KR100703882B1 - Mobile robot capable of pose sensing with a single camera and method thereof - Google Patents

Abstract

The present invention relates to a mobile robot capable of single camera-based pose recognition and a method thereof. According to the present invention, at least two milestones attached to the bottom surface are recognized by one camera, and the absolute coordinates of the corresponding point and the absolute coordinates of the origin of the own coordinate system and its coordinates are obtained by obtaining the absolute coordinates of the corresponding point and the coordinates of the mobile robot's own coordinates from the milestones. Find the pose of the mobile robot displayed by the degree of rotation. As a result, the mobile robot can recognize its own pose using only one camera, an expensive spatial sensor for extracting three-dimensional coordinates, a plurality of cameras, and the complexity of the design and the size of data to be processed. It can solve the problem, and can increase the processing speed of the pose recognition of the mobile robot.

Mobile Robot, Localization, Navigation, pose, Pose Recognition, Position Recognition, Camera, Image Coordinate

Description

Mobile robot capable of pose sensing with a single camera and method

1 is a view provided for explaining the operation of a mobile robot according to the present invention;

2 is a block diagram of a mobile robot capable of recognizing a single camera based on the present invention;

3 is a flowchart provided to explain a pose recognition method of a mobile robot according to one embodiment of the present invention;

4 is a view provided for explaining a method of obtaining a robot coordinate from an image coordinate of a milestone point;

FIG. 5 is a diagram showing a part of a robot coordinate system provided for explaining a method of obtaining y _r among robot coordinates of a milestone point; and

6 is a diagram illustrating a pose of a mobile robot in a global coordinate system.

The present invention relates to pose recognition of a mobile robot, and in particular, a single camera-based pose recognition capable of recognizing its position and rotation on a plane by recognizing a milestone of a floor surface through one camera. The present invention relates to a mobile robot and a method thereof.

A mobile robot performs a given task by moving itself in space rather than in a fixed position. The mobile robot moves the parts and work tools necessary for the production of the product to the required position, and may also assemble the moved parts to produce the product. Recently, many examples of the use of mobile robots in homes as well as in industrial fields have been published. At home, have your mobile robot cleaned or moved.

There are various researches on navigation of mobile robots that perform a given task by moving a space by themselves instead of a fixed position.

In order for the mobile robot to search for a specific target point or to move within a predetermined space, to recognize the location of the mobile robot and to perform a predetermined purpose, various functions must be provided. Such features include avoiding obstacles, identifying specific indicators or your current pose and path planning.

In the case of an intelligent mobile robot that cleans the floor while moving along the floor at home, accurate pose recognition will be important. Here, the pose includes its localization and rotation.

Conventionally, a spatial sensor such as an acoustic detector (sonar, sonar), a laser (laser) and a stereo camera is used to extract three-dimensional coordinates of the surface. Among these, the pose recognition of the camera-based mobile robot mainly extracted the three-dimensional coordinates of the surface using a stereo camera.

In order to estimate the pose of the robot, the mobile robot uses an algorithm that estimates the position of the robot using geometric data simultaneously to measure the displacement of the robot wheel along with the sensor information.

As such, the pose recognition of the conventional mobile robot essentially requires the extraction of three-dimensional coordinates of the surface, and thus requires an expensive spatial sensor. Furthermore, since geometric data is used together with sensor information, components such as encoders and signal processing algorithms for them are additionally needed.

Accordingly, an object of the present invention is to provide a mobile robot and a method capable of recognizing a single camera-based pose capable of recognizing its position and rotation degree on a plane by recognizing a milestone of a floor through one camera. have.

In order to achieve the above object, according to the present invention, the global coordinates and the degree of rotation of the origin of the robot coordinate system which is a coordinate system whose position is the origin on a global coordinate system which is a preset at least two-dimensional coordinate system on the floor plane A mobile robot capable of recognizing a pose may include a plurality of milestones, a single camera, an image processor, and a controller, which are attached to an arbitrary point on the bottom surface and visually display information including global coordinates of the point. Include.

The camera is fixed to the mobile robot to generate and output an image including at least two milestones, and the image processor extracts the global coordinates of the milestone points from the milestone images included in the received image. The controller calculates the pose by calculating a relationship between the robot coordinates of the milestone point and the global coordinates of the milestone point calculated based on the geometric relationship in which the milestone points are recognized in the robot coordinate system.

The camera is installed on the front of the mobile robot at a predetermined angle toward the bottom surface, it is preferable that the running direction of the mobile robot and the direction of the camera's gaze coincide with each other.

The milestone is preferably at least one of a bar code, a series of Arabic numerals, and a character in displaying information including the global coordinates of the point.

In an embodiment, the image processor extracts the image coordinates of the milestone image on the image coordinate system, which is a two-dimensional coordinate system applied to the image, and outputs the image coordinates to the controller, wherein the controller is a three-dimensional image between the robot coordinate system and the image coordinate system. The robot coordinates of the milestone point are calculated based on a predetermined geometric relationship.

Here, the robot coordinate system is formed by two axes on the bottom surface and the other axis orthogonal to each other and the two axes on the bottom surface coincide with the traveling direction of the mobile robot and the direction perpendicular to the traveling direction after passing through the origin. An axis perpendicular to the traveling direction of the robot coordinate system and one axis of the image coordinate system are parallel to each other.

When the relationship between the global coordinates of the milestone point and the robot coordinates of the milestone point is expressed by the following equation, the pose becomes the global coordinates (P _X , P _Y ) of the origin of the robot coordinate system, and with respect to the global coordinate system. The state is rotated by θ.

Here, the global coordinate of the milestone point is (x _w , y _w ), and the robot coordinate is (x _r , y _r ).

According to another embodiment of the present invention, the controller may further include a driving unit configured to control the mobile robot to travel according to a control command of the controller, wherein the controller calculates the pose by at least one of periodic and aperiodic intervals while the mobile robot is traveling. The control command is transmitted to the driving unit so that the mobile robot travels based on the calculated pose.

Furthermore, the bottom surface may be cleaned by further including a dust collecting device for collecting dust on the bottom surface facing the air while inhaling air.

According to another embodiment of the present invention, the global coordinates and the degree of rotation of the origin of the robot coordinate system which is a coordinate system whose origin is the coordinates on the global coordinate system which is a preset at least two-dimensional coordinate system on the floor plane which is a predetermined plane The pose recognition method of the mobile robot shown includes generating an image, extracting global coordinates, and calculating a pose.

The generating of the image may include at least two of the milestones of a plurality of milestones visually displayed with information including a global coordinate of the point and attached to an arbitrary point on the bottom surface by using a camera fixed on the mobile robot. Create an image that includes.

Extracting the global coordinates extracts the global coordinates of the milestone point from the milestone image included in the image.

The calculating of the pose may include calculating the pose by calculating a relationship between the robot coordinates of the milestone points and the global coordinates of the milestone points calculated based on a geometric relationship in which the milestone points are recognized in the robot coordinate system. Calculate

The invention can be implemented with methods, apparatus and systems. Hereinafter, the present invention will be described in detail with reference to the drawings.

1 is a view provided for explaining the operation of the mobile robot according to the present invention.

Referring to FIG. 1, the mobile robot 200 of the present invention may move in a predetermined manner along a flat bottom surface, and may include its own camera 203 to recognize its pose. A pose refers to its position and degree of rotation, where the position refers to a coordinate on a given absolute coordinate system.

The mobile robot 200 first extracts the absolute coordinates of a specific point located in front of itself through the camera 203. Subsequently, the mobile robot 200 based on the geometric relationship in which the specific point is recognized in its viewpoint (coordinate system whose origin is the current position), and the absolute point of view (coordinate system) and absolute from the absolute coordinates of the extracted specific point. Calculate your pose by knowing the relationship of the coordinate system. In other words, the pose of the mobile robot 200 is the position and rotation degree of the origin of its coordinate system on the absolute coordinate system.

Hereinafter, various coordinate systems used for proper description of the present invention will be briefly described. In the present invention, three coordinate systems including a global coordinate system, an image coordinate system, and a robot coordinate system are used.

The global coordinate system is an absolute coordinate based on a preset origin, and the mobile robot 200 obtains its global coordinate through the method of the present invention and recognizes its pose accordingly. In FIG. 1, the global coordinate system is X _W , Y _W , respectively. And Z _W as each axis.

The video coordinate system is a coordinate set by setting a predetermined position in the image generated by the camera 203 of the mobile robot 200 as an origin, and basically two-dimensional coordinates are sufficient. Accordingly, the image coordinate system is formed by two lines included in the imaging surface (not shown) of the camera 203 and perpendicular to each other. In this case, the direction that the imaging surface (not shown) faces corresponds to the traveling direction of the mobile robot 200.

However, when a three-dimensional coordinate system is used for proper explanation, a coordinate system formed by a line perpendicular to the imaging surface (not shown) passing through two lines on the imaging surface (not shown) and the two lines may be used. In the video coordinate system, X _I and Y _I And Z _I are represented by each axis.

The video coordinate system changes in correspondence with the global coordinate system according to the position of the camera 203 fixed to the mobile robot 200 and the movement of the mobile robot 200. However, since the relative positions of the camera 203 and the mobile robot 200 are fixed, the robot coordinate system and the image coordinate system described below will have a fixed correspondence.

The robot coordinate system is a coordinate system of the mobile robot 200 which has a predetermined portion of the mobile robot 200 as its origin, and the corresponding relationship between the global coordinate system and the robot coordinate system changes according to the traveling of the mobile robot 200.

Preferably, one axis of the robot coordinate system is a driving direction of the mobile robot 200. In Figure 1, the robot coordinate system is X _R , Y _R , respectively. And Z _R as each axis. However, one axis of the robot coordinate system does not necessarily have to be the driving direction, and the corresponding relationship between the image coordinate system and the global coordinate system may be adjusted accordingly.

The correlation between the global coordinate system, the robot coordinate system and the image coordinate system is as described above. In summary, the robot coordinate system and the image coordinate system have a predetermined fixed correspondence and are changed on the global coordinate system.

One axis of the robot coordinate system included on the bottom surface and one axis of the image coordinate system are preferably parallel. However, it is not limited thereto. The geometric correspondence between the robot coordinate system and the image coordinate system can be adjusted.

Furthermore, one axis of the robot coordinate system included on the bottom surface preferably coincides with the traveling direction of the mobile robot 200. However, the present invention is not limited thereto, and only the driving algorithm of the mobile robot 200 using the calculated pose may be compensated.

For convenience of explanation of the present invention, hereinafter, X _R of the robot coordinate system And X _W of the global coordinate system and the plane formed by Y _R And planes formed by Y _W are parallel to each other. In addition, X _R of the robot coordinate system and X _I of the image coordinate system are parallel to each other, and Y _R of the robot coordinate system is parallel to each other. And Z _R are the Y plane and the image coordinate system to generate _I And the planes generated by Z _I are parallel to each other.

The mobile robot 200 recognizes its position by determining where its origin, that is, the origin of the robot coordinate system, is located in the global coordinate system, and how much of one axis in the robot coordinate system is rotated in one axis of the global coordinate system.

To this end, the mobile robot 200 recognizes its own pose by reading at least two 'signals' displayed on the floor using the camera 203. A milestone is a marker that has the coordinates of its marked bottom position, attached to or printed on the bottom. In FIG. 1, milestones are marked at two points A and B on the bottom surface.

The milestones are displayed on the bottom surface in such a manner that the predetermined number includes attachment, printing, etc. at predetermined intervals, and should be displayed so that at least two are read through the camera 203. The milestone has information about its global coordinates in various ways, and the method may correspond to a method that is visually recognizable such as a bar code, a series of Arabic numerals or letters.

The camera 203 attached to the mobile robot 200 is fixed at a predetermined angle with respect to the floor surface, and reads the global coordinates of at least two milestones on the floor surface, and the corresponding global coordinates of the milestones are robot coordinate systems. By determining where the image is located, the global coordinate and the degree of rotation of the origin of the robot coordinate system in the global coordinate system are determined.

Through this method, the mobile robot 200 can recognize its own pose even if it displays a relatively small number of milestones, and the number of milestones displayed in the predetermined space is the camera 203 installed on the mobile robot 200. Can vary depending on the height and angle.

Hereinafter, each part of the mobile robot 200 will be described with reference to FIG. 2.

2 is a block diagram of a mobile camera capable of position recognition based on a single camera according to the present invention.

Hereinafter, a method of using two milestones for proper description of the present invention will be described. For this purpose, it is assumed that two milestones marked at positions A and B of FIG. 1 are simultaneously read through the camera 203.

2, the mobile robot 200 includes a sensor 201, a camera 203, an image processor 205, a driver 207, a storage medium 209, and a controller 211. And a user interface unit 213.

The mobile robot 200 of FIG. 2 illustrates only a basic configuration of the mobile robot of the present invention, and omits various configurations according to the role of the mobile robot. For example, in the case of a mobile robot for cleaning, a dust collecting device (not shown) for collecting dust on the opposite floor while inhaling air may be controlled by the controller 211.

The sensor unit 201 includes various sensors and generates information for tasks such as obstacle detection and mileage detection. Examples of the sensor included in the sensor unit 201 may include an obstacle detecting sensor that detects an obstacle by transmitting a signal to the outside and receiving a reflected signal, and a traveling distance detecting sensor capable of measuring a driving distance.

The camera 203 is installed in the mobile robot 200 at a fixed angle with respect to the floor surface so as to capture an image in front of the camera 203 as in the example of FIG. 1, and preferably the global coordinate system X _W , Y _W. The mobile robot 200 is installed on the front surface of the mobile robot 200 so that the traveling direction of the mobile robot 203 projected on the plane coincides with the gaze direction of the camera 200. However, the overall algorithm for calculating the pose of the mobile robot 200 from the captured image is the same even if the camera 203 is not installed in the front or the driving direction and the viewing direction do not coincide. The algorithm for controlling the movement of 200 may vary.

Since the width of the bottom surface that can be generated as one image varies depending on a predetermined angle at which the camera 203 is installed, the number of milestones to be displayed on the floor may vary.

The camera 203 generates an image including at least two milestones from the bottom surface according to the movement of the mobile robot 200. Therefore, the camera 203 transmits the image including the milestones indicated by A and B to the image processor 205.

The image processor 205 extracts the image of the milestones indicated by A and B from the image received from the camera 203, and extracts the global coordinates of the bottom points A and B from the extracted image of the milestone. In addition, the image processor 205 extracts image coordinates of A and B points in the received image. The method of extracting the global coordinates from the image received from the camera 203 and the method of extracting the image coordinates by the image processor 205 may use techniques known in the art and the image processing field of the present invention.

The image processor 205 outputs the extracted image coordinates and global coordinates of the A and B points to the controller 211.

The driving unit 207 controls the mobile robot 200 to move according to the command of the controller 211. The driving unit 207 includes a motor (not shown) for rotating the wheels, respectively, and a power transmission unit (not shown) for transmitting power generated by the motor. The driving unit 207 controls the respective motors according to the control command of the control unit 211 to perform an operation including forward, backward and direction change.

The storage medium 209 corresponds to predetermined active and inactive memory and other hard disks, and various programs related to the operation of the control unit 211 are stored, information about the global coordinate system, and global coordinates. Map information of the workspace and the pose information calculated while driving can be stored.

The control unit 211 controls the overall operation of the mobile robot 200 including forward, backward and rotation, and at the same time receives a predetermined control command from the user through the user interface unit 213 or performs predetermined information. To the user.

The controller 211 extracts its own pose based on the information received from the image processor 205. Information received from the image processor 205 includes global coordinates and image coordinates of two milestones included in the image generated by the camera 203. That is, in FIG. 1, global coordinates and image coordinates of the bottom surfaces A and B are input to the controller 211.

The controller 211 obtains the global coordinates and the degree of rotation of the mobile robot 200 from the received global coordinates and the image coordinates.

The control unit 211 controls the driving unit 207 or the like to perform its own operation based on the obtained pose of the user.

The user interface unit 213 includes an input unit (not shown) and a display unit (not shown) for providing an interface between the mobile robot 200 and the user. In addition, the user interface unit 213 may include an interface for a predetermined wired or wireless connection, and thus may receive a user's command or provide predetermined information to the user through a wired or wireless connection.

Hereinafter, a pose recognition method of the mobile robot of the present invention will be described with reference to FIG. 3.

3 is a flowchart provided to explain a pose recognition method of a mobile robot according to an embodiment of the present invention. The mobile robot 200 can know its pose in the global coordinate system by reading the milestone at the current position, and does not need history management for the previous movement path.

The camera 203 of the mobile robot 200 generates an image including the milestones displayed on the bottom surfaces A and B and transmits the image to the image processor 205 (S301).

The image processor 205 extracts the global coordinates and the image coordinates of the A and B positions from the received image. For example, if the milestone is in the form of a barcode, the global coordinates of A and B are obtained from the barcode included in the image, and the image coordinates of the position of the landmark on the received image are obtained. The image processor 205 transfers the obtained global coordinates and image coordinates of the two A and B points to the controller 211 (S303).

The controller 211 obtains a robot coordinate from the image coordinates of two points A and B based on a predetermined correspondence, and then obtains a pose of the mobile robot 200 by applying the calculated coordinates to the robot coordinates and the global coordinates. (S305).

Hereinafter, the step S305 of the step of FIG. 3 will be described in more detail with reference to FIGS. 4 to 6. First, a method of obtaining robot coordinates of two milestones by a predetermined method based on the image coordinates of two milestones will be described.

4 to 6, X _R of the robot coordinate system and X _I of the image coordinate system are parallel to each other, and Y _R of the robot coordinate system is parallel to each other. And Y _I of the plane and image coordinate system formed by Z _R And the planes formed by Z _I are parallel to each other.

However, the present invention is not limited to this case, but it is sufficient to maintain a predetermined fixed correspondence between the image coordinate system and the robot coordinate system.

First, FIG. 4 is a diagram provided to explain a method of obtaining robot coordinates from image coordinates of a milestone point. Referring to FIGS. 1 to 4, a method of obtaining x _r of robot coordinates (x _r , y _r ) from the image coordinates. Explain. Describe with one milestone for proper explanation.

4 (a) shows an image c of the camera 203 projected onto the image coordinate system, and shows that the image coordinate at point A is (x _i , y _i ).

Figure 4 (b) shows two axes X _R , Y _R A denotes a robot coordinate system, based on, x _r is would display the coordinate of (x _r, y _r) on the robot coordinate system of the point A, and displays only the center x _r.

Considering the position of the camera 203 and the position of the image coordinate system and the position of the robot coordinate system of FIG. 1, x _i and x _r have a linear relationship, and as x _i increases, x _r also increases, so that Equation 1 and Have the same relationship.

Where K is a linear coefficient and P is a constant. However, since x _r = 0 when x _i = 0, P = 0. The image projected on the image coordinate system has a characteristic that K changes linearly with y _i since the substantial x _i becomes larger as y _i increases according to the characteristics of the lens of the camera 203. D of FIG. 3 (b) shows it.

Therefore, the final relationship between x _i and x _r is shown in Equation 2 below.

Here, a and b are mapping variables, and are obtained by actual measurement of an image and an actual distance projected on the image coordinates.

Hereinafter, a method of obtaining y _r among robot coordinates (x _r , y _r ) will be described with reference to FIG. 5.

FIG. 5 is a view showing a part of a robot coordinate system provided for explaining a method of obtaining y _r among robot coordinates at a milestone point, and among robot coordinates (x _r , y _r ) from image coordinates with reference to FIGS. 1 to 5. Describes how to find y _r .

5 is a cross section of a robot coordinate system in which X _R = 0, and the horizontal and vertical axes are Y _R and Z _R , which are one axis of the robot coordinate system.

Here, h is the height of the center of the camera 203, e is the image plane (imaginary plane) showing the state of the image captured by the camera 203 in a virtual plane, y _off is the center of the mobile robot 200 (robot coordinate system origin) and Distance on the Y _R axis from the camera 203 position, y _rc is the value at which the center of the image plane e is projected on the Y _R axis, ψ is one end angle of the isosceles triangle formed by points h, y _s , p, φ is the angle of the camera 203. In the above, h, y _s , y _rc are the values calculated | required actually.

The robot coordinate y _r to be obtained is obtained as in Equation 3 below according to the geometry of FIG. 5.

Where γ is offy _off hy _s to be. λ is ∠y _s hy _r to be.

이며,

와

는 실제로 측정한 거리이다. Since the triangle formed by the center of the camera 203 and the image plane e and the isosceles triangle formed by the points h, y _s and p have a similar relationship, y _p = my _i . Where m is the similarity ratio

Is,

Wow

Is the actual measured distance.

By the geometric relationship in Fig. 5, γ and λ are as shown in the following equations (4) and (5).

,

이다.here,

,

to be.

Therefore, Equation 3 for obtaining the robot coordinate y _r may be expressed as in Equation 6 below.

이고, h, y_s, y_off는 실측으로 구해지므로, 수학식 6은 영상좌표계 상의 y_i에 의해 고유하게 계산되는 y_r의 비선형 방정식이 된다.here,

Since h, y _s and y _off are found by measurement, Equation 6 becomes a nonlinear equation of y _r that is uniquely calculated by y _i on the image coordinate system.

Hereinafter, a method of obtaining a pose of the mobile robot 200 using the obtained two coordinates of the robot coordinates will be described.

6 is a diagram illustrating a pose of a mobile robot in a global coordinate system.

Referring to FIG. 6, the global coordinate system is displayed based on two axes X _W and Y _W , and the robot coordinate system is displayed on two axes X _R and Y _R based on the moving direction of the robot indicated by a thick arrow. It became.

The image coordinates of the two milestones marked at points A and B are (x _i1 , y _i1 ) and (x _i1 , y _i1 ), respectively, and the global coordinates are (x _w1 , y _w1 ) and (x _w2 , y _w2 ), respectively. Seems to be. And, when the global coordinates of the mobile robot 200 to obtain the origin of the robot coordinate system, the position of the mobile robot is (P _X , P _Y ), the mobile robot is rotated by θ from the axis X _W.

Therefore, the pose of the mobile robot can be expressed as (P _X , P _{Y ,} θ).

A matrix for mapping an arbitrary point of the robot coordinate system to the coordinates of the global coordinate system is expressed by Equation 7 below. Where _W T ^R is the mapping matrix.

Since α + θ = 90 °, the relationship between the global coordinate system and the robot coordinate system is expressed by Equation 8 below.

Using global coordinates (x _w1 , y _w1 ), (x _w2 , y _w2 ) and robot coordinates (x _r1 , y _r1 ), (x _r2 , y _r2 ) (P _X , P _Y ) to obtain Equations 9 and 10 below.

Here, θ is the rotation angle of the mobile robot 200, as shown in the following equation (11).

In this way, the current global coordinates and the degree of rotation of the mobile robot 200 can be obtained. Therefore, the pose of the mobile robot becomes (P _X , P _{Y ,} θ).

While the above has been shown and described with respect to preferred embodiments of the invention, the invention is not limited to the specific embodiments described above, it is usually in the technical field to which the invention belongs without departing from the spirit of the invention claimed in the claims. Various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or prospect of the present invention.

As described above, according to the present invention, the mobile robot can recognize its pose by using only one camera, and its position and rotation degree without managing the history of the previous movement path. Can be recognized immediately.

The mobile robot according to the present invention does not require various spatial sensors for pose recognition, and does not use a plurality of cameras such as a stereo camera, thereby relieving the complexity of the design and the size of the data to be processed. Accordingly, the moving speed and the processing speed of the mobile robot can be further increased.

In addition, the manufacturing cost of the mobile robot can be reduced by not using expensive sensors or cameras.

The mobile robot pose recognition method of the present invention can perform intelligent work using its pose recognition by applying to a robot navigation field. For example, in the case of a cleaning mobile robot, the robot accurately grasps its pose so that the robot does not cross the path once passed and prevents the unscanned portion from occurring in the predetermined cleaning area.

Claims (14)

A mobile robot capable of recognizing poses representing the global coordinates and the degree of rotation of the origin of the robot coordinate system, which is a coordinate system whose origin is the origin, on a global coordinate system which is a predetermined two-dimensional coordinate system on the floor plane that is a predetermined plane. for robots, A plurality of milestones provided at any point on the bottom surface and visually displaying information including global coordinates of the point; A camera fixed to the mobile robot to generate an image including at least two milestones; An image processor extracting a global coordinate of the milestone point from the milestone image included in the image received from the camera; And And a controller configured to calculate the pose by calculating a relationship between the robot coordinates of the milestone points and the global coordinates of the milestone points, calculated based on the geometric relationship in which the milestone points are recognized in the robot coordinate system. Wherein the milestone, in displaying the information including the global coordinates of the point mobile robot capable of pose recognition, characterized in that displayed by at least one of a bar code, a series of Arabic numerals, characters. The method of claim 1, The camera is installed on the front of the mobile robot at a predetermined angle toward the bottom surface, A moving robot capable of pose recognition, characterized in that the traveling direction of the mobile robot and the gaze direction of the camera coincide with each other. delete The method of claim 1, The image processor extracts the image coordinates of the milestone image on the image coordinate system, which is a two-dimensional coordinate system applied to the image, and outputs the image coordinates to the controller. And the controller calculates a robot coordinate of the milestone point based on a predetermined geometric relationship between the robot coordinate system and the image coordinate system in three dimensions. The method of claim 4, wherein The robot coordinate system is formed by two axes on the bottom surface and the other one orthogonal to each other, the two axes on the bottom surface that pass through the origin and coincide with the traveling direction of the mobile robot and the direction perpendicular to the traveling direction, A robot capable of recognizing poses, wherein an axis perpendicular to the traveling direction of the robot coordinate system and one axis of the image coordinate system are parallel to each other. The method of claim 5, When the relationship between the global coordinates of each milestone point and the robot coordinates of the milestone point is expressed by the following equation, The pose is a global coordinate (P _X , P _Y ) of the origin of the robot coordinate system, the robot can pose recognition, characterized in that rotated by the θ relative to the global coordinate system.

Here, the global coordinate of the milestone point is (x _w , y _w ), the robot coordinate is (x _r , y _r ). The method of claim 1, And a driving unit which controls the mobile robot to travel according to a control command of the controller. The control unit may calculate the pose at least one of periodic and non-periodic intervals while the mobile robot is driving, and transmit the control command to the driving unit so that the mobile robot travels based on the calculated pose. Mobile robot with pose recognition. The method of claim 7, wherein And a dust collecting device for collecting dust on the bottom surface while inhaling air, wherein the bottom surface is cleaned. In the pose recognition method of the mobile robot which shows the global coordinate and the rotation degree of the said origin of the robot coordinate system which is a coordinate system whose origin is an origin, on the global coordinate system which is a preset two-dimensional coordinate system on the floor surface which is predetermined plane, Generating an image comprising at least two of the milestones of a plurality of milestones visually displayed with information including global coordinates of the points and attached to an arbitrary point on the bottom surface using a camera fixed on the mobile robot; step; Extracting a global coordinate of the milestone point from the milestone image included in the image; And Calculating the pose by calculating a relationship between the robot coordinates of the milestone point and the global coordinates of the milestone point calculated on the basis of the geometric relationship in which the milestone points are recognized in the robot coordinate system; Wherein the milestone, in displaying the information including the global coordinates of the point at least one of a bar code, a series of Arabic numerals, a character to display the pose of the mobile robot. delete The method of claim 9, Extracting image coordinates of the milestone image on the image coordinate system which is a two-dimensional coordinate system applied to the generated image; Robot coordinates of the milestone point in the step of calculating the pose is calculated based on a predetermined geometric relationship between the three-dimensional robot coordinate system and the image coordinate system. The method of claim 11, The robot coordinate system is formed by two axes on the bottom surface and the other one orthogonal to each other, the two axes on the bottom surface that pass through the origin and coincide with the traveling direction of the mobile robot and the direction perpendicular to the traveling direction, And an axis perpendicular to the traveling direction of the robot coordinate system and one axis of the image coordinate system are parallel to each other. The method of claim 12, When the relationship between the global coordinates of each milestone point and the robot coordinates of the milestone point is expressed by the following equation, And the pose is a global coordinate (P _X , P _Y ) of the origin of the robot coordinate system and is rotated by θ relative to the global coordinate system.

Here, the global coordinate of the milestone point is (x _w , y _w ), the robot coordinate is (x _r , y _r ). The method of claim 9, Repeating calculating the pose with at least one of periodic and aperiodic intervals while the mobile robot is traveling; And And moving the mobile robot based on the calculated pose.

Images