KR101037379B1 - The moving robot exploration system founded the distance information of a surrounding circumstance required from distance sensor and the exploration method using the moving robot exploration system - Google Patents

Abstract

The present invention relates to an exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor. The mobile robot exploration system acquires terrain information through a distance sensor and generates a phase map by itself. It is about exploration method to move.

An exploration method using a mobile robot sensing system based on distance information of a surrounding environment obtained from a distance sensor according to the present invention includes a distance information input step of receiving distance information of a surrounding environment from a distance sensor unit, and a map generation unit for distance information. A node extraction step of obtaining node information from the information input at the input step, a first child node extraction step of finding a first child node among node information input at the node extraction step of the map generator; Concave node extraction step to find concave node among the first child node, select nodes to be used as the phase map in the first child node and generate the topology map Global phase map generation step and target point (ta) to select node to move rget point) selection step.

Mobile Robot Exploration System, Topology Map, Concave Node, Exploration, Distance Sensor

Description

The moving robot exploration system founded the distance information of a surrounding circumstance required from distance sensor and the exploration method using the moving robot exploration system }

 The present invention relates to an exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor, wherein the mobile robot exploration system obtains distance information of the surrounding environment and extracts node information to determine the surrounding environment. It is about how to explore.

Beginning with industrial robots, the area has recently been expanded in the form of intelligent robots that can operate in any environment. Intelligent robots can be applied to various fields such as service, military, security, and dangerous area exploration, and it is essential to secure autonomous driving technology of such robots. Among them, exploration technology that moves while acquiring information and generating a map by itself is a key element when the robot is located in a space where information about the surrounding environment is not available.

 The robot's exploration technology is basically made with map generation, and there are various methods depending on which map and which kind of sensor are used. Recent research mainly uses topological maps, which can easily describe maps, and sensors that obtain distance information such as lasers and ultrasonic sensors.

Conventional techniques for generating maps and performing exploration on the basis of such sensors are as follows.

 Domestic patents include a patent for creating a phase map by applying a thinning algorithm to the surrounding environment information obtained using a distance sensor (application number 10-2005-0057716) and a robot generating a phase map while moving along a wall. Patent for a method of performing autonomous driving (application number 10-2006-0003787) and a method for generating a phase map based on a path traveled by a robot using a distance sensor and an obstacle recognition sensor (publication number 10 -2004-0087171. The first patent has the disadvantage that efficiency is inferior because the robot makes the topographic map grid map at the same time. In the second patent, it moves along the wall, so it acquires the distance information and the spatial information in generating the topographic map. It can be seen that the efficiency is poor. In addition, the third patent has a disadvantage in that the efficiency of exploration of the entire environment is reduced because it moves simply avoiding obstacles without any strategy in exploration.

 Among the published papers, similar technologies to this patent are as follows. B. Yamauchi presented a method to perform an exploration while generating a grid map based on a distance sensor. (B. Yamauchi, "A Frontier-Based Approach for Autonomous Exploration," IEEE CIRA, 1997.) This method uses grid maps to create maps based on one coordinate system, as well as the disadvantage of large map information. Therefore, there was a problem that errors accumulated continuously during the exploration. H. Choset et al. Proposed a method of performing node exploration using GVG (generalized voronoi graph) on the data obtained from distance sensors (H. Choset and K. Nagatani, "Topological simultaneous localization and mapping). (SLAM): towards exact localization without explicit localization, "IEEE Transactions on Robotics and Automation, vol. 17, no. 2, pp. 125-137, 2001.). This method has the disadvantage that the exploration method is not efficient because it simply moves to the nearest node order during the exploration.

The present invention has been made to solve the inconvenience of the prior art as described above, the robot obtains environmental information using only the distance sensor without prior information about the space where the robot is located to generate a map and perform exploration efficiently To provide a way.

In addition, another object of the present invention is to generate a topographic map of a small map size and at the same time using a local coordinate system, to reduce the errors caused by the robot to perform the exploration, and to perform the exploration by visiting the nodes blocked three sides in advance, It is to provide an exploration method that increases the time or distance required to visit all the spaces.

As a technical idea for achieving the above object, in the present invention, in the mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor as a first aspect, a camera unit for receiving an image of the external surrounding environment ( 100); A robot body unit 200 for storing external motion and object models, inputting distance information, and generating a global phase map; The mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor, characterized in that it comprises a robot moving unit 300 for moving the mobile robot exploration system is presented.

In the first aspect,

The robot body 200,

A money machine 210 which is a mechanical arm for external operation; A distance sensor unit 220 mounted to the outside to input distance information of the surrounding environment; A database 230 located therein and storing an object model for object recognition; Characterized in that it comprises a map generation unit 240 for generating a global phase map using the distance information input from the distance sensor unit 220,

In the first aspect,

The robot moving part 300, characterized in that the wheel or caterpillar line,

The distance sensor unit 220 is characterized in that to obtain the distance information of all directions,

In the first aspect,

The global phase map is characterized by using a local coordinate system,

In the first aspect,

The mobile robot exploration system,

It is characterized by exploring the surrounding area by obtaining the surrounding environment information from the distance sensor in the state of not having prior information on the surrounding environment.

As a second aspect, an exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor, comprising: a first step (S110) of receiving distance information from a distance sensor unit 220; Extracting node information capable of moving the mobile robot exploration system from the input distance information (S120); A third step (S130) of extracting a first connection node (first child node) from node information capable of moving the received mobile robot exploration system; Step 4 (S140) of finding a concave node among the extracted first child nodes; A fifth step (S150) of making the global topographical map of the first child nodes capable of moving the mobile robot exploration system; Based on the distance information of the surrounding environment obtained from the distance sensor, comprising a sixth step (S160) for determining a target point which is the position to which the mobile robot exploration system moves next from the global phase map. An exploration method using a mobile robot exploration system is presented.

In the second aspect,

The distance sensor unit 220 is characterized in that to obtain the distance information of all directions,

In the second aspect,

The global phase map is characterized by using a local coordinate system,

In the second aspect,

The mobile robot exploration system, characterized in that to explore the surrounding area by obtaining the surrounding environment information from the distance sensor in the state that does not have prior information about the surrounding environment,

In the second aspect,

In the second step, the surrounding environment information obtained from the distance sensor unit is stored in the form of an image, a skeletalization algorithm is applied thereto, and a mask of Equation 1 is applied to the generated skeletal image to find a movable node. Features,

In the second aspect,

In the third step, a first child node is applied to each arc extending from the current position of the mobile robot exploration system in the skeletal image by applying a mask (2) suitable according to the direction of the arc. Characterized by finding

In the second aspect,

Step 4 is a mask of equation (2) according to the direction of the arc used when searching for the primary connection node in consideration of the characteristic that the end of the concave node is connected to two end nodes and one branch node (branch node) Finding the position of the concave node by applying the,

In the second aspect,

The skeletalization algorithm is a morphological image processing algorithm or a brushfire algorithm.

An exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor according to the present invention obtains distance information of a surrounding environment, extracts node information, and generates a phase map to explore the surrounding environment by itself. Can be. In addition, the concept of a concave node can be used to increase the efficiency in terms of time and distance in exploring the environment.

Hereinafter, with reference to the accompanying drawings will be described in detail the configuration and operation of the embodiment of the present invention.

In order to achieve the above object, the exploration method using the mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor according to the present invention moves the surrounding environment by itself without the prior information about the surrounding environment. The present invention relates to a method for generating a map, and an invention for an exploration method in which a mobile robot exploration system obtains terrain information by a distance sensor and moves while generating global phase map information.

An exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor according to the present invention includes a distance information input step of receiving distance information of a surrounding environment from a distance sensor unit 220, and map generation. A node extraction step of acquiring node information from the information input in the additional distance information input step, and a first child node finding a first child node among the node information input in the node extraction step of the map generator. Concave node extraction step to find concave node among extraction step and first child node, node to be used as phase map in first child node Global phase map generation step for generating map and target point for selecting node to move mobile robot exploration system in topology map Including the (target point) selection stage.

The mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor according to the present invention and the exploration method using the same will be described in more detail.

1 is a structural diagram of a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor.

Mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor according to the present invention, as shown in Figure 1 including a camera unit 100, the robot main body 200 and the robot moving unit 300 It is composed.

The camera unit 100 receives an image of an external surrounding environment. The robot main unit 200 stores external motion and object models, inputs distance information, and generates a global phase map. The robot moving unit 300 moves the mobile robot exploration system.

The robot main body 200 has a manipulator 210, which is a mechanical arm for external operation, and a distance sensor unit 220 for inputting distance information of the surrounding environment, mounted on the outside, and an object for object recognition therein. And a map generator 240 (not shown) for generating the global phase map using the database 230 storing the model and distance information input from the distance sensor 220.

Ideally, the distance sensor unit 220 may be mounted one by one before and after the robot body 200 to obtain an omnidirectional distance at a time.

The robot moving unit 300 may use any device capable of moving the main body of the mobile robot exploration system including a wheel or a caterpillar as a configuration for moving the mobile robot exploration system along a moving path.

2 is a flowchart of an exploration method using a mobile robot exploration system based on distance information of a surrounding environment obtained from a distance sensor.

In the exploration method using the mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor according to the present invention, as shown in FIG. 2, step 1 of receiving distance information from the distance sensor unit 220 (S110). )Wow; Extracting node information capable of moving the mobile robot exploration system from the input distance information (S120); A third step (S130) of extracting a first connection node (first child node) from node information capable of moving the received mobile robot exploration system; Step 4 (S140) of finding a concave node among the extracted first child nodes; A fifth step (S150) of making the global topographical map of the first child nodes capable of moving the mobile robot exploration system; And a sixth step S160 of determining a target point, which is a position to which the mobile robot exploration system moves next from the global topological map. Therefore, in the present invention, basically, the environmental distance information is first obtained, the node information is extracted based on this, a global topology map is generated, and the mobile robot exploration system is moved.

The distance sensor unit 220 acquires omnidirectional distance information and connects the acquired distance information in angular order to generate a closed area and store the binary image. 100mm of distance information generates an image corresponding to 1 pixel of a binary image. And the inside of the closed region of the generated binary image is filled using morphological image processing algorithm. The skeleton image of FIG. 3 is extracted by applying a brushfire algorithm to the filled lung region image.

It is necessary to extract information to be used as a node of the phase map from the obtained skeleton image. Node information is found by applying a mask of Equation 1 to skeleton information obtained as shown in FIG. 3.

When the mask of Equation 1 is applied to the skeleton image, node information may be classified as shown in FIG. 4. The result as shown in (a) of FIG. 4 means an end node, the result as shown in (b) means a branch node, and the result as shown in (c) means an arc. . In the above results, a branch node having a result as shown in (b) is used as a node of the topology map. The result of using the mask of Equation 1 is found in a lump form as shown in FIG. You need to find the center point from each chunk of branch nodes. First, among the image processing algorithms, a labeling algorithm is used to classify the chunks. The labeling algorithm is one of the commonly used image processing algorithms. The labeling algorithm searches for adjacent pixels and binds them together. Then, one branch node is found in each chunk as shown in FIG. 6 using an algorithm for finding the center of gravity in each chunk. As a result, information about the branch node was extracted from the skeleton information. Among them, the mobile robot exploration system must go through the process of filtering the nodes generated in the non-movable space. This can be found by applying a mask of the size of the mobile robot exploration system to each node as shown in FIG. 7 by examining whether there is a part overlapping with the closed region image generated from the laser data.

Exploration of the mobile robot exploration system proceeds based on the nodes obtained in the node extraction step. First, when the mobile robot exploration system starts the exploration without any information about the surrounding environment, it performs the initialization process. When the initialization process starts, the mobile robot exploration system performs the node extraction step and moves to the nearest branch node among the branch nodes obtained in the node extraction step. The shifted position is selected as (0,0,0) which is the initial point of the global phase map.

After setting the origin, the mobile robot exploration system obtains omnidirectional distance information of the surrounding environment through the distance sensor unit 220. The node extraction step is performed from the obtained omnidirectional distance information. The node to be moved next is selected from branch node information acquired in the node extraction step. In the present invention, only the first child node connected to the current position of the mobile robot exploration system among the branch nodes obtained from the skeleton image is used as the node to which the mobile robot exploration system moves. At this time, there is a high possibility that a branch node does not occur at the location of the current mobile robot exploration system. Therefore, it is necessary to select a reference point for performing the search of the first child node. The reference point uses all branch points that exist within a radius of 1 m from the location of the current mobile robot exploration system. If a branch node does not exist within a radius of 1m, a pixel point of the nearest skeleton image is used as a reference point. After selecting the reference point, the first child node (first child node) is searched. Applying the mask of Equation 1 to the reference point it can be seen in which direction the arc (arc) extends from the reference point. When the direction of an arc is obtained, a branch node is searched along the direction. At this time, in order to prevent returning to the pixel where the search has already been performed, a method of using a mask having a different form according to the direction of an arc as in Equation 2 is proposed.

The mask proposed by Equation 2 is used in the form as shown in FIG. The mask is applied to determine the node type of the pixel to which the mask is applied as in the node extraction step. If the result of applying the mask is 2 or more, it is determined to be a branch node, 1 is an arc, and 0 is an end node. In addition, when the current pixel is an arc, the direction is obtained from the masking result, and the mask of the corresponding direction is selected from the masks of Equation 2 to continuously search for the first child node. .

After the search for the first child node is completed, a process of searching for a concave node is performed. As shown in FIG. 9, a concave node refers to a node with three blocked surfaces, and as shown in FIG. 10, a primary connection node includes two end nodes and one branch node. Finding a convex node is similar to finding a first child node. For each first child node, one branch node and two end nodes are found by searching for nodes that are connected first using the mask proposed in Equation 2. If so, select the concave node.

It is necessary to filter out the nodes represented in the same space among the first child node acquired at the position of the current mobile robot exploration system and the nodes of the topology map stored in the database of the mobile robot exploration system. This process uses two criteria. First, we find the node closest to the first child node by comparing it with all nodes of the global topological map. If a node that is different from the current position and a distance less than a D _{Threshold, which} is arbitrarily determined by the user, is determined to be a duplicate position with the node. Second, the angle difference between the first child node and the node closest to the node and the angle difference between the nodes that are primarily connected to the node is compared to be less than or equal to θ _{Threshold, which} is arbitrarily determined by the user. In this case, it is determined to be an unnecessary node. An example of its use is shown in FIG. The first child node except for unnecessary node information is stored as a global topological map. Then select a target point. The target point is determined based on whether the node is a concave node and the distance between nodes. If the first child node does not exist, the nearest node among the nodes not visited in the global topological map is selected as the target point. At this time, the distance to the node is calculated using Dijkstra's algorithm. The exploration is terminated when there are no more nodes visited on the topological map.

Although the embodiments of the present invention have been described above with reference to the accompanying drawings, those skilled in the art to which the present invention pertains may be embodied in other specific forms without changing the technical spirit or essential features of the present invention. You will understand that. Accordingly, the embodiments described above are exemplary in all respects and not restrictive.

 1 is a structural diagram of a mobile robot exploration system for performing exploration and generating a phase map based on distance information according to the present invention.

2 is a basic flowchart of a method of performing an exploration and generating a phase map based on distance information according to the present invention.

3 is a result of skeletalization after generating a binary image of the omnidirectional distance information acquired by the mobile robot exploration system in the present invention.

4 is a conceptual diagram of a branch node, an end node, and an arc search using a mask according to the present invention.

5 is a conceptual diagram of a result of performing a branch node search using a mask according to the present invention.

FIG. 6 is a conceptual diagram of a result of finding a center point among agglomerated branch nodes according to the present invention.

FIG. 7 is a conceptual diagram of a result of filtering out branch nodes that are not movable in the mobile robot exploration system according to the present invention.

8 is a conceptual diagram of the use of masks to find a primary connection node in accordance with the present invention.

9 is a conceptual diagram of a region where a concave node is found according to the present invention.

10 is a conceptual diagram of characteristics in skeleton information of a concave node according to the present invention.

11 is a conceptual diagram for the extraction of overlapping nodes according to the present invention.

              <Description of Major Symbols in Drawing>

100: camera portion 200: robot body portion

210: Money Purifier 220: Distance Sensor

230: database 240: map generator

300: robot moving part

Claims (15)

delete delete delete delete delete delete delete In the exploration method using the mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor, A first step of receiving distance information from the distance sensor unit; Extracting node information capable of moving the mobile robot exploration system from the input distance information; Extracting a first child node from the node information capable of moving the received mobile robot exploration system; Finding a concave node among the extracted first child nodes; Creating a global topology map of the first child nodes capable of moving the mobile robot exploration system; And a sixth step of determining, from the global topological map, a target point, which is a position to which the mobile robot exploration system moves to next. The second step, Storing the surrounding environment information obtained from the distance sensor in an image form, applying a morphology image processing algorithm and a brushfire algorithm, and applying a mask of Equation 1 to the generated skeleton image to find a movable node. An exploration method using a mobile robot exploration system based on distance information of surrounding environment obtained from a characteristic distance sensor. [Equation 1]

The method according to claim 8, The distance sensor unit is an exploration method using a mobile robot detection system based on the distance information of the surrounding environment obtained from the distance sensor, characterized in that to obtain the distance information of the omnidirectional. The method according to claim 8, The global phase map is an exploration method using a mobile robot exploration system based on the distance information of the surrounding environment obtained from a distance sensor, characterized in that using a local coordinate system. The method according to claim 8 The mobile robot exploration system, An exploration method using a mobile robot exploration system based on the distance information of the surrounding environment obtained from the distance sensor, wherein the surrounding environment is obtained by obtaining the surrounding environment information from the distance sensor without having prior information on the surrounding environment. delete The method according to claim 8, The third step, In the skeletal image, the first child node is found for each arc extending from the current position of the mobile robot exploration system by applying a mask according to Equation 2 according to the direction of the arc. Exploration method using mobile robot exploration system based on distance information of surrounding environment obtained from distance sensor. [Equation 2]

The method according to claim 8, The fourth step, Considering that the concave node is connected to two end nodes and one branch node, a concave is applied by applying a mask according to Equation 2 according to the direction of an arc used when searching for a primary connection node. An exploration method using a mobile robot exploration system based on distance information of the surrounding environment obtained from a distance sensor characterized in finding the position of a node. [Equation 2]

delete

Images