METHOD AND APPARATUS TO DETERMINE ROBOT LOCATION USING OMNI-DIRECTIONAL IMAGE

• US 20080075357A1
• Filed: 07/31/2007
• Published: 03/27/2008
• Est. Priority Date: 09/26/2006
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Roh, Kyung Shik, Min, Seung Ki, Han, Woosup, Yoon, Sukjune
• Granted Patent: US 8,498,488 B2
• Time in Patent Office: Days
• US Class Current: 382/153
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: G06V 20/10 Terrestrial scenes scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle

Abstract

A method to determine the location of a robot using an omni-directional image, the method including acquiring an omni-directional image from a robot, extracting a predetermined current line from the acquired omni-directional image, calculating a correlation coefficient between the extracted current line of the robot and each landmark line of pre-stored nodes using a Fast Fourier Transform (FFT), and performing a stochastic approach method of a particle filtering process on a basis of the calculated correlation coefficient to recognize a location of the robot.

Claims

1. A method to locate a robot using an omni-directional image, the method comprising:

acquiring the omni-directional image from a robot;

extracting a predetermined current line from the acquired omni-directional image;

calculating a correlation coefficient between the extracted current line of the robot and each landmark line of pre-stored nodes using a Fast Fourier Transform (FFT);

selecting a node at which the calculated correlation coefficient is equal to or higher than a predetermined value; and

recognizing a location of the robot on a basis of the selected node.

2. The method of claim 1, wherein the predetermined current line is a circumference spaced apart from an origin of the omni-directional image by a predetermined location.

3. The method of claim 1, wherein the correlation coefficient between the current line of the robot and each landmark line of the pre-stored nodes is calculated by the following equation:

$\begin{array}{cc}\mathrm{\rho <br><br>}<br><br>\left(\mathrm{\tau <br><br>}\right)=\frac{C_{\mathrm{xy}}<br><br>\left(\mathrm{\tau <br><br>}\right)}{\sqrt{C_{\mathrm{xx}}<br><br>\left(0\right)·<br><br>C_{\mathrm{yy}}<br><br>\left(0\right)}}& \left[\mathrm{Equation}\right]\end{array}$

wherein ρ

(τ

) is a correlation coefficient, τ

is a rotation angle of the robot, C_xy is a cross-correlation value, and C_xx and C_yy are correlation values.

4. The method of claim 1, further comprising:

selecting M nodes at which the calculated correlation coefficient is equal to or higher than a predetermined value;

modifying a current line of the robot such that same spatial objects are located at a same location on a basis of the landmark lines of the selected M nodes to create a wrapped current line;

calculating a correlation coefficient between the wrapped current line of the robot and the landmark lines of the M nodes;

selecting N nodes at which the calculated correlation coefficient is equal to or higher than a predetermined correlation coefficient such that N<

M; and

recognizing a location of the robot on the basis of the selected N nodes.

5. The method of claim 4, wherein the modifying of the current line comprises:

extracting an edge of the current line of the robot and an edge of each landmark line of the M nodes;

matching the edge of the current line of the robot with the edge of the landmark line of the M nodes; and

modifying the current line of the robot on a basis of the matched edge, such that the same spatial objects are located at the same location of the current line.

6. The method of claim 5, wherein the edge is indicative of an end point at which each RGB of the current line changes by a predetermined value or more.

7. A method to locate a robot using an omni-directional image, the method comprising:

acquiring an omni-directional image from a robot;

extracting a predetermined current line from the acquired omni-directional image;

calculating a correlation coefficient between the extracted current line of the robot and each landmark line of pre-stored nodes using a Fast Fourier Transform (FFT); and

performing a stochastic approach method of a particle filtering process on a basis of the calculated correlation coefficient to recognize a location of the robot.

8. The method of claim 7, further comprising:

randomly dispersing particles on an overall robot-locatable area on a map on a basis of a specific node determined by the calculated correlation coefficient;

extracting particles on the basis of a probability value indicating that the robot is located at the dispersed particles; and

recognizing a current location of the robot on the basis of the extracted particles.

9. The method of claim 8, further comprising:

uniformly dispersing the particles on an overall robot-locatable area expected by both a

selecting N nodes at which the calculated correlation coefficient is equal to or higher than a predetermined correlation coefficient such that N<

M; and

recognizing a location of the robot on a basis of the selected N nodes.

14. The method of claim 13, wherein the modifying of the current line comprises:

extracting an edge of the current line of the robot and an edge of each landmark line of the M nodes;

matching the edge of the current line of the robot with the edge of the landmark line of the M nodes; and

modifying the current line of the robot on the basis of the matched edge, such that the same spatial objects are located at the same location of the current line to create the wrapped current line.

10. The method of claim 8, wherein a probability value indicating that the robot is located at the dispersed particles increases in proportion to the calculated correlation coefficient.

11. The method of claim 7, further comprising:

randomly dispersing particles on an overall robot-locatable area on a map on a basis of a specific node determined by the calculated correlation coefficient;

assigning a weight on a basis of a probability value indicating that the robot is located at the dispersed particles;

randomly extracting the dispersed particles on a basis of a value to which the weight is added; and

recognizing a current location of the robot on a basis of the randomly extracted particles.

12. The method of claim 7, wherein the correlation coefficient between the current line of the robot and each landmark line of the pre-stored nodes is calculated by the following equation:

$\begin{array}{cc}\mathrm{\rho <br><br>}<br><br>\left(\mathrm{\tau <br><br>}\right)=\frac{C_{\mathrm{xy}}<br><br>\left(\mathrm{\tau <br><br>}\right)}{\sqrt{C_{\mathrm{xx}}<br><br>\left(0\right)·<br><br>C_{\mathrm{yy}}<br><br>\left(0\right)}}& \left[\mathrm{Equation}\right]\end{array}$

wherein ρ

(τ

) is a correlation coefficient, τ

is a rotation angle of the robot, C_xy is a cross-correlation value, and C_xx and C_yy are correlation values.

13. The method of claim 7, further comprising:

selecting M nodes at which the calculated correlation coefficient is equal to or higher than a predetermined value;

modifying a current line of the robot such that the same spatial objects are located at the same location on the basis of the landmark lines of the selected M nodes to create a wrapped current line;

calculating a correlation coefficient between the wrapped current line of the robot and the landmark lines of the M nodes;

selecting a second number of nodes among the first number of nodes at which the second calculated correlation coefficient is equal to or higher than a second predetermined value; and

recognizing a location of the robot on the basis of the selected second number of nodes.

15. A method to locate a robot with an omni directional camera mounted thereon, the method comprising:

setting a number landmark lines for a predetermined number of nodes within a robot-locatable area;

acquiring an omni directional image from the robot;

extracting a current line from the acquired image;

calculating a correlation coefficient between the current line and the landmark lines for each node; and

determining a location of the robot based on the calculated correlation coefficient.

16. The method of claim 15, wherein the calculating of the correlation coefficient uses a Fast Fourier Transform (FFT) method.

17. The method of claim 15, wherein the determining of the location comprises:

selecting a first number of nodes at which the calculated correlation coefficient is equal to or higher than a predetermined value;

modifying the current line into a plurality of wrapped current lines such that spatial objects in each wrapped current line are at the same location with corresponding spatial objects in each of the landmark lines for the first number of nodes;

calculating a second correlation coefficient between the wrapped current lines and the landmark lines of the first number of nodes;

previous location of the robot and a driving command of the robot.

18. The method of claim 17, wherein the modifying of the current line comprises:

extracting an edge of the current line and an edge of each landmark line of the first number of nodes; and

matching the edge of the current line with the edge of the landmark lines.

19. The method of claim 17, wherein the predetermined value and the second predetermined value are the same.

20. The method of claim 18, wherein the edge is an end point at which each RGB of the current line changes by at least a third predetermined value.

21. A robot locating apparatus, comprising:

a robot body;

an omni-directional camera mounted on the robot body to acquire an omni-directional image; and

a controller to extract a predetermined current line from the acquired omni-directional image, to calculate a correlation coefficient between the extracted current line of the robot and each landmark line of pre-stored nodes of a robot locatable area using a Fast Fourier Transform (FFT), and to select a node at which the calculated correlation coefficient is equal to or higher than a predetermined value to recognize a location of the robot based on the selected node.

22. The apparatus of claim 21, wherein the controller is mounted apart from the robot body and communicates with at least one of the robot body and the omni-directional camera.