Techniques for image recognition-based aerial vehicle navigation

1. A control terminal for controlling an unmanned aerial vehicle (UAV), comprising:

a processor; and

a storage medium storing instructions that, when executed by the processor, cause the processor to:

render an image on a user interface of the control terminal, the image being captured by an imaging device coupled to the UAV and being associated with a view of the imaging device;

detect, via the user interface, a gesture-based input including one or more reference points in the image, the gesture-based input indicating a view change of the imaging device;

determine a type of the gesture-based input by analyzing the one or more reference points; and

generate control data based on the type of the gesture-based input to control at least one of a parameter of the UAV or a parameter of the imaging device, for the view change of the imaging device, the parameter of the UAV including at least one of a velocity of the UAV or an acceleration of the UAV, and the parameter of the imaging device including at least one of a velocity of the imaging device or an acceleration of the imaging device.

2. The control terminal of claim 1 , wherein the one or more reference points comprise a single reference point, or multiple reference points including at least an initial point and a release point.

3. The control terminal of claim 2 , wherein:

the single reference point is associated with the view of the imaging device; or

the initial point is associated with the view of the imaging device and the release point is associated with a subsequent view of the imaging device.

4. The control terminal of claim 2 , wherein determining the type of the gesture-based input by analyzing the one or more reference points comprises:

determining a total number of the one or more reference points; or

determining coordinates of the one or more reference points in the image.

5. The control terminal of claim 1 , wherein the type of the gesture-based input comprises a tap gesture, a swipe gesture, a pinch gesture, or a zoom gesture.

6. The control terminal of claim 5 , wherein:

the type of the gesture-based input is the tap gesture when the one or more reference points comprise a single reference point;

the type of the gesture-based input is the swipe gesture when the one or more reference points comprise multiple reference points including at least one initial point and at least one release point; or

the type of the gesture-based input is the pinch gesture or the zoom gesture when the one or more reference points comprise multiple reference points including at least two initial points and at least two release points.

7. The control terminal of claim 1 , wherein generating the control data comprises:

determining a magnitude or direction of the gesture-based input based on coordinates of the one or more reference points in the image; and

generating the control data based on the magnitude or direction of the gesture-based input to control the parameter of the UAV or the parameter of the imaging device, the parameter of the UAV further including at least one of a position, an orientation, or an attitude of the UAV, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude, a zoom factor, a focus length, or an operational parameter of the imaging device.

8. The control terminal of claim 1 , wherein the one or more reference points comprise a single reference point and generating the control data comprises:

identifying a feature region in the image that corresponds to coordinates of the single reference point, wherein a size of the feature region is configurable; and

generating the control data indicating a distance between the feature region and a next single reference point to control the parameter of the UAV or the parameter of the imaging device, the parameter of the UAV further including at least one of a position, an orientation, or an attitude of the UAV, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude, a zoom factor, a focus length, or an operational parameter of the imaging device.

9. The control terminal of claim 8 , wherein the processor is configured to stop generating the control data in response to: (1) the feature region overlaying the next single reference point, (2) coordinates of the feature region falling within a threshold deviation of the coordinates of the next single reference point, or (3) a magnitude of a control vector being smaller than a threshold value.

10. The control terminal of claim 1 , wherein the one or more reference points comprise multiple reference points including at least an initial point and a release point, and generating the control data comprises:

identifying a feature region in the image that corresponds to coordinates of the initial point, wherein a size of the feature region is configurable; and

generating the control data indicating a distance between the feature region and the release point to control the parameter of the UAV or the parameter of the imaging device, the parameter of the UAV further including at least one of a position, an orientation, or an attitude of the UAV, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude, a zoom factor, a focus length, or an operational parameter of the imaging device.

11. The control terminal of claim 10 , wherein the processor is configured to stop generating the control data in response to: (1) the feature region overlaying the release point, (2) coordinates of the feature region falling within a threshold deviation of the coordinates of the release point, or (3) a magnitude of a control vector being smaller than a threshold value.

12. The control terminal of claim 1 , wherein the gesture-based input is generated by detecting movement of a stylus, or a user's fingers, eyes, or head.

13. The method of claim 1 , wherein the control data include a control vector that is determined based on the one or more references points, the control vector including:

an X component for controlling a yaw angle of the UAV or a yaw angle of the imaging device; and

a Y component for controlling a pitch angle of the UAV or a pitch angle of the imaging device.

14. A method for controlling a movable object, comprising:

rendering, by a processor, an image on a user interface of a control terminal, the image being captured by an imaging device coupled to the movable object and being associated with a view of the imaging device;

detecting, by the processor via the user interface, a gesture-based input including one or more reference points in the image, the gesture-based input indicating a view change of the imaging device;

determining, by the processor, a type of the gesture-based input by analyzing the one or more reference points; and

generating, by the processor, control data based on the type of the gesture-based input to control at least one of a parameter of the movable object or a parameter of the imaging device for the view change of the imaging device, the parameter of the movable object including at least one of a velocity of the movable object or an acceleration of the movable object, and the parameter of the imaging device including at least one of a velocity of the imaging device or an acceleration of the imaging device.

15. The method of claim 14 , wherein determining the type of the gesture-based input by analyzing the one or more reference points comprises:

determining a total number of the one or more reference points; or

determining coordinates of the one or more reference points in the image.

16. The method of claim 14 , wherein the one or more reference points comprise a single reference point and generating the control data comprises:

identifying a feature region in the image that corresponds to coordinates of the single reference point, wherein a size of the feature region is configurable; and

generating the control data indicating a distance between the feature region and a next single reference point to control the parameter of the movable object or the parameter of the imaging device, the parameter of the movable object further including at least one of a position, an orientation, or an attitude of the movable object, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude a zoom factor, a focus length, or an operational parameter of the imaging device.

17. The method of claim 14 , wherein the one or more reference points comprise multiple reference points including at least an initial point and a release point, and generating the control data comprises:

identifying a feature region in the image that corresponds to coordinates of the initial point, wherein a size of the feature region is configurable; and

generating the control data indicating a distance between the feature region and the release point to control the parameter of the movable object or the parameter of the imaging device, the parameter of the movable object further including at least one of a position, an orientation, or an attitude of the movable object, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude a zoom factor, a focus length, or an operational parameter of the imaging device.

18. A non-transitory computer-readable medium including instructions stored thereon which, when executed by a processor, cause the processor to:

render an image on a user interface of a control terminal, the image being captured by an imaging device coupled to a movable object and being associated with a view of the imaging device;

detect, via the user interface, a gesture-based input including one or more reference points in the image, the gesture-based input indicating a view change of the imaging device;

determine a type of the gesture-based input by analyzing the one or more reference points; and

generate control data based on the type of the gesture-based input to control at least one of a parameter of the movable object or a parameter of the imaging device, for the view change of the imaging device, the parameter of the movable object including at least one of a velocity of the movable object or an acceleration of the movable object, and the parameter of the imaging device including at least one of a velocity of the imaging device or an acceleration of the imaging device.

19. The non-transitory computer-readable medium of claim 18 , wherein the one or more reference points comprise a single reference point and the instructions further cause the processor to:

identify a feature region in the image that corresponds to coordinates of the single reference point, wherein a size of the feature region is configurable; and

generate the control data indicating a distance between the feature region and a next single reference point to control the parameter of the movable object or the parameter of the imaging device, the parameter of the movable object further including at least one of a position, an orientation, or an attitude of the movable object, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude, a zoom factor, a focus length, or an operational parameter of the imaging device.

20. The non-transitory computer-readable medium of claim 18 , wherein the one or more reference points comprise multiple reference points including at least an initial point and a release point, and the instructions further cause the processor to:

identify a feature region in the image that corresponds to coordinates of the initial point, wherein a size of the feature region is configurable; and

generate the control data indicating a distance between the feature region and the release point to control the parameter of the movable object or the parameter of the imaging device, the parameter of the movable object further including at least one of a position, an orientation, or an attitude of the movable object, and the parameter of the imaging device further including at least one of a position, an orientation, an attitude, a zoom factor, a focus length, or an operational parameter of the imaging device.