JP2013109325A - Map creation method using group of intelligent robots and device for the same - Google Patents

Abstract

An apparatus capable of generating an entire environment map by fusing information from each robot is provided.
A relative position between robots autonomously traveling in a search range where a map is to be created is detected, a search area for each robot is divided based on the detected relative position, and a map is generated from each robot. After acquiring 208 each local information for the search area used for the search, the information is merged to generate 210 the entire environment map, so that the search range for creating the map in the informal environment It is possible to effectively suppress the search time from being excessively reduced, and to reduce the merging time of the local information 202 acquired from each robot through selective matching.
[Selection] Figure 2

Description

  The present invention relates to a technique for constructing a map using intelligent robots of a crowd. More specifically, the present invention divides an entire area (entire area) for which a map is to be created into a plurality of search areas for each robot. The present invention relates to a map construction method and apparatus using a crowd intelligent robot suitable for generating an entire environmental map using local information acquired from an area.

  As is well known, the most basic function of a mobile object such as an autonomous mobile robot is that the mobile object must be able to move to a desired target point without a collision. Such a function is an autonomous mobile robot. The position measurement technique and the map construction technique are used.

  Here, as a method for recognizing the map configuration and the position of the moving body, it is known to use, for example, a SLAM (Simultaneous Localization And Mapping) algorithm.

  As described above, in the conventional typical technique using the SLAM algorithm, research for recognizing a position in a two-dimensional formal environment and recognizing the environment has become the mainstream. Currently, it has been applied to products.

  However, since the conventional technology for recognizing the position of the moving body and the surrounding environment using the SLAM algorithm is limited to a formal environment, when applied to an informal environment, the range is wide and enormous. There is a problem that it takes a long time to search.

Republic of Korea Open Patent 2010-0005488

  Therefore, the present invention has been made in view of the above circumstances, and an object of the present invention is to detect a relative position between robots that autonomously travel in a search range where a map is to be created, and based on the detected relative position. The search area for each robot is divided and local information on the search area used for map generation is obtained from each robot, and then the information is merged to generate the entire environment map. It is in.

  According to one aspect of the present invention, a process of measuring a relative distance between robots to detect a relative position between the robots, a process of dividing a search area for each robot based on the detected relative positions, A process of transmitting each divided search area information to each corresponding robot, a process of acquiring local information used for generating a map from each robot, and each acquired local information. A map construction method using intelligent robots of the crowd including the process of generating the entire environmental map.

  According to another aspect, the present invention provides a relative position acquisition block that detects and shares a relative position between robots, and a search area division block that divides a search area for each robot based on the detected relative position. A crowd including a local information collection block for acquiring local information used for generating a map from each robot, and an integrated environment map generation block for generating an entire environment map by fusing each acquired local information A map construction device using an intelligent robot is provided.

  According to the present invention, the relative position between the robots autonomously traveling in the search range for which a map is to be created is detected, the search area for each robot is divided based on the detected relative position, and the map is obtained from each robot. The search range for creating maps in informal environments is obtained by acquiring local information for each search area used to generate the map and then fusing the information to generate the entire environment map. It is possible to effectively suppress the search time from being reduced and reduce the merge time of local information acquired from each robot through selective matching.

It is a conceptual diagram explaining the example which produces | generates the whole environment map using the local information acquired from each robot, after dividing | segmenting a search range into the search area for many each robot by this invention. 1 is a block configuration diagram of a map construction device using a crowd intelligent robot according to the present invention. FIG. It is a division | segmentation example figure of the search area explaining the example which divides | segments and allocates a search range to the search area for each robot by this invention. FIG. 5 is a flow chart illustrating a main process of constructing a map using an intelligent robot of a crowd according to the present invention.

  The gist of the technology of the present invention is that, unlike the above-described conventional technology for recognizing the position of the moving body and the surrounding environment using the SLAM algorithm in a formal environment, the vehicle travels autonomously in the search range (search area) to be created. The relative position between each robot is detected, the search area for each robot is divided based on the detected relative position, and local information on the search area used for generating a map (map) from each robot is respectively obtained. After acquisition, this information is merged to generate the entire environmental map, and the present invention effectively improves the problems of the conventional method by realizing the target through such technical means. it can.

  Here, the creation of the entire environment map by merging the local information of the divided search areas can be performed using, for example, a SLAM algorithm.

  In the following description of the present invention, when it is determined that there is a possibility that specific descriptions regarding known functions or configurations may unnecessarily obscure the gist of the present invention, detailed description thereof will be omitted. . The terms to be described later are defined in consideration of the functions of the present invention, and it is needless to say that the terms may be changed depending on the intention or practice of the user or the operator. Therefore, the definition should be made based on the technical idea described throughout the present specification.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a conceptual diagram illustrating an example in which an entire environment map is generated using local information acquired from each robot after the search range is divided into a number of search sections for each robot according to the present invention.

  Referring to FIG. 1, the conceptual diagram includes one robot 111 that creates (generates) an entire environment map in a search range (search region) and four robots 112 to 115 that have crowd intelligence in the entire search range. The robot 111 measures the relative distance between the robots 112 to 115 to detect the relative position between the robots, and divides the search section for each robot based on the state position thus detected. Here, each robot is defined as a mobile body provided with wireless communication means for wirelessly transmitting local information (for example, ultrasonic waves, infrared rays, lasers, cameras, odometry, etc.) acquired during traveling in the search range. obtain. And sensor information such as ultrasonic waves, infrared rays, lasers, cameras, odometry, etc. can be commonly defined (represented) as video data.

  For example, a search section 121 can be assigned to the robot 112, a search section 122 can be assigned to the robot 113, a search section 123 can be assigned to the robot 114, and a search section 124 can be assigned to the robot 115. The robots 112 to 115 wirelessly transmit the local information acquired by searching the search section assigned to the robot 112 to the robot 111. Here, the local information acquired by each of the robots 112 to 115 and wirelessly transmitted to the robot 111 may include, for example, ultrasonic information, infrared information, laser information, camera information, odometry, and the like.

  At this time, it is preferable to increase or decrease the number of robots participating in the creation of the search range map in proportion to the size of the search range so as to prevent excessive search time.

  FIG. 2 is a block diagram of the map construction apparatus using the intelligent robot of the crowd according to the present invention, and includes an interface block 202, a relative position acquisition block 204, a search area division block 206, a local information collection block 208, and an integrated environment map. The integrated environment map generation block 210 may include a feature point extraction block 2102, a merge matching block 2104, a map creation block 2106, and the like.

  Referring to FIG. 2, the relative position acquisition block 204 transmits / receives wireless data to / from the robots 112 to 115 through the interface block 202 to measure the relative distances of the robots 112 to 115 with respect to the robot 111. It is possible to provide a function to detect (recognize) and share the relative position between robots (relative coordinate values generated based on the relative distance between robots) based on the measured relative distance. The relative position information between the robots is transmitted to the search area division block 206.

  Here, the relative distance between each robot can be measured using, for example, a TWR (Two Way Ranging) technique which is one of RTT (Round Trip Time) techniques using a network-based WPAN or WLAN signal.

  Next, the search area division block 206 divides the entire search range into search areas for each robot based on the relative position transmitted from the relative position acquisition block 204. For example, as shown in FIG. When it is assumed that one robot 311 and four crowd intelligence robots 312 to 315 for creating the entire environment map belong to 320, the search area 321 is divided (assigned) to the robot 312, and the robot 313. The search area 322 may be divided (assigned), the robot 314 may be divided (assigned), and the robot 315 may be divided (assigned).

  Here, the search section may be specified as an area relatively close to the position of the corresponding robot, and adjacent search areas may be specified to overlap each other by a certain area, but this is to improve the accuracy of the created map. It is. That is, in FIG. 3, reference numeral OL1 indicates an overlapping area where the search sections of the robot 312 and robot 313 overlap, and reference numeral OL2 indicates an overlapping area where the search sections of the robot 313 and robot 314 overlap. OL3 indicates an overlapping area where the search sections of the robot 314 and the robot 315 overlap.

  The search area division block 206 wirelessly transmits each search area information assigned to each robot to the corresponding robot through the interface block 202 so that each robot can search the search area assigned to itself. The search section information for each robot can be transmitted to the local information collection block 208.

  The local information collection block 208 acquires local information (for example, ultrasonic information, infrared information, laser information, camera information, odometry, etc.) used for generating a map from each of the robots 112 to 115, that is, Local information acquired by each robot 112-115 through traveling in the search area assigned to each robot 112 is collected through the interface block 202 and transmitted to the integrated environment map generation block 210 together with the molecular information and search section information for each robot. Can provide functions such as

  Next, the integrated environment map generation block 210 can provide a function of, for example, generating the entire environment map by merging the local information for each robot transmitted from the local information collection block 208 through execution of the SLAM algorithm. For this, a feature point extraction block 2102, a merge matching block 2104, a map creation block 2106, and the like can be included.

  First, the feature point extraction block 2102 uses, for example, a split and merge algorithm and / or a SIRF (Scale Invariant Feature Transform) algorithm to extract feature points (for example, straight lines, curves, identifications) from each local information. The feature points extracted here may mean landmarks or feature points with respect to a specific pattern.

  Next, the merge matching block 2104 determines whether the local information is information of a nearby robot, that is, through matching between the search section information for each robot and the local information for each robot, as shown in FIG. As described above, it is determined whether the video information corresponds to the overlap areas OL1, OL2, OL3. When the video information is determined to be the video information corresponding to the overlap area, matching between the corresponding robots is performed. Provides functions such as merging feature points. Here, merging the feature points of the portions that match each other through the matching operation is to reduce the amount of calculation in creating the map. That is, the merge matching block 2104, after confirming the presence or absence of video information for the same information through the matching operation, merges the same information into one, otherwise excludes it from the merge target.

  Finally, the map creation block 2106 combines the local information acquired by each robot using the feature points extracted through the feature point extraction block 2102 and the feature points merged through the merge matching block 2104, so that Functions such as creating the entire environmental map can be provided.

  Next, a series of processes for creating the entire environmental map using the crowd intelligent robots through the map construction apparatus having the above-described configuration will be described.

  FIG. 4 is a flowchart illustrating a main process for constructing a map using an intelligent robot of a crowd according to the present invention.

  Referring to FIG. 4, the relative position acquisition block 204 transmits / receives wireless data to / from the robots 112 to 115 through the interface block 202 to measure the relative distances of the robots 112 to 115 with reference to the robot 111. The relative position between the robots based on the relative distance (relative coordinate value generated based on the relative distance between the robots) is detected and transmitted to the search section division block 206 (step 402). Here, the relative distance between the robots can be measured using a TWR technique, which is one of RTT techniques using a network-based WPAN or WLAN signal, for example.

  In response to this, the search area dividing block 206 divides the entire search area into search areas for each robot based on the detected relative positions between the robots (step 404). As an example, as shown in FIG. 3, when it is assumed that one robot 311 and four crowd intelligence robots 312 to 315 for creating the entire environment map belong to the entire search range 320, Divides (assigns) the search area 321, divides (assigns) the search area 322 to the robot 313, divides (assigns) the search area 323 to the robot 314, and divides (assigns) the search area 324 to the robot 315. )it can. Here, the search section may be specified as an area relatively close to the position of the corresponding robot, and adjacent search areas may be specified to overlap each other by a certain area.

  Next, in the search area division block 206, each search area information assigned to each robot is wirelessly transmitted to each corresponding robot through the interface block 202 (step 406). As a result, each robot is assigned to itself. Local information is collected while traveling in the search section.

  That is, each robot in the search range collects local information (for example, ultrasonic information, infrared information, laser information, camera information, odometry, etc.) while traveling in the search section assigned to itself, and this collected The local information is wirelessly transmitted to the robot that creates the entire environment map. Such local information for each robot is received by the local information collection block 208 through the interface block 202 (local information acquisition) (step 408).

  In the local information collection block 208, local information acquired from each robot is transmitted to the integrated environment map generation block 210 together with the segment molecule information and search section information for each robot. First, in the feature point extraction block 2102, the local information is extracted from each local information. Although feature points (eg, straight lines, curves, specific patterns, etc.) are extracted (step 410), extraction of such feature points may be performed using, for example, a separation and combination algorithm and / or a SIRF algorithm.

  Next, the merge matching block 2104 determines whether the local information is information on a nearby robot, that is, through matching between the search section information for each robot and the local information for each robot, as shown in FIG. Then, it is determined whether the video information corresponds to the overlapping areas OL1, OL2, OL3 (step 412). However, as a result of the determination, it is determined that the local information is not the information of the nearby robot (video information of the overlapping area). If so, the corresponding information is excluded from the merge target (step 414).

  If it is determined in the step (412) that the local information is information on a nearby robot (video information of the overlapping area), the merge matching block 2104 performs local information matching between the robots to match each other. Merge matching is performed to merge the feature points of the part to be merged (step 416). That is, the merge matching block 2104, after confirming the presence or absence of video information for the same information through the matching operation, merges the same information into one, otherwise excludes it from the merge target.

  Next, the map creation block 2106 creates the entire environment map for the search range by fusing the local information acquired by each robot using the extracted feature points and the merged feature points (step 418).

  In the above description, a preferred embodiment of the present invention has been presented and described. However, the present invention is not necessarily limited thereto, and any person having ordinary knowledge in the technical field to which the present invention belongs can be used. It can be easily understood that various substitutions, modifications and changes can be made without departing from the technical idea of the present invention.

202 Interface block 204 Relative position acquisition block 206 Search section division block 208 Local information collection block 210 Integrated environment map generation block 2102 Feature point extraction block 2104 Merge matching block 2106 Map generation block

Claims (20)

Measuring the relative distance between each robot to detect the relative position between each robot;
Dividing the search area for each robot based on the detected relative position;
Transmitting the divided search area information to the corresponding robots;
Obtaining local information used for generating a map from each robot;
A method of constructing a map using intelligent robots of the crowd, including a process of generating an entire environmental map by fusing each acquired local information. The relative distance is
The map construction method using the intelligent robot of the crowd according to claim 1, wherein the map construction is measured through a TWR technique. The relative position is
The map construction method using the intelligent robot of the crowd according to claim 2, wherein the coordinate values are generated based on a relative distance between the robots. Each of the divided search areas is
The map construction method using the intelligent robot of the crowd according to claim 1, wherein adjacent areas are designated to overlap. The local information is
The map construction method using the intelligent robot of the crowd according to claim 1, comprising ultrasonic waves, infrared rays, lasers, cameras, and odometry. The process of generating the entire environmental map is as follows:
A process of extracting feature points from each local information;
Checking whether each local information is local information of a nearby robot;
When it is information of the nearby robot, the process of performing matching between the corresponding robots and merging the feature points of the matching parts,
The map construction method using the intelligent robot according to claim 1, further comprising: creating the entire environment map using the extracted feature points and the merged feature points. The feature points are extracted as follows:
The map construction method using the intelligent robot of the crowd according to claim 6, wherein the map construction is performed through a separation and combination algorithm. The feature points are extracted as follows:
The map construction method using the intelligent robot of the crowd according to claim 6, wherein the map construction is performed through a SIRF algorithm. The feature points are
The map construction method using the intelligent robot of the crowd according to claim 7 or 8, wherein the map is a feature point with respect to a landmark or a specific pattern. The entire environmental map is
The map construction method using the intelligent robot according to claim 1, wherein the map construction method is created using a SLAM algorithm. A relative position acquisition block that detects and shares the relative position between each robot;
A search area dividing block that divides the search area for each robot based on the detected relative position;
A local information collection block for acquiring local information used for generating a map from each robot;
A map construction apparatus using an intelligent robot of a crowd comprising: an integrated environment map generation block that generates an entire environment map by fusing each acquired local information. The relative position acquisition block is:
The map construction apparatus using the intelligent robot of the crowd according to claim 11, wherein the relative position is detected through a relative distance measured through a TWR technique. The relative position is
The map construction apparatus using the intelligent robot of the crowd according to claim 12, wherein the coordinate values are generated based on a relative distance between the robots. The search area division block is:
The map construction apparatus using the intelligent robot according to claim 11, wherein the search areas are specified so that adjacent areas overlap each other. The local information is
The map construction apparatus using the intelligent robot of the crowd according to claim 11, comprising an ultrasonic wave, an infrared ray, a laser, a camera, and odometry. The integrated environment map generation block includes:
A feature point extraction block for extracting feature points from the local information;
When the local information is information on nearby robots, a matching matching block that performs matching between the corresponding robots and merges feature points of parts that match each other, and
The map construction apparatus using the intelligent robot according to claim 11, further comprising: a map creation block that creates the entire environment map using the extracted feature points and the merged feature points. The feature point extraction block includes:
The map construction apparatus using the intelligent robot of the crowd according to claim 16, wherein the feature points are extracted through a separation and combination algorithm. The feature point extraction block includes:
The map construction apparatus using the crowd intelligent robot according to claim 16, wherein the feature points are extracted through an SIRF algorithm. The feature points are
The map construction apparatus using the intelligent robot of the crowd according to claim 17 or 18, wherein the map construction apparatus is a feature point with respect to a landmark or a specific pattern. The integrated environment map generation block includes:
The map construction apparatus using the intelligent robot according to claim 11, wherein the entire environment map is generated using a SLAM algorithm.

Images