KR20110078544A - Motion implementing method of vision system grafted legged mobile robot - Google Patents

Abstract

PURPOSE: A motion performing method of a vision-system-associated walking robot is provided so that if characters recognized by the camera of a walking robot match those stored in an image processing board, then corresponding commands are output and corresponding motion can be performed. CONSTITUTION: A motion performing method of a vision-system-associated walking robot is as follows. A character recognized by the camera of a walking robot is compared with that stored in an image processing board(S100,S110). If the two characters match, a vision board outputs a corresponding command to a motion board(S120). A program built in the motion board outputs a driving signal for driving each motor(S130). A robot performs corresponding motion with the driving of the motor(S140).

Description

MOTION IMPLEMENTING METHOD OF VISION SYSTEM GRAFTED LEGGED MOBILE ROBOT}

The present invention relates to a method for performing a motion of a vision system-type pedestrian robot, and more particularly, to a motion system for performing a vision system-type pedestrian robot using a vision system.

In general, a walking robot has 20 to 25 motors, which act as human joints, to be connected to each other to input movement displacement and speed, respectively, to realize human-like movement. This means that humans don't walk by any method when walking, but robots connect each other at the appropriate moments and play them back to enable human-like movement. In other words, when walking, one motion (step 1) that tilts in the opposite direction while lifting one foot and two motions (step 2) can be generated by extending the opposite foot at the same time as the other foot, You can create two symmetrical motions by pushing again. Here, you can input the speed to one motion (each step) to create the speed to execute the step to show the naturalness of the motion. If you play with), the robot walks. When creating an animation, it's the same as connecting each frame and playing it. The other important thing is that the speed of executing the next frame at different speeds is different.

These walking robots literally start from walking robots and require the addition of various motions, but the hardware and software become complicated to solve them. However, the development of the robotics industry requires the addition of more natural and diverse motion.

The present invention has been made to solve the above problems, the object of which is to recognize the character with a camera installed on the walking robot and compare with the character stored in the image processing board and outputs a corresponding command if the corresponding motion to perform the corresponding motion Therefore, the present invention provides a method of performing a motion of a vision system-integrated pedestrian robot that can realize an intelligent pedestrian robot.

In addition, another object of the present invention, the motion of the walking robot can be performed by performing arithmetic, writing, drawing, playing, dancing, etc. of the vision system grafted walking robot that can add a more natural and various functions The present invention provides a method for performing motion.

In order to achieve the above object, the present invention includes a first step of comparing the character recognized by the camera installed in the walking robot with the character stored in the vision board for imaging; A second step of outputting a command corresponding to the vision board by the vision board if the two letters compared are the same; A third step of outputting a driving signal for driving each motor by a program embedded in the motion board; And a fourth step in which the walking robot performs a corresponding motion by driving the motor.

In addition, the second step of the present invention is characterized by recognizing the presented number and the operation symbol through the vision and direct operation.

In addition, the second step of the present invention is characterized by recognizing the characters presented through the vision and writing directly.

In addition, the second step in the present invention is characterized in that the self-recognized and directly draw the picture presented using the vision.

In addition, the second step in the present invention is characterized by recognizing the presented sheet music using a vision and playing the instrument directly, while producing a corresponding sound in the state of recognizing the scale in advance.

In addition, the second step of the present invention is characterized in that the character of the song title or singer name is presented using a vision, the dance operation is performed in a state in which the dance operation is recognized in advance.

In addition, the character recognition using the vision in the present invention, the step of horizontally converting the character string that is the input image; Enlarging or reducing the size of an image such that a character string input from the camera is equal to a size of a preset character string; And recognizing a character by template matching by comparing the input image with an image database stored in the vision board.

Such a motion performance method of the vision system grafted walking robot of the present invention, after recognition of the character with a camera installed in the walking robot, compared with the character stored in the image processing board and outputs a corresponding command if the match is performed to perform the motion An intelligent robot can be implemented, and the motion of the walking robot can be performed by performing calculations, writing, drawing, playing, dancing, etc., thereby adding more natural and various functions.

Hereinafter, with reference to the accompanying drawings, a method for performing a motion of the vision system grafted walking robot of the present invention will be described with reference to the accompanying drawings.

According to an exemplary embodiment of the present invention, a motion performing method of the vision system-type pedestrian robot may include a character recognition step (S100), a character comparison step (S110), a command output step (S120), and a driving signal. An output step S130 and a motion performing step S140 are included.

The character recognition step (S100) is a step of recognizing a character through a camera (CCD camera) installed in the walking robot. At this time, the character recognition using the vision, as shown in FIG. 2 includes the step of horizontally converting the character string (S102), the zooming (resizing) step (S104) and the character recognition step by template matching (S106) do.

In particular, similar character recognition errors when recognizing characters including numbers change the font of the numbers and finely adjust the correlation coefficient so that the recognition difference between each number can be accurately recognized, and the character recognition error due to the difference in illumination is binarized. Adjust the coefficients of the technique and find the illuminance at the time when the recognition is best, so that the original letter shape is recognized as it is.

In the step of converting the character string horizontally (S102), the height of the character string is calculated while rotating the character image of the image received from the camera. In this case, the character string becomes horizontal when the height of the character string is minimum. The method of calculating the height of a character string is obtained by obtaining a line histogram and detecting a region in which the line histogram is larger than a threshold.

The enlargement / reduction step (S104) enlarges / reduces the size of the input image at a constant ratio such that the size (width × length) of the character string input from the camera is equal to the size of the character string stored in the embedded board as a vision. .

In the character recognition step S106 of the template matching, an image (character) input from the camera is recognized by comparing the image input from the camera with an image database stored in the non-dedicated embedded board. At this time, the walking bots perform operations (compute, write, draw, play and dance) based on the recognized results.

The character comparing step S110 is a step of comparing the recognized character as shown in FIG. 1 with a character stored in an image vision board by a program.

The command output step (S120) is a step of comparing a recognized character with a character stored in the vision board for an image, and if the two characters are the same, the vision board outputs a command corresponding to the motion board. In this case, the vision board uses an embedded board. (See FIG. 3)

The driving signal output step (S130) is a step in which a program embedded in the motion board outputs a driving signal to drive a motor installed for each moving part of the walking robot.

The motion performing step (S140) is a step in which the walking robot performs the corresponding motion by driving the motor, and it is desirable to implement an efficient and smooth movement by finding an appropriate displacement ratio and speed of each motor.

On the other hand, the motion performed by the walking robot through the step (S100 ~ S130) is performing arithmetic, writing, drawing, playing and dancing.

That is, the calculation of the motion performed by the walking robot recognizes the presentation number and the operation code input from the camera as shown in FIG. 4A through the vision board and directly calculates the result by the program, and then fixes it on the walking robot hand. Will be marked with a pen.

And writing is to recognize the presentation characters input from the camera through the vision board as shown in Figure 4b and to write with a pen fixed to the walking robot hand by a program, the central processing unit (CPU of the control unit) ) Recognizes each consonant and vowel and commands them to write the corresponding article and implements the action.

And the drawing is a self-recognized presentation picture input from the camera as shown in Figure 4c through the vision board to draw a picture directly with a pen fixed to the walking robot hand, unlike the above writing already built It recognizes the picture itself, not the program process, and implements the action immediately.

And as shown in Figure 4d, the musical instrument input from the camera through the vision board to recognize the presented score, and fixed the bar to the walking robot hand to play the instrument directly, the control in a similar manner to the writing above The CPU recognizes the scale and makes the corresponding sound.

In addition, as shown in FIG. 4E, the dancing recognizes the written song title or the name of the singer name input from the camera through the vision board, but implements the dance operation in a state in which the dance movement is recognized in advance.

At this time, when the song title or the singer's name is recognized, the dancing is a walking robot that dances according to the program because the motion corresponding to the character, that is, the same choreography according to the song is programmed.

For example, the decision items when setting a program for the motion of the walking robot are as follows.

① Servo motor number: Adopted program uses hexadecimal number. Motor can control 25 motors at the same time up to 18 with 0 as the first motor.

② Motion data: The rotation position of the motor can be controlled within the range of hexadecimal 0000 to 03ff. 7 steps from 0 to 6 can be controlled simultaneously on one screen.

③ Page setup window: You can see the current page number and pages are from 0 to 7f in hexadecimal. The main usage menus are Toque, PlayCount, PageStep, NextPage, and ExitPage, which will give you a percentage of torque, and you can decide how many steps are used and how many steps are used in the current page. You can then use Nextpage and ExitPage to enter the current page and go to or from which page.

④ Message output window: Displays the response after the command of the current page to detect an error or the user to know the current status.

⑤ Command Input Window: To use the program, the user can enter a command.

⑥ Communication setting window: Enter the port number and the communication speed to connect the computer and the board of the walking robot.

⑦ Mode selection button: This motion edit can be used to control the walking robot with a computer or remote control. It is a distinguishing button for this control and helps to save and manage the total pages and steps of user input values.

Furthermore, a dance motion of a walking robot using motion edit is described as an example.

① Dance after standing in the water, return to place: put your hands on the floor → raise your legs → store your strength → quickly bounce your arms and back → return to your home position.

Here, the hand on the floor is made by using the centrifugal force to raise the leg by using the momentary leg and the reaction of the hand as a preparation operation for standing up as shown in Figure 6a.

And lift the leg as shown in Figure 6b exerts a momentary force by using centrifugal force to raise the leg perpendicular to the floor and bounce the body to the beat using the motor of the hand and leg.

And to store the force is a preparatory action to return to its original position as shown in Figure 6c to slowly store the force for rotating the body close to the ground.

Quickly arm and back bounces and centers the foot in contact with the ground with the reserve force and centrifugal force generated by stretching the arms and legs (waist) at a rapid rate as shown in FIG. 6D.

In the detailed description of the present invention described above with reference to the preferred embodiment of the present invention, the scope of protection of the present invention is not limited to the above embodiment, and those skilled in the art of the present invention It will be understood that various modifications and variations can be made in the present invention without departing from the spirit and scope of the invention.

1 is a block diagram illustrating a method of performing a motion of a vision system-grafted walking robot according to the present invention.

2 is a block diagram showing a character recognition sequence using vision in the method for performing a motion of the vision system-grafted walking robot.

Figure 3 is a photograph showing an embedded board applied to the motion performance method of the vision system grafted walking robot.

4A to 4E are photographs showing the motion of the walking robot according to the motion performing method of the vision system-grafted walking robot.

5 is a photograph illustrating a program by a motion performing method of the vision system grafted walking robot.

6a to 6d is a motion picture of a walking robot showing a dance motion by the motion performing method of the vision system grafted walking robot.

Claims (7)

A first step of comparing a character recognized by a camera installed in the walking robot with a character stored in an image vision board; A second step of outputting a command corresponding to the vision board by the vision board if the two letters compared are the same; A third step of outputting a driving signal for driving each motor by a program embedded in the motion board; And And a fourth step in which the robot performs the corresponding motion by driving the motor. The method of claim 1, The second step is to perform the motion of the vision system-type pedestrian robot, characterized in that to recognize the presented number and the operation symbol through the vision and directly calculate. The method of claim 1, The second step is to perform the motion of the vision system-type pedestrian robot, characterized in that the recognition of the presented characters through the vision and write directly. The method of claim 1, The second step is to perform the motion of the vision system-type pedestrian robot, characterized in that the self-recognition of the presented picture using the vision and draw the picture directly. The method of claim 1, The second step is to perform the motion of the vision system-type pedestrian robot, characterized in that by using the vision to recognize the presented sheet music and play the instrument directly, the corresponding sound in the state in advance to recognize the scale. The method of claim 1, The second step is to recognize the character of the song title or singer name presented using a vision, the motion performance method of the vision system grafted walking robot, characterized in that performing the dance operation in a state in which the dance movement is recognized in advance . The method according to any one of claims 3 to 6, Character recognition using the vision, Horizontally converting a character string that is an input image; Enlarging or reducing the size of an image such that a character string input from the camera is equal to a size of a preset character string; And Recognizing characters by template matching comparing the input image and the image database stored in the vision board;

Images