination of the movable device, in the case that a non-zero inclination of the movable device is detected, determining a corrected position and/or pose of the movable device, based on the determined position and/or pose of the movable device and on the detected inclination of the movable device, and navigating the movable device based on the corrected position and/or pose of the movable device.

2. The method according to claim 1, wherein the step of capturing at least one fiducial comprises capturing at least one fiducial mounted on a ceiling at or near a position of the movable device.

3. The method according to claim 1, wherein the vision-based detector is oriented in a direction which covers part of a ceiling at or near a position of the movable device.

4. The method according to claim 3, wherein the vision-based detector is oriented in a direction which covers a direction being substantially normal to the surface along which the movable device moves.

5. The method according to claim 1, wherein the movable device further comprises an inertial measurement unit, and wherein the step of detecting an inclination of the movable device is performed by means of the inertial measurement unit.

6. The method according to claim 1, wherein the step of detecting an inclination of the movable device is performed by means of detections obtained by the vision-based detector.

7. The method according to claim 1, wherein the step of capturing at least one fiducial comprises the vision-based detector capturing a machine readable code forming part of the at least one fiducial.

8. The method according to claim 1, wherein the step of capturing at least one fiducial comprises capturing at least two fiducials, and wherein the step of determining a position and/or a pose of the movable device is based on a recognition of a combination of the captured fiducials.

9. The method according to claim 1, wherein the step of detecting an inclination of the movable device comprises determining an angle of inclination of the movable device, and wherein the step of determining a corrected position and/or pose of the movable device comprises compensating the determined position and/or or pose in accordance with the determined angle of inclination.

10. The method according to claim 1, further comprising the step of obtaining an image of one or more features arranged at or near the position of the movable device, by means of the vision-based detector, and wherein the step of determining a position and/or a pose of the movable device is further based on the obtained image.

11. The method according to claim 1, wherein the step of capturing at least one fiducial comprises capturing at least one fiducial mounted on a wall at or near a position of the movable device.

12. A movable device comprising a moving system arranged to move the movable device along a surface, a control system arranged to control movement of the movable device by means of the moving system, and a vision-based detector, wherein the control system is adapted to navigate the movable device in accordance with a method comprising the steps of: capturing at least one visually recognisable fiducial mounted at or near a position of the movable device, by means of the vision-based detector, and recognising the at least one fiducial, determining a position and/or a pose of the movable device, based on the recognised at least one fiducial and on a known position and/or orientation of the recognised at least one fiducial, detecting an inclination of the movable device, in the case that a non-zero inclination of the movable device is detected, determining a corrected position and/or pose of the movable device, based on the determined position and/or pose of the movable device and on the detected inclination of the movable device, and navigating the movable device based on the corrected position and/or pose of the movable device.

13. The movable device according to claim 12, further comprising an inertial measurement unit configured to detect an inclination of the movable device.

14. The movable device according to claim 11, wherein the vision-based detector is a camera.

15. The movable device according to claims 12, wherein at least the vision-based detector is arranged in a nugget mounted on the movable device.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The invention will now be described in further detail with reference to the accompanying drawings in which:

[0060] FIG. 1 is a schematic illustration of a movable device according to a first embodiment of the invention,

[0061] FIG. 2 is a schematic illustration of a movable device according to a second embodiment of the invention,

[0062] FIG. 3 shows fiducials for use in a method according to an embodiment of the invention,

[0063] FIG. 4 shows the movable device of FIG. 1 capturing a number of fiducials,

[0064] FIG. 5 shows the movable device of FIG. 1 on a horizontal surface and on a sloped surface,

[0065] FIGS. 6 and 7 illustrate the movable device of FIG. 1 capturing a fiducial from a horizontal surface and a sloped surface, respectively, and

[0066] FIG. 8 is a flow chart illustrating a method according to an embodiment of the invention.

DETAILED DESCRIPTION

[0067] FIG. 1 is a schematic illustration of a movable device 1 according to a first embodiment of the invention. The movable device 1 comprises a moving system in the form of wheels 2, and the movements of the movable device 1 are controlled by means of a control system (not shown). The movable device 1 further comprises a vision-based detector 3 mounted on an upper part of the movable device 1. The vision-based detector 3 may, e.g., be in the form of a camera.

[0068] The vision-based detector 3 defines a detection field, illustrated by lines 4, spanning an angle. The detection field covers a direction pointing directly vertically upwards, i.e. normal to a horizontal surface along which the movable device 1 moves. Thereby the vision-based detector 3 is capable of capturing features arranged above the movable device 1, e.g. features mounted on or forming part of a ceiling. The vision-based detector 3 can be used for determining a position and/or a pose of the movable device 1, in a manner which will be described in further detail below.

[0069] FIG. 2 is a schematic illustration of a movable device 1 according to a second embodiment of the invention. The movable device 1 of FIG. 2 is very similar to the movable device of FIG. 1, and it will therefore not be described in detail here. However, in the movable device 1 of FIG. 2, the vision-based detector 3 is mounted on an inclined surface 5 of the movable device 1. Thereby the detection field 4 of the vision-based detector 3 is directed more in a forwards direction than is the case for the movable device 1 of FIG. 1. This allows the vision-based detector 3 to also capture features arranged in front of the movable device 1 along a direction of movement. Thereby the vision-based detector 3 can be used for further navigation tasks, such as proximity determination. However, the detection field 4 still covers the direction being substantially normal to the horizontal surface along which the movable device 1 moves.

[0070] It should be noted that even though FIG. 2 shows that the vision-based detector 3 is directed in such a manner that it is allowed to capture features arranged in front of the movable device 1, it is within the scope of protection of the present invention that the inclined surface 5 could be arranged on the movable device 1 in a manner which allows the vision-based detector 3 to captures features behind the movable device 1, i.e. in a direction opposite to the direction of movement of the movable device 1, and/or sideways relative to the movable device 1, i.e. in a direction which is transverse to the direction of movement of the movable device 1. It is also within the scope of protection of the present invention that the vision-based detector 3 is mounted on the movable device 1 in a manner which does not allow the vision-based detector 3 to capture features arranged directly above the movable device 1.

[0071] FIG. 3 shows fiducials 6 for use in a method according to an embodiment of the invention. The fiducials 6 are in the form of two-dimensional machine readable codes, allowing them to be recognised when captured by a suitable vision-based detector, such as a camera. In FIG. 3 ten fiducials 6 are shown, each being different from any of the other fiducials 6. Thereby the fiducials 6 of FIG. 3 can all be uniquely identified. Accordingly, if the fiducials 6 of FIG. 3 are mounted at specific positions within a site where a movable device is supposed to move, the movable device will be able to determine its position and/or pose by capturing and recognising a given fiducial.

[0072] The machine readable codes of the fiducials 6 may further define commands for a movable device. In this case, when a vision-based detector of a movable device captures a given fiducial, this will cause the movable device to perform the commands defined by the machine readable code. Such command may, e.g., include ‘stop’, ‘slow down’, ‘speed up’, ‘turn left’, ‘turn right’, ‘generate audible signal’, ‘generate visual signal’, etc.

[0073] The machine readable codes of the fiducials 6 of FIG. 3 provide an appearance to each fiducial 6 which is not rotational symmetric. Thereby the fiducials 6 will not appear in the same manner when they are viewed from different angles. This further adds to the possibility of uniquely determining a position of a movable device which captures one of the fiducials 6.

[0074] FIG. 4 shows the movable device 1 of FIG. 1 and four fiducials 6 mounted on a ceiling above the movable device 1. The fiducials 6 could, e.g., be of the kind illustrated in FIG. 3. It can be seen that the detection field 4 of the vision-based detector 3 covers a direction, illustrated by dashed line 7, which is normal to a horizontal surface, along which the movable device 1 moves. Accordingly, the vision-based detector 3 is capable of capturing the fiducials 6 mounted on the ceiling above the movable device 1 as the movable device 1 moves along the horizontal surface.

[0075] It can be seen that, in the position of the movable device 1 illustrated in FIG. 4, the vision-based detector 3 captures two neighbouring fiducials 6. The fiducials 6 are recognisable, and thereby the movable device 1 is capable of determining its position, and possibly its pose, based on the recognition of the captured fiducials 6. To this end it is assumed that the movable device 1 is not inclining, that the vision-based detector 6 is thereby arranged substantially horizontally, and the line 7 is therefore arranged substantially vertically. However, if the movable device 1 is, for some reason, inclined, e.g. due to the movable device 1 moving along a sloped surface, or due to an object being stuck on one of the wheels 2, this assumption is not true, and therefore the position of the movable device 1 is determined incorrectly. Accordingly, when this is the case it is necessary to correct the determined position in order to take the inclination of the movable device 1 into account.

[0076] FIG. 5 shows the movable device 1 of FIG. 1 moving along a horizontal surface 8, as described above with reference to FIG. 4, and along an inclined surface 9, respectively. It can be seen that when the movable device 1 moves along the inclined surface 9, the detection field 4 of the vision-based detector 3 is tilted as compared to the situation where the movable device 1 moves along the horizontal surface 8. This is illustrated by the line 7, which indicates the direction which is normal to the surface 9 along which the movable device 1 moves, defining an angle with respect to vertical.

[0077] If the inclination of the movable device 1 is not taken into account when determining the position and/or pose of the movable device 1, based on at least one recognised fiducial 6, the determination of the position will be incorrect. More particularly, in the example illustrated in FIG. 5, the uncorrected determined position of the movable device 1 would be further to the right than the actual position of the movable device 1. The larger the inclination, the larger the deviation between the determined position and the actual position of the movable device 1.

[0078] In order to improve the accuracy of the determination of the position and/or the pose of the movable device 1, the inclination of the movable device 1 is detected. This could, e.g., be done by means of an inertial measurement unit (IMU) arranged in or mounted on the movable device 1. As an alternative, the inclination may, e.g., be detected by means of the vision-based detector 3, e.g. in the manner described below with reference to FIGS. 6 and 7. In any event, an angle of inclination is preferably determined.

[0079] Next, the determined position and/or pose of the movable device 1 is corrected, based on the detected inclination. In the example illustrated in FIG. 5, the determined position should be moved towards the left by a distance, Δx=D.Math.sin(α), where D is the distance between the vision-based detector 3 and the ceiling, along the line 7, and α is the angle of inclination.

[0080] When the determined position of the movable device 1 is corrected in this manner, a more accurate position determination is obtained, and thereby it is possible to navigate the movable device 1 in a more precise manner. Since the correction is based on a detection of the inclination of the movable device 1, the accurate position determination can be obtained easily, and regardless of the direction or magnitude of the inclination of the movable device 1.

[0081] FIGS. 6 and 7 illustrate detection of an inclination of the movable device 1 of FIG. 1 by means of the vision-based detector 3. In FIG. 6, the movable device 1 moves along a horizontal surface 8 and captures a fiducial 6 mounted immediately above the vision-based detector 3. Since the direction 7 being normal to the surface 8 along which the movable device 1 moves is substantially vertical, the vision-based detector 3 captures an undistorted image of the fiducial 6, as illustrated in the upper part of FIG. 6.

[0082] In FIG. 7, the movable device 1 moves along an inclined surface 9, and therefore the direction 7 being normal to the surface 9 along which the movable device 1 moves forms an angle with respect to vertical. Therefore the vision-based detector 3 captures a distorted image of the fiducial 6, caused by the inclination, as illustrated in the upper part of FIG. 7. The angle of inclination can be derived from an analysis of the distortion of the captured image. Accordingly, the inclination of the movable device 1 can be detected by means of the vision-based detector 3.

[0083] FIG. 8 is a flow chart illustrating a method according to an embodiment of the invention. The process is started at step 10. At step 11 at least one fiducial is captured by means of a vision-based detector mounted on or forming part of a movable device, and the fiducial(s) is/are recognised.

[0084] At step 12 a position and/or a pose of the movable device is determined, based on the recognised fiducial(s).

[0085] At step 13 it is investigated whether or not the movable device is inclined. If this is not the case, it can be concluded that the determined position and/or pose is correct and reliable, and therefore the process is forwarded to step 14, where the movable device is navigated based on the determined position and/or pose.

[0086] In the case that step 13 reveals that the movable device is inclined, the process is forwarded to step 15, where an angle of inclination of the movable device is determined. At step 16, a corrected position and/or pose is determined, i.e. the position and/or pose of the movable device, which was determined at step 12, is corrected on the basis of the determined angle of inclination. Thereby the inclination of the movable device is taken into account in the determination of the position and/or pose of the movable device, and accordingly a more precise determination of the position and/or pose is obtained.

[0087] Finally, the process is forwarded to step 14, where the movable device is navigated based on the corrected determination of the position and/or pose of the movable device.