AUTOMATIC RECOVERY METHOD FOR AN UNMANNED AERIAL VEHICLE

• US 20130325222A1
• Filed: 12/22/2010
• Published: 12/05/2013
• Est. Priority Date: 12/14/2010
• Status: Active Grant

• Current Assignee: Korean Air Lines Company Limited
• Original Assignee: Korean Air Lines Company Limited
• Inventors: Roh, Eun Jung
• Granted Patent: US 9,207,681 B2
• Time in Patent Office: Days
• US Class Current: 701/18
• CPC Class Codes: B64C 39/024 of the remote controlled vehicle type, i.e. RPV
• CPC Class Codes: B64F 1/02 for arresting aircraft, e.g. nets or cables
• CPC Class Codes: B64F 1/027 using net or mesh
• CPC Class Codes: B64U 70/30 for capturing UAVs in flight by ground or sea-based arresting gear, e.g. by a cable or a net
• CPC Class Codes: G05D 1/0676 specially adapted for landing

Abstract

The present invention relates to an automatic recovery method for an unmanned aerial vehicle (UAV), and more particularly to a vision-based automatic recovery method for an unmanned aerial vehicle (UAV) using a reference trajectory, in which information from a front vision camera and information from navigation equipment embedded in the unmanned aerial vehicle are combined by a weighting method.

The vision-based automatic recovery method using a reference trajectory for an unmanned aerial vehicle includes a reference trajectory generating step of generating the reference trajectory connecting an align point and a recovery point on an on-board based on the recovery point given to the unmanned aerial vehicle through wireless communication; an image processing step of taking an image containing a symbol (e.g., a net) through a front vision camera embedded in the unmanned aerial vehicle, determining the position and size of the symbol in the taken image through image processing computer, and calculating a distance between the symbol and the unmanned aerial vehicle; an attitude command issuing step of generating an attitude command for centering the symbol based on the position of the symbol at the center of the image taken in the image processing step; and a virtual trajectory generating step of generating a virtual trajectory by combining the reference trajectory generated in the reference trajectory generating step and the attitude command generated in the attitude command generating step, in which the virtual trajectory is generated by moving a weight to the attitude command as going from the align point to the recovery point.

Claims

1. A vision-based automatic recovery method using a reference trajectory for an unmanned aerial vehicle, the method comprising:

a reference trajectory generating step of generating the reference trajectory connecting an align point and a recovery point on an on-board based on the recovery point given to the unmanned aerial vehicle through wireless communication;

an image processing step of taking an image containing a symbol (e.g., a net) through a front vision camera embedded in the unmanned aerial vehicle, determining the position and size of the symbol in the taken image through image processing computer, and calculating a distance between the symbol and the unmanned aerial vehicle;

an attitude command issuing step of generating an attitude command for centering the symbol based on the position of the symbol at the center of the image taken in the image processing step; and

a virtual trajectory generating step of generating a virtual trajectory by combining the reference trajectory generated in the reference trajectory generating step and the attitude command generated in the attitude command generating step, in which the virtual trajectory is generated by moving a weight to the attitude command as going from the align point to the recovery point.

2. The method according to claim 1, wherein the reference trajectory is set up to pass one or more waypoints from the align point to the recovery point, in which a virtual reference trajectory for going toward the next waypoint is generated before passing each waypoint, and a guide control command is issued to follow the generated virtual reference trajectory based on information of embedded navigation equipment.

3. The method according to claim 1, wherein the attitude command comprises a command to calculate the azimuth angle ψ

_VPUCmd and the elevation angle θ

_VPUCmd between the positions of the symbol in the center of the taken image, and change a direction of the unmanned aerial vehicle as much as the calculated azimuth angle ψ

_VPUCmd and elevation angle θ

_VPUCmd.

4. The method according to claim 1, wherein the moving the weight in the virtual trajectory generating step comprises linearly increasing weight of the attitude command from 0% to 100% as going from the align point to the recovery point.

5. The method according to claim 1, wherein distance calculation for moving the weight in the virtual trajectory generating step comprises using a distance between the symbol and the unmanned aerial vehicle calculated in the image processing step.

6. The method according to claim 1, wherein the generation of the virtual trajectory improves precision by following the reference trajectory through navigation information in an initial align section and generating an attitude command based on high-precise image information in a final section, i.e., at the recovery point.