KR101141601B1 - Quick map building apparatus based on vision natural landmark - Google Patents

Abstract

The present invention relates to an image-based natural marker mapping apparatus and method thereof, and more particularly, it relates to an image-based natural marker mapping apparatus and method thereof that calculates a position using a rotation number of a wheel, measures distance information through a laser sensor, extracts a natural marker through a vision sensor, By creating a map to be used for the estimation technique, a moving object including a mobile robot can quickly and accurately create an image-based natural marker for recognizing the surrounding environment in a public facility, a hospital, a factory or the like. In addition, the present invention provides an image-based natural marker mapping apparatus and a method thereof, allowing a user to draw a cart-shaped moving object, create a map based on physical dimensions and a natural marker map based on the image, Based natural marker map can be saved, and there is an advantage that it can be utilized for estimating the exact position of a moving object such as a mobile robot.

Mapping, natural markers, vision sensors, mobile robots

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to an image-based natural marker mapping apparatus,

The present invention relates to an image-based natural marking apparatus and method thereof, and more particularly, to an image-based natural marking apparatus and method thereof, And a device capable of creating a map for utilization in a position estimation technique of a mobile robot.

The present invention was derived from research carried out by the Ministry of Knowledge Economy and the Institute of Information and Communication Technology as a part of the IT Growth Driving Technology Development Project [Task Number: 2005-S-092-03, Project Name: USN-based Ubiquitous Robotic Space Technology Development ].

As is well known, as the interest in robot technology increases, the utilization rate thereof is gradually diversified. In particular, autonomous mobile robots are widely used in various fields such as helping handicapped persons, transferring logistics from factories, space exploration work, nuclear waste disposal plant, or work in a dangerous environment such as deep sea. We are performing on behalf of human beings.

Such an autonomous mobile robot replaces dangerous environment work, and can be used as a substitute for non-dangerous work such as an unmanned vacuum cleaner or an unmanned lawn mower. Therefore, It will give the company a high value-added market due to its mass industrialization.

Until now, autonomous mobile robots have been inferior to humans in their cognitive abilities and reasoning abilities, and their functions are very limited. Therefore, researches have been conducted to improve the intelligence of robots in many fields.

Here, the autonomous mobile robot must be capable of moving to a desired destination without collision, and in order to move freely in the living space, it is necessary to have an ability to determine where his / her own position is. These functions are performed by position measurement techniques and mapping techniques. These two functions are very related to each other like the relationship between 'eggs and chickens'. If the map is erroneously written, the positioning performance based on it will be significantly decreased. If the map is created based on the position, The error will become even bigger. This can result in satisfying the other one technique if it is possible to provide the robots with satisfactory performance even if one of the accurate mapping technique or the position measurement technique is satisfied.

Recently, researchers and developers are mainly studying the accurate location of robots by providing accurate maps to robots. Depending on the type of map provided at this time, methods of using artificial markers and using natural markers Respectively.

First, the method of using the artificial markers in the form of a map is as follows. First, a mounting material recognizable by the robot is installed in the surrounding environment, and the initial position information is provided to the robot. When the robot subsequently recognizes the artificial marker, In this method, the position of the robot is precisely estimated when the robot detects the artificial marker because the position of the artificial marker is precisely known in advance.

Next, the method of using the natural markers in the form of the map is to use the markers characteristic of the normal environment without additional artificial markers in the environment that the robot travels, Simultaneous localization and mapping (SLAM) technology. The purpose of SLAM technology is to store the natural markers in a map form while estimating the accurate map and position by estimating their position at the same time as the robot travels the unknown environment by itself.

However, the method of using the artificial marker in the form of the map according to the prior art as described above requires a lot of manpower to attach the artificial marker and requires a lot of time and cost since the position of the initial artificial marker must be calculated in advance There is a problem that it consumes. In addition, since the method using the natural markers in the form of the map is a robot having limited intelligence, it is difficult to judge what is the characteristic of the environment, and thus there is a problem that the strong natural markers can not be extracted.

SUMMARY OF THE INVENTION Accordingly, the present invention has been made keeping in mind the above problems occurring in the prior art, and it is an object of the present invention to provide a method and apparatus for calculating a position where a moving object is moved by the number of revolutions of a wheel, Dimensional map based on the created physical dimensions, coordinates of the image-based map are created, and the coordinates of the image-based map are created by extracting the natural markers derived from the robust features that represent the information of the environment through the vision sensor, Based map marking device capable of generating an indoor map for use in a position estimation technology of a mobile robot by synchronizing the position of the mobile robot with the position of the mobile robot.

According to an aspect of the present invention, there is provided an apparatus for mapping an image-based natural markers, comprising: a wheel encoder for measuring the number of revolutions of a plurality of wheels, a laser sensor for detecting distance information, and a vision sensor for detecting image information A moving object detection information processing unit for calculating a moving position of the moving object based on the number of wheel rotations measured by the wheel encoder and extracting an image based natural marker from the detected image information; Based map by using the calculated movement position and the sensed distance information, and creating an indoor map by coordinate-synchronizing the created physical-dimension-based map and the image-based map created by the natural markers Wherein the moving object detection information processing unit detects the moving object based on the measured rotation numbers, A distance information acquisition unit for acquiring the sensed distance information, an image information acquisition unit for acquiring the sensed image information, and a display unit for acquiring image-based natural markers from the sensed image information, And a natural marker extracting unit for extracting the natural markers.

delete

In the present invention, the position is calculated by the number of revolutions of the wheel, the distance information is measured through the laser sensor, the natural marker is extracted through the vision sensor, and the map is utilized for the position estimation technology of the mobile robot. A moving object including a mobile robot can quickly and accurately create an image-based natural marker for recognizing the surrounding environment in a public facility, a hospital, a factory, and the like.

In addition, the present invention provides an image-based natural marker mapping apparatus and method, wherein a user draws a moving object (e.g., a cart-shaped moving object) and creates a map based on physical dimensions and a natural marker map based on the image, It is possible to save the time required for constructing the image-based natural marker map and can be used for estimating the exact position of the moving object such as the mobile robot. In particular, if the mobile robot is used in a mobile robot used for various services in various other indoor spaces such as public facilities, the mobile robot can accurately estimate its own position.

Hereinafter, the operation principle of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. The following terms are defined in consideration of the functions of the present invention, and these may be changed according to the intention of the user, the operator, or the like. Therefore, the definition should be based on the contents throughout this specification.

FIG. 1 is a view showing a cart-shaped mobile device for mapping an image-based natural marking according to an embodiment of the present invention. In this cart-shaped mobile device, The first and second wheels S1 and S2 are used to measure the distance and direction in which the first wheel S1 and the second wheel S2, the third wheel S3 located on the rear wheel side, A wheel encoder 11 for measuring the number of revolutions of the wheel, a laser sensor 13 for detecting the distance to objects in the surrounding environment, and an image sensor And a vision sensor 15 mounted for this purpose. It is preferable that the wheel encoders 11 are mounted in the same number as the wheels so as to correspond to the first and second wheels S1 and S2 positioned on the front wheel side.

FIG. 2 is a block diagram of an image-based natural marker mapping apparatus according to an exemplary embodiment of the present invention, which includes a moving object hardware unit 10, a moving object detection information processing unit 30, a map creating unit 50, .

The moving body hardware unit 10 may include a wheel encoder 11, a laser sensor 13, a vision sensor 15, and the like as a moving object sensing unit as shown in FIG.

The wheel encoder 11 measures the number of rotations of the first wheel S1 and the second wheel S2 and then outputs the measured number of rotations of the first wheel S1 to the moving object detection information processing unit 30, And supplies the measured rotation number of the second wheel S2 to the rotation number acquisition unit 33 of the second wheel in the moving object detection information processing unit 30 .

The laser sensor 13 detects the distance to objects in the surrounding environment by using a laser and provides the detected distance information to the distance information obtaining unit 37 in the moving object detection information processing unit 30. [

The vision sensor 15 senses an image while being moved by the user and provides the sensed image information to the image information acquisition unit 39 in the moving object detection information processing unit 30. [

4, the moving object detection information processing unit 30 includes a first wheel speed obtaining unit 31, a second wheel speed obtaining unit 33, a position calculating unit 35, a distance information obtaining unit 37, an image information obtaining unit 39, a natural marker extracting unit 41, and the like.

The first wheel rotational speed obtaining unit 31 obtains the rotational speed of the first wheel S1 input from the wheel encoder 11 in the moving body hardware unit 10 and provides the rotational speed of the first wheel S1 to the position calculating unit 35.

The second wheel rotational speed acquiring unit 33 acquires the rotational speed of the second wheel S2 input from the wheel encoder 11 in the moving body hardware unit 10 and provides the rotational speed of the second wheel S2 to the position calculating unit 35. [

The position calculating unit 35 calculates the position of the first and second wheels S1 and S2 based on the rotational speeds of the first and second wheels S1 and S2 input from the first wheel rotational speed obtaining unit 31 and the second wheel rotational speed obtaining unit 33, And provides the calculated position to the physical dimension-based map unit 51 in the map creating unit 50. [

The distance information obtaining unit 37 obtains the distance information inputted from the laser sensor 13 in the moving body hardware unit 10 and provides the obtained distance information to the physical dimension based map unit 51 in the map creating unit 50.

The image information obtaining unit 39 obtains image information input from the vision sensor 15 in the moving body hardware unit 10 and provides the image information to the natural marker extracting unit 41.

The natural marker extracting unit 41 extracts the natural marker image from the image information input from the image information acquiring unit 39 by using image information of a strong characteristic that represents the information of the environment, And supplies the extracted image information to the image-based guidance unit 53 in the map-making unit 50. The image-

The map creating unit 50 may include a physical dimension based map unit 51, an image based map unit 53, a coordinate synchronizing unit 55 and an indoor map creating unit 57 as shown in FIG.

The physical dimension-based guidance unit 51 uses the position of the moving object inputted from the position calculation unit 35 in the moving object detection information processing unit 30 and the distance information inputted from the distance information obtaining unit 37, I.e., a raster map, a vector map, a topology map, and the like, and supplies the generated coordinates to the coordinate synchronization unit 55. [

The image-based guidance unit 53 creates a database based on the image-based natural markers input from the natural marker extracting unit 41 in the moving object detection information processing unit 30, and provides the image-based map to the coordinate synchronization unit 55.

The coordinate synchronization unit 55 synchronizes the physical dimension based map input from the physical dimension based map unit 51 with the image based map including any natural markers inputted from the image based map unit 53, And supplies the synchronized coordinates to the indoor map creator 57 so that the physical location information of the image including the virtual space exists.

The indoor map creating unit 57 creates an indoor map using the physical location information of the image including the arbitrary natural markers synchronized with the coordinates inputted from the coordinate synchronizing unit 55. [

Accordingly, in the present invention, the position is calculated by the number of revolutions of the wheel, the distance information is measured through the laser sensor, the natural marker is extracted through the vision sensor, and the map is utilized for the position estimation technique of the mobile robot, A moving object including a mobile robot can quickly and accurately create an image-based natural marker for recognizing the surrounding environment in a space (for example, a public facility, a hospital, a factory, etc.).

Next, an image-based natural marker mapping process will be described in an embodiment of the present invention having the above-described configuration.

FIG. 6 is a flowchart sequentially illustrating an image-based natural marker mapping method according to an embodiment of the present invention.

First, the user must draw environment information on a cart-shaped mobile device equipped with a laser sensor 13 and a vision sensor 15 as shown in FIG.

That is, in the wheel encoder 11 mounted on the cart-shaped mobile device, the number of rotations of the first wheel S1 and the second wheel S2 mounted on the mobile device is measured (S601) The measured rotation number of the first wheel S1 is provided to the rotation number acquisition unit 31 of the first wheel in the moving object detection information processing unit 30 and the rotation number of the measured second wheel S2 And is provided to the rotation speed acquiring unit 33 of the second wheel in the processing unit 30. [

The first wheel rotation speed acquiring unit 31 acquires the rotation speed of the first wheel S1 input from the wheel encoder 11 in step S603 and provides the rotation speed to the position calculating unit 35 in step S604, The second wheel rotation speed acquisition unit 33 acquires the rotation speed of the second wheel S2 input from the wheel encoder 11 in step S605 and provides the rotation speed to the position calculation unit 35 in step S607.

The position calculating unit 35 calculates the rotational speeds of the first and second wheels S1 and S2 input from the rotational speed acquiring unit 31 of the first wheel and the rotational speed acquiring unit 33 of the second wheel, (S609), and the calculated position of the mobile body is provided to the physical dimension-based guidance unit 51 (S611).

The laser sensor 13 detects the distance to objects in the surrounding environment using a laser (S613), and provides the detected distance information to the distance information obtaining unit 37. [ Then, the distance information obtaining unit 37 obtains the distance information inputted from the laser sensor 13 (S615) and provides the distance information to the physical dimension-based guidance unit 51 (S617).

Based on the position at which the moving object is input from the position calculation unit 35 and the distance information input from the distance information acquisition unit 37, the physical dimension-based map unit 51 generates a map based on a water rich number, (S619), and provides them to the coordinate synchronizing unit 55 (S621).

In the vision sensor 15, an image is sensed while being moved by a user (S623), and the sensed image information is provided to the image information acquiring unit 39. [ Then, the image information obtaining unit 39 obtains the image information input from the vision sensor 15 (S625) and provides the image information to the natural marker extracting unit 41 (S627).

The natural marker extracting unit 41 extracts image information of a strong characteristic that is representative of environment information from the image information input from the image information obtaining unit 39 as a natural marker (S629), extracts the natural marker To the image-based guidance unit 53 (S631). Here, the natural markers can be extracted by SHIFT, SIFT, principal components analysis SIFT, speed up robust feature (SURF), gradient location and orientation histogram (GLOH), local energy based shape histogram (LESH) Detection of canine, canny-deriche, differential, sobel, corner detection, ridge detection, blob detection (laglacian of gaussian (LoG), difference of gaussians (DoG), determinant of hessian (DoH) In one embodiment, an image feature point extraction algorithm such as SHIFT is used to judge whether the image is well characterized by the number of feature points extracted from the image information and the shape of the feature point, etc. The natural markers can be extracted.

The image-based map unit 53 creates an image-based map by converting the natural markers inputted from the natural marker extracting unit 41 into a database and creates an image-based map (S633). The image-based map including the created arbitrary natural markers And supplies it to the synchronization unit 55 (S635).

The coordinate synchronization unit 55 synchronizes the physical dimension based map input from the physical dimension based map unit 51 with the image based map including any natural mark input from the image based map unit 53, (S637) so that the physical location information of the image exists, and the physical location information of the image including the synchronized arbitrary natural mark is stored in the indoor map creating unit 57 (S639).

The indoor map creator 57 uses the physical location information of the image including the arbitrary natural markers synchronized with the coordinates inputted from the coordinate synchronizer 55 to display the image based nature When it is judged that the mark is formed by the image-based map sufficiently, the indoor map is created (S641). Here, the indoor map created by the indoor map creating unit 57 can be applied to the position estimation technology of the robot or the moving body in the future. For example, when a mobile robot is kidnapped, its position can be estimated when a natural marker registered in an image-based map matches a current image of the vision sensor.

The image-based natural marker mapping method according to the present invention can be realized by a computer program. The code and code segments that make up this computer program can be easily deduced by a computer programmer in the field. In addition, the computer program is stored in a computer-readable medium and readable and executed by a computer to implement an image-based natural marker mapping method. The information storage medium includes a magnetic recording medium, an optical recording medium, and a carrier wave medium.

As described above, the present invention provides an image-based natural marker mapping apparatus and method, wherein a user draws a cart-shaped moving object, creates a map based on physical dimensions and a natural marker map based on the image, It is possible to save the time required for constructing the image-based natural marker map, and can be used for estimating the exact position of the moving object such as a mobile robot. In particular, if the mobile robot is used in a mobile robot used for various services in various other indoor spaces such as public facilities, the mobile robot can accurately estimate its position.

The present invention has been described with reference to some embodiments thereof. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. Therefore, the disclosed embodiments should be considered in an illustrative rather than a restrictive sense. The scope of the present invention is defined by the appended claims rather than by the foregoing description, and all differences within the scope of equivalents thereof should be construed as being included in the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a cart-type mobile device for mapping an image-based natural markers according to an embodiment of the present invention;

FIG. 2 is a block diagram of an image-based natural marker mapping apparatus according to an embodiment of the present invention;

FIG. 3 is a block diagram showing the moving body hardware unit shown in FIG. 2 in more detail;

FIG. 4 is a block diagram showing the moving object detection information processing unit shown in FIG. 2 in detail;

FIG. 5 is a block diagram showing in detail the map creating unit shown in FIG. 2,

FIG. 6 is a flowchart sequentially illustrating an image-based natural marker mapping method according to an embodiment of the present invention; FIG.

Description of the Related Art

10: Moving body hardware part 11: Wheel encoder

13: Laser sensor 15: Vision sensor

30: moving object detection information processing unit 31: rotational speed acquiring unit of first wheel

33: second wheel speed acquiring unit 35: position calculating unit

37: Distance information acquisition unit 39: Image information acquisition unit

41: Natural marker extracting unit 50: Mapping unit

51: Physical Dimension Based Leadership Unit 53:

55: coordinate synchronizer 57: indoor map maker

Claims (10)

A moving object detection unit including a wheel encoder for measuring a rotation number of each of a plurality of wheels, a laser sensor for detecting distance information, and a vision sensor for detecting image information, A moving object detection information processing unit for calculating a moving position of the moving object based on the number of wheel rotations measured by the wheel encoder and extracting an image based natural marker from the detected image information, A map creating unit for creating a physical map based on the calculated moving position and the detected distance information and creating an indoor map by coordinate-synchronizing the created physical map based on the created physical map with the image- ≪ / RTI & Wherein the moving object detection information processing unit comprises: A position calculating unit for calculating a position at which the moving body moves based on the measured rotation speed, A distance information obtaining unit for obtaining the detected distance information, An image information acquisition unit for acquiring the sensed image information; And a natural marker extracting unit for extracting an image-based natural marker from the sensed image information Image based natural marking device. delete The method according to claim 1, Wherein the wheel encoders are mounted in the same number as the wheels so as to correspond to wheels located on the front wheel side of the moving body. delete The method according to claim 1, Wherein, A physical dimension-based map unit for creating a physical dimension-based map using the calculated position and the sensed distance information; An image-based map unit for creating an image-based map by converting the extracted image-based natural markers into a database, And an indoor map creating unit for creating an indoor map by coordinate-synchronizing the created physical-dimension-based map and the image-based map Image based natural marking device. 6. The method of claim 5, The natural markers, SHIFT, SIFT, PCA-SIFT, SIFF, gradient location and orientation histogram (GLOH), local energy based shape histogram (LESH), edge detection (canny, canny-deriche, differential) , the image extracted by at least one of corbel detection, corner detection, ridge detection, blob detection (laplacian of gaussian (LoG), difference of gaussians (DoG), determinant of hessian (DoH) Based natural marker mapping device. 6. The method of claim 5, Wherein the indoor- Wherein the indoor map is created when the image-based natural markers are created in the image-based map. delete delete delete

Images