KR101104225B1 - Method and Apparatus for Storing Map Data for Robot - Google Patents

Abstract

Embodiments of the present invention relate to a method and apparatus for storing terrain data for a robot.

An embodiment of the present invention, the edge point extraction unit for extracting the edge point from the grid map, the Voronoi diagram generator for generating a Voronoi diagram based on the edge point, a node search unit for searching for a node from the Voronoi diagram, and the node The present invention provides a terrain data storage device for a robot including a terrain data generation storage unit for generating terrain data by scanning the periphery of the space and storing the generated terrain data.

According to an embodiment of the present invention, while allowing the robot to correctly recognize the space or position of the robot so as to allow normal operation of the robot, it is possible to efficiently acquire the terrain data by acquiring only the terrain data for a point necessary for the external terrain. As a result, it is possible to reduce the storage space by efficiently storing the acquired terrain data.

Robot, terrain data

Description

Method and apparatus for storing terrain data for robots {Method and Apparatus for Storing Map Data for Robot}

Embodiments of the present invention relate to a method and apparatus for storing terrain data for a robot. More specifically, while precisely recognizing the space or position of the robot to enable the normal operation of the robot, it is possible to efficiently acquire the terrain data by acquiring only the terrain data for the point necessary for external terrain identification, The present invention relates to a method and apparatus for storing terrain data for a robot that efficiently stores terrain data to reduce storage space.

 As a technology for space recognition and location recognition of a robot, a lot of research is being conducted on a technology for generating a map and simultaneously performing a location recognition in a robot's unknown environment. This technique is called SLAM (Simultaneous Localization and Mapping).

As described above, the conventional technique for space recognition and location recognition of the robot acquires and stores terrain data about an external environment using a sensor or the like, analyzes the terrain data to create a map of the surrounding space, Based on this, the location of the robot is estimated on the map.

However, in the conventional technology for space recognition and location recognition of the robot, when the space recognition or location recognition of the robot, the terrain data obtained in order to generate the map by grasping the external terrain is necessary to identify the external terrain (eg : Acquisition of terrain data that can include not only terrain data for areas with heavy terrain changes, such as crossroads and doors, but also terrain data for unnecessary points (e.g., long-distance changes such as long passages) that do not need to be acquired. There is a problem of inefficiency. In addition, the conventional technique described above also has a problem of requiring a large amount of storage space for storing the acquired terrain data in order to more accurately recognize the space or position recognition of the robot.

In this background, an object of the embodiment of the present invention, while accurately performing the spatial recognition or location recognition of the robot to enable the normal operation of the robot, while obtaining only the topographical data for the point necessary for the external topography, efficient terrain data Is to enable acquisition.

In addition, another object of the embodiment of the present invention is to reduce the storage space by efficiently storing the terrain data obtained for spatial recognition or location recognition of the robot.

One embodiment of the present invention, the edge point extraction unit for extracting the edge point from the grid map; A Voronoi diagram generator for generating a Voronoi diagram based on the edge point; A node searching unit searching for a node from the Voronoi diagram; And a terrain data generation storage unit configured to scan the periphery of the node to generate terrain data and to store the generated terrain data.

In addition, an embodiment of the present invention, the edge point extraction step of extracting the edge point from the grid map; Generating a Voronoi diagram based on the edge point; A node searching step of searching for a node from the Voronoi diagram; And generating terrain data by scanning the periphery of the node, and generating and storing the generated terrain data.

As described above, according to an exemplary embodiment of the present invention, while the space recognition or position recognition of the robot is accurately performed so that the normal operation of the robot can be performed, the terrain data is efficiently obtained by obtaining only the terrain data for the point necessary for the external terrain. There is an effect that enables the acquisition of.

In addition, according to an embodiment of the present invention, by effectively storing the terrain data obtained for spatial recognition or location recognition of the robot, there is an effect of reducing the storage space.

Hereinafter, some embodiments of the present invention will be described in detail through exemplary drawings. In adding reference numerals to the components of each drawing, it should be noted that the same reference numerals are assigned to the same elements as much as possible even though they are shown in different drawings. In the following description of the present invention, a 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.

In addition, in describing the component of this invention, terms, such as 1st, 2nd, A, B, (a), (b), can be used. These terms are only for distinguishing the components from other components, and the nature, order or order of the components are not limited by the terms. If a component is described as being "connected", "coupled" or "connected" to another component, that component may be directly connected to or connected to that other component, but there may be another configuration between each component. It is to be understood that the elements may be "connected", "coupled" or "connected".

1 is a block diagram of a terrain data storage device 100 for a robot according to an embodiment of the present invention.

As shown in FIG. 1, the terrain data storage device 100 for a robot according to an embodiment of the present invention includes an edge point extractor 110, a Voronoi diagram generator 120, and a node searcher 130. ) And the terrain data generation storage 140.

The edge point extractor 110 extracts an edge point from a grid map. The Voronoi diagram generator 120 generates a Voronoi diagram based on the extracted edge point. The node search unit 130 searches for a node from the generated Voronoi diagram. The terrain data generation storage 140 scans the periphery of the found node to generate terrain data and stores the generated terrain data.

Since it is difficult to generate a Voronoi diagram directly from the grid map, in an embodiment of the present invention, an edge point is extracted from the grid map and a Voronoi diagram is generated using the extracted edge point.

As described above, in order to more easily generate a Voronoi diagram, the above-described edge point extractor 110 extracting an edge point from a grid map is an example of edge point extraction, which is a pre-stored case of edge point case. ), We can extract the edge point from the grid map.

After the edge points are extracted, the Voronoi diagram generator 120 generates the Voronoi diagram using the extracted edge points. The Voronoi diagram generated as described above may be quite complicated. If the generated Voronoi diagram is very complex, it is difficult to obtain necessary information or data (eg, terrain data) from the complex Voronoi diagram. Thus, processing using multiple conditions can produce simplified Voronoi diagrams that fit the actual terrain.

As an example of processing a complex Voronoi diagram and processing it into a simplified Voronoi diagram suitable for real terrain, the Voronoi diagram generator 120 generates a plurality of points based on the extracted edge points, and removes the points. According to the method, one or more points are removed from the generated points to connect the remaining points to generate a connection path, and a Voronoi diagram including the generated connection path can be generated.

The above-mentioned point removal method may include one or more of a method of removing a point larger than a grid map, a method of removing a point occurring inside a wall of the grid map, a method of removing a duplicate point, and the like. .

In addition, the above-mentioned connecting path includes lines passing through the center of the path on the grid map. Since the robot makes a path that does not collide with an obstacle such as a wall, as described above, the connection path including the lines passing through the center of the path is similar to the movement path through which the robot moves. In addition, the connection path can recognize the terrain of the map or find out information about the terrain.

After the Voronoi diagram is generated by the Voronoi diagram generation unit 120, the node search unit 130 searches for a point that can proceed in three or more directions in the Voronoi diagram as a node.

The searched node means the place where the topographical change is severe, such as the point of intersection or door, and among all the points of the whole map, it is the point that best reflects the topography of the whole map and most accurately for the normal operation of the robot. It is also the point to be recognized. This is because in the long passage where there is almost no change in the terrain, even if the terrain data does not exist, the robot's sensor can be used to know its direction online so that normal driving is possible. If there is not enough terrain data in the node, the robot may not be able to operate normally.

As described above, after the node is searched for in the Voronoi diagram, the terrain data generation storage 140 generates and stores the terrain data through a scan of the surrounding of the found node.

As an example of generation and storage of terrain data, the terrain data generation storage unit 140 scans the periphery of the retrieved node to recognize surrounding obstacles, obtains geographic information on the recognized surrounding obstacles, and obtains the obtained geographic information. Create and store terrain data based on

The above-mentioned geographic information on the surrounding obstacles includes the position information of the recognized surrounding obstacles and the distance information between the recognized surrounding obstacles and the robot, wherein the recognized surrounding obstacles are fixed obstacles such as walls, or meet when the robot moves. May be a moving obstacle, such as a moving object.

In the case of the terrain data generated by the terrain data generation and storage unit 140 described above, since similar information is displayed in many points, data loss and information amount may be degraded. Accordingly, the terrain data generation storage 140 may generate simplified terrain data in order to have a large amount of information with a small amount of data.

As an example, the terrain data generation storage 140 may scan the surrounding of the searched node to recognize the surrounding obstacle, obtain geographic information about the recognized surrounding obstacle, and based on the acquired geographic information, the recognized surrounding obstacle This vertex is a vertex, and the lines connecting each vertex are followed by a polyline, and the polyline is simplified by polyline simplication. The simplified polyline may be generated and stored as simplified terrain data. As such, by simply generating and storing the terrain data, the size of the terrain data to be stored is also significantly reduced.

The above-mentioned "Polyline Simplication" technique is a "Vertext Reduction technique" to reduce the number of vertices by removing one or more near vertices of the near vertices associated with each vertex according to a specific condition, and the above One or more of the 'polyline approximation techniques' that approximate the generated polyline such that all vertices in the resulting polyline are within a defined distance range. The vertex removal technique removes the vertices from one vertex by a specific condition, for example, by removing vertices within a specified radius from one vertex and repeating this process to reduce the vertices. . The polyline approximation technique connects the two ends of a polyline with a straight line to make an initially approximate simple line, then calculates the distance to the approximated approximation line for all vertices and checks whether the value falls within the specified range. If all vertices are not within the defined range, all vertices are determined by creating a new approximation line that connects the vertex with the largest deviation from the approximation line and the previous end point, and calculates the distance to the approximate line for all vertices. It is a technique to approximate to have a value in a range. This polyline approximation technique is also known as Douglas-Peucker (DP) algorithm.

The terrain data generation storage 140 described above may normalize the generated terrain data by the number of nodes and store the generated terrain data as a numeric value between 0 and 1. FIG. This allows a more simplified map representation using only a combination of numbers and significantly reduces the size of the terrain data that needs to be stored.

That is, the above-described terrain data generation and storage unit 140 simplifies and stores the terrain data on the surroundings of the node with only some point information, normalizes the information about the moving obstacles generated in the process of obtaining the terrain data by the number of nodes, and Store the size as a numeric value between 0 and 1.

The terrain data storage device 100 for a robot according to an embodiment of the present invention may further include a map generator 150 for regenerating a map based on the terrain data generated and stored in the terrain data generation storage 140. Can be. The map regenerated by the map generator 150 is similar to the original grid map.

In the above-described terrain data storage device 100 for a robot according to an embodiment of the present invention, SLAM (Simultaneous Location and Mapping) technology is applied to simultaneously construct a map and perform location recognition in an unknown environment. It may be a robot or a device included in such a robot.

In addition, the terrain data storage device 100 for a robot according to an embodiment of the present invention may be linked with various sensors for acquiring information about an external environment, and may analyze the information acquired by such a sensor to analyze the surroundings. You can map a space and estimate where your location is on the map. The terrain data storage device 100 for a robot according to an embodiment of the present invention may include a GPS or RFID for acquiring absolute coordinate information, a vision or laser scanner for simple input information, etc. to estimate the position of the robot. And the like.

2 is a flowchart illustrating a terrain data storage method provided by the terrain data storage device 100 for a robot according to an embodiment of the present invention.

2, the terrain data storage method for a robot according to an embodiment of the present invention, the edge point extraction step of extracting the edge point from the grid map (S200); A Voronoi diagram generation step S202 for generating a Voronoi diagram based on the extracted edge points; A node searching step (S204) of searching for a node from the generated Voronoi diagram; And a terrain data generation storage step (S206) of scanning the surroundings of the found node to generate terrain data, and storing the generated terrain data.

In the above, the terrain data storage device 100 for a robot according to an embodiment of the present invention, and a terrain data storage method for a robot provided therefrom have been described. Hereinafter, FIGS. 3 to 12 will be described. A terrain data storage method for a robot will be described by way of example.

3 is a grid map applied to exemplarily describe a method of storing terrain data for a robot according to an embodiment of the present invention. Referring to FIG. 3, the dark portion becomes a fixed obstacle such as a wall, and the bright portion becomes a passage through which the robot can move.

In the terrain data storage method for a robot according to an embodiment of the present invention, first, an edge point to be used for generating a Voronoi diagram is extracted from the grid map illustrated in FIG. 3.

The terrain data storage device 100 for a robot according to an embodiment of the present invention now stores, as an example of point extraction, various cases in which edge points may occur (that is, various edge point cases). Various edge point cases stored in advance can be inspected, and points on the grid map corresponding to the edge point cases stored in advance as a result of the inspection can be extracted as edge points. In Fig. 4, eight examples of edge point cases examined to extract edge points are shown. In the eight edge point cases of FIG. 4, the portion marked 'x' corresponds to the edge point.

5 is a diagram illustrating an edge point extracted through inspection of edge point cases on a map. It can be seen that the edge point (indicated by 'x') extracted in FIG. 5 is a boundary between the light and dark portions in the grid map of FIG. 3.

After the edge points are extracted as shown in FIG. 5, the Voronoi diagram is generated using the extracted edge points, and the Voronoi diagram thus generated is illustrated in FIG. 6. It can be seen that the Voronoi diagram shown in FIG. 6 is a very complicated diagram. In this complex Voronoi diagram, it is quite difficult to get the information you want, so processing a complex Voronoi diagram using different conditions will result in a simplified Voronoi diagram that fits the real terrain. For example, complex Voronoi diagrams can be simplified by removing Voronoi points larger than the size of the map and removing Voronoi points that occur inside the wall. In addition, by using the connection of each point, all the overlapping points can be removed and the connection path can be determined. As such, a Voronoi diagram in which the complex Voronoi diagram shown in FIG. 6 is simplified is illustratively shown in FIG.

Referring to FIG. 7, the Voronoi diagram is followed by lines 700 passing through the center of the passage, which is the bright part, which is generally the line 700 passing through the center of the passage, since the robot makes a path that will not hit walls or obstacles. These are similar to the path that a real robot travels. These Voronoi diagrams provide insights into the terrain on the map.

As shown in FIG. 7, in order to identify the terrain through the Voronoi diagram generated as shown in FIG. 7, the terrain data is obtained by searching each point, and it may be inefficient to obtain the terrain data at all points. This is because, in the case of a passage (a part highlighted in FIG. 7) where there is almost no terrain change, even if the terrain data does not exist, since the direction of travel can be directly known by using the sensor of the robot online, the point where such terrain change is not severe. Acquiring topographical data from to may be inefficient in terms of processing and storage data volume, and the like. Therefore, it is more efficient to obtain the terrain data only at the point where the terrain changes are severe (such as a 'node') without obtaining the terrain data for a point such as a passage having almost no such terrain change. In other words, by grasping the severe point of the terrain change, that is, the surrounding terrain of the node, the entire terrain can be indirectly grasped, and the robot can be fully operated through this. The aforementioned node is a point where the robot can move in three or more directions, as shown in the four node cases ((a), (b), (c) and (o)) illustrated in FIG. 8. It is a point where the topographical change is severe, such as a corner point of a passageway or an intersection of two or more passageways.

FIG. 9 is a diagram illustrating 23 nodes retrieved from the Voronoi diagram exemplarily shown in FIG. 7. The 23 nodes (including node 900) shown in FIG. 9 are indicated by small square boxes on the connection line 700.

As shown in FIG. 9, since the terrain data around the found node has important information for identifying the terrain, the terrain data is scanned and generated and stored. The terrain data obtained by scanning the periphery of the retrieved node is exemplarily illustrated in FIG. 10. FIG. 10 is a graph illustrating terrain data obtained by scanning the periphery of a specific node among searched nodes. In the graph of FIG. 10, the X-axis and the Y-axis denote distances from peripheral obstacles such as walls, and each point indicated by a '+' symbol on the graph indicates a position of a fixed peripheral obstacle. Referring to FIG. 10, if the distance unit of the X-axis and the Y-axis is m, there is an obstacle 1000 at a point approximately 5m away from the (0,0) point where the robot is located in the X-axis direction and the Y-axis direction, respectively. It can be seen that various obstacles are recognized in the X-axis direction and the Y-axis direction based on (1000).

Referring to FIG. 10, in the case of terrain data obtained by scanning the periphery of a node, similar information is displayed at many points, which results in a loss of data loss and efficiency of information amount. Therefore, it is necessary to simplify and store the scanned terrain data obtained so as to have a larger amount of information while using a smaller amount of data. This simplified topographical data is illustratively shown in FIG. 11.

FIG. 11A, which is a graph of simplified terrain data with respect to a specific node, is compared to FIG. 10, and it is understood that each point (ie, obstacles) is connected to the line 1100 and simplified. In this case, as an example of the simplification technique, the aforementioned polyline simplification technique is used. FIG. 11B is an enlarged graph of a portion indicated by a box in FIG. 11A. If all of the lines (meaning simplified terrain data) shown in FIG. 11A are connected, a map as shown in FIG. 12 may be regenerated.

As described above, according to the exemplary embodiment of the present invention, while the space recognition or the position recognition of the robot is precisely performed to enable the normal operation of the robot, only the terrain data for the point necessary for the external terrain is obtained, and the terrain data is efficiently There is an effect that enables the acquisition of.

In addition, according to an embodiment of the present invention, there is an effect of reducing the storage space by efficiently storing the terrain data obtained for spatial recognition or location recognition of the robot.

In the above description, all elements constituting the embodiments of the present invention are described as being combined or operating in combination, but the present invention is not necessarily limited to the embodiments. In other words, within the scope of the present invention, all of the components may be selectively operated in combination with one or more. In addition, although all of the components may be implemented in one independent hardware, each or all of the components may be selectively combined to perform some or all functions combined in one or a plurality of hardware. It may be implemented as a computer program having a. Codes and code segments constituting the computer program may be easily inferred by those skilled in the art. Such a computer program may be stored in a computer readable storage medium and read and executed by a computer, thereby implementing embodiments of the present invention. The storage medium of the computer program may include a magnetic recording medium, an optical recording medium, a carrier wave medium, and the like.

In addition, the terms "comprise", "comprise" or "having" described above mean that the corresponding component may be included, unless otherwise stated, and thus excludes other components. It should be construed that it may further include other components instead. All terms, including technical and scientific terms, have the same meanings as commonly understood by one of ordinary skill in the art unless otherwise defined. Terms commonly used, such as terms defined in a dictionary, should be interpreted to coincide with the contextual meaning of the related art, and shall not be construed in an ideal or excessively formal sense unless explicitly defined in the present invention.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention but to describe the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

1 is a block diagram of a terrain data storage device for a robot according to an embodiment of the present invention;

2 is a flowchart illustrating a method of storing terrain data for a robot according to an embodiment of the present invention;

3 is a diagram illustrating a grid map for applying a terrain data storage method for a robot according to an embodiment of the present invention;

4 illustratively illustrates an edge point case being examined to extract edge points from a grid map;

5 is a diagram illustrating an edge point extracted from a grid map by way of example;

6 is an exemplary diagram showing a Voronoi diagram;

7 illustrates a simplified Voronoi diagram illustratively;

8 illustratively illustrates a node case for retrieving nodes from a Voronoi diagram;

9 shows an example of a node retrieved from a Voronoi diagram;

10 is a diagram illustrating terrain data generated by scanning the periphery of a node;

11 is a diagram illustrating terrain data generated through a polyline simplification technique.

12 is a diagram exemplarily illustrating a map reproduced based on terrain data.

<Description of Symbols for Main Parts of Drawings>

100: terrain data storage

110: edge point extraction unit

120: Voronoi diagram generator

130: node search unit

140: terrain data generation storage

150: map generator

Claims (13)

An edge point extracting unit extracting an edge point from the grid map; A Voronoi diagram generator for generating a Voronoi diagram based on the edge point; A node searching unit searching for a node from the Voronoi diagram; And A terrain data generation storage unit configured to scan the periphery of the node to generate terrain data, and store the generated terrain data; The Voronoi diagram generator, Generate a connection path by generating a plurality of points based on the extracted edge point, connecting one or more points by removing one or more points from the generated plurality of points according to a point removal method, and including the generated connection path. Terrain data storage device for a robot, characterized in that for generating the Voronoi diagram. The method of claim 1, The edge point extraction unit, Terrain data storage device for a robot, characterized in that for extracting the edge point from the grid map through the inspection of a pre-stored edge point case. delete The method of claim 1, The point removal method, Terrain data for a robot comprising at least one of a method for removing a point larger than the grid map, a method for removing a point occurring inside a wall of the grid map, and a method for removing a duplicated point. Storage device. The method of claim 1, The generated connection path is, Terrain data storage device for a robot comprising lines passing through a center of a passage on the grid map. The method of claim 1, The node search unit, Terrain data storage device for a robot, characterized in that for searching the nodes that can proceed in more than three directions in the Voronoi diagram. The method of claim 1, The terrain data generation storage unit, Scanning the periphery of the node to recognize surrounding obstacles, obtain geographic information on the recognized surrounding obstacles, and generate and store the terrain data based on the acquired geographic information. Storage device. The method of claim 7, wherein The geographic information, And location information of the recognized peripheral obstacle and distance information between the recognized peripheral obstacle and the robot, wherein the recognized peripheral obstacle is a fixed obstacle or a moving obstacle. The method of claim 7, wherein The terrain data generation storage unit, Scans the periphery of the node to recognize peripheral obstacles, obtains geographic information for the recognized peripheral obstacles, and based on the acquired geographic information, the point where the recognized peripheral obstacles are located as a vertex, Lines connecting each vertex are then created to generate a polyline, the generated polyline is simplified through polyline simplication, and the simplified polyline is generated as the terrain data. Terrain data storage device for a robot, characterized in that for storing. The method of claim 9, The polyline simplification technique, Vertex Reduction, which reduces the number of vertices by removing one or more near vertices of one or more of the near vertices associated with each vertex, and generating the vertices so that all vertices in the generated polyline are within a defined distance range. And at least one of a polyline approximation technique for approximating the drawn polyline. The method of claim 1, The terrain data generation storage unit, The terrain data storage device for a robot according to claim 1, wherein the generated terrain data is normalized by the number of nodes and stored as a numeric value between 0 and 1. FIG. The method of claim 1, And a map generator for regenerating a map based on the stored terrain data. An edge point extraction step of extracting an edge point from the grid map; Generating a Voronoi diagram based on the edge point; A node searching step of searching for a node from the Voronoi diagram; And Generating a terrain data by scanning the periphery of the node, and storing and generating the terrain data; Generate a connection path by generating a plurality of points based on the extracted edge point, connecting one or more points by removing one or more points from the generated plurality of points according to a point removal method, and including the generated connection path. Terrain data storage method for a robot, characterized in that for generating the Voronoi diagram.

Images