CN112393719B - Grid semantic map generation method and device and storage equipment - Google Patents
Grid semantic map generation method and device and storage equipment Download PDFInfo
- Publication number
- CN112393719B CN112393719B CN201910739998.XA CN201910739998A CN112393719B CN 112393719 B CN112393719 B CN 112393719B CN 201910739998 A CN201910739998 A CN 201910739998A CN 112393719 B CN112393719 B CN 112393719B
- Authority
- CN
- China
- Prior art keywords
- node
- map
- visual
- current
- relative position
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/005—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 with correlation of navigation data from several sources, e.g. map or contour matching
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B29/00—Maps; Plans; Charts; Diagrams, e.g. route diagram
- G09B29/003—Maps
- G09B29/005—Map projections or methods associated specifically therewith
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Navigation (AREA)
Abstract
The application provides a grid semantic map generation method, which comprises the following steps: generating a global grid map of the current environment, and determining the final position of a node in the current environment on the global grid map; in the process of generating the global grid map, obtaining the relative position of the visual semantic information of the current environment and the current node; and embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map. The map generated by the grid semantic map generation method provided by the application not only can reflect the position information of the object in the current environment, but also can reflect the image information of the object in the current environment, so that the condition of the corresponding environment of the map can be reflected more accurately.
Description
Technical Field
The application relates to the field of map construction, in particular to a method and a device for generating a grid semantic map and storage equipment.
Background
With the progress and development of science and technology, more and more self-moving devices with navigation functions appear in the daily life of people, and are used for providing services in the aspects of navigation and the like for people, such as: self-moving navigation robots, and the like. The working principle of the self-moving navigation robot is as follows: the method comprises the steps of obtaining a navigation instruction, displaying a pre-stored grid map of the environment according to the navigation instruction, obtaining a marked starting point and a marked target point on the grid map, generating a path plan from the starting point to the target point on the grid map according to the starting point and the target point, further generating the grid map-based navigation map according to the path plan, and performing self-movement according to a selected navigation route, so as to lead related personnel to the target point from the starting point.
A grid map used when a self-moving device having a self-moving navigation function performs route planning is generally generated from laser scan data obtained from a laser scanner device mounted on the self-moving device. The grid map can only reflect certain obstacle information in the working environment of the mobile device, and cannot accurately reflect the specific situation of the environment corresponding to the map.
Disclosure of Invention
The application provides a method for generating a grid semantic map, which can enable the generated map to reflect the situation of the corresponding environment of the map more accurately.
The application provides a grid semantic map generation method, which comprises the following steps:
generating a global grid map of the current environment, and determining the final position of each node in the current environment on the global grid map;
obtaining the relative position of the visual semantic information of the current environment and a current node in the process of generating a global grid map;
and embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map.
Optionally, the method further includes:
and the navigation map generating unit is used for generating the navigation map of the self-moving equipment according to the global grid semantic map.
Optionally, the generating a global grid map of the current environment and determining a final position of a node in the current environment on the grid map include:
step 1: taking a node where the mobile equipment is located at present as a current node;
and 2, step: obtaining laser scanning data acquired by the self-moving equipment at the current node;
and step 3: constructing a local grid map of the current node according to the laser scanning data acquired by the self-moving equipment at the current node and the laser scanning number acquired by the self-moving equipment at all nodes before the current node;
and 4, step 4: judging whether all nodes before the current node have similar nodes of the current node; if so, performing closed-loop optimization on the local grid map of the current node, and adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node;
and 5: and (4) repeating and sequentially executing the steps 1-4 until the current node is the final node, finishing the closed-loop optimization of the global grid map, and determining the final positions of all the nodes on the global grid map.
Optionally, the determining whether similar nodes of the current node exist in all nodes before the current node includes:
judging whether all nodes before the current node have similar nodes of the current node or not according to whether the similarity of the laser scanning data or the visual semantic information of the current node and any node before is equal to or greater than a similarity threshold value or not;
if the similarity is equal to or greater than the similarity threshold, the similar node of the current node exists; otherwise, the similar node of the current node does not exist.
Optionally, the obtaining the relative position between the visual semantic information of the current environment and the current node includes:
taking a node where the mobile equipment is located at present as a current node;
visual semantic information in the process that a mobile device moves from a current node to a next node of the current node is obtained, and the relative position of the visual semantic information and the current node is determined.
Optionally, the method further includes: establishing a mapping relation between the visual semantic information and the nodes;
the embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map, including:
and finding the visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map.
Optionally, before obtaining the relative position between the visual semantic information of the current environment and the current node, the method further includes:
obtaining visual information of the current environment through a visual recognition device, wherein the visual information is image information of an object in the current environment;
and identifying the image information to obtain the visual semantic information of the current environment.
Optionally, the obtaining of the relative position between the visual semantic information of the current environment and the current node further includes;
obtaining a relative position of the visual recognition device and the current node;
obtaining the relative position of the visual identification device and the visual semantic information of the current environment;
and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the current node and the relative position of the visual recognition device and the semantic information of the current environment.
Optionally, marking a starting point and a target point of the self-moving device in the global grid semantic map;
generating a path plan from the starting point to the target point in the global grid semantic map according to the starting point to the target point;
and constructing the navigation map according to the path plan.
In another aspect of the present application, a grid semantic map generating apparatus is provided, including:
the final position determining unit is used for generating a global grid map of the current environment and determining the final position of each node in the current environment on the global grid map;
the relative position obtaining unit is used for obtaining the relative position of the visual semantic information of the current environment and the current node in the process of generating the global grid map;
and the grid semantic map generating unit is used for embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map.
In another aspect of the present application, there is provided a storage device storing a program of a grid semantic map generating method, the program being executed by a processor and performing the steps of:
generating a global grid map of the current environment, and determining the final position of each node in the current environment on the global grid map;
in the process of generating a global grid map, obtaining the relative position of the visual semantic information of the current environment and a current node;
and embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map.
Compared with the prior art, the method has the following advantages:
according to the grid semantic map generating method, in the process of generating the global grid map, the relative positions of the visual semantic information of the current environment and the nodes are obtained; and embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map. The map generated by the grid semantic map generation method provided by the application not only can reflect the position information of the object in the current environment, but also can reflect the image information of the object in the current environment, so that the grid semantic map can more accurately reflect the condition of the corresponding environment of the map, and more effective path planning can be conveniently made.
Drawings
Fig. 1 is a flowchart of a method for generating a grid semantic map according to a first embodiment of the present disclosure.
Fig. 2 is a flowchart of a method for determining a node location according to a first embodiment of the present disclosure.
Fig. 3 is a flowchart of a method for generating a navigation map according to a first embodiment of the present application.
Fig. 4 is a schematic diagram of a grid semantic map generating apparatus according to a second embodiment of the present application.
Fig. 5 is a schematic diagram of a storage device usage environment according to a third embodiment of the present application.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. This application is capable of implementation in many different ways than those herein set forth and of similar import by those skilled in the art without departing from the spirit of this application and is therefore not limited to the specific implementations disclosed below.
A first embodiment of the present application provides a method for generating a grid semantic map, which is described below with reference to fig. 1 to 3.
As shown in fig. 1, in step S101, a global grid map of the current environment is generated, and the final positions of the nodes in the current environment on the global grid map are determined.
The process of generating the global grid map comprises the following steps: firstly, laser scanning data of the current environment are collected in the moving process of the self-moving equipment through a laser scanning device installed on the self-moving equipment. The current environment is a working environment where the self-moving device is located, taking the self-moving device as an example of the self-moving navigation robot, if the self-moving navigation robot is used in a shopping mall, the current environment is the environment of the shopping mall where the self-moving navigation robot is located, and if the self-moving navigation robot is used in an airport, the current environment is the environment of the airport where the self-moving navigation robot is located. Then, during the self-moving process of the self-moving device in the current environment, a local grid map is continuously generated at nodes in the current environment by using laser scanning data. And finally, when the current node is the final node in the current environment, generating a global grid map.
In addition, the local grid map or the global grid map may be a two-dimensional grid map or a three-dimensional grid map. The local grid map is generated at the current node before the mobile device reaches the final node in the current environment, and the global grid map is generated at the final node after the mobile device reaches the final node in the current environment. Each node in the current environment is a node selected by related personnel in the current environment in advance, or a node which is automatically determined by self-moving equipment according to a preset node generation rule after the self-moving equipment moves for a certain distance and angle, the node is used as a reference point, the position information of an object in the current environment is expressed by the relative position of the object and a certain node, and the relative position of the object and the certain node is fixed in the process of generating the global grid map according to laser scanning data.
In the process of generating the global grid map according to the laser scanning data, when the mobile device moves to a certain node for the second time due to the movement deviation of the mobile device and the like, the corresponding position of the node on the grid map may be different. Therefore, in the process of generating the global grid map, whether the current node is the same as the previous node or not needs to be continuously judged according to the laser scanning data of the nodes, if so, the grid map is locally optimized, and the positions of the nodes including the current node and the nodes before the current node on the grid map are adjusted, so that in the process of generating the global grid map, the positions of the nodes on the grid map are continuously changed, and the final positions of all the nodes in the current environment on the global grid map can be determined after the global map is subjected to closed-loop optimization.
It should be noted that the grid map in the present application is also referred to as a grating image, and is an occupied grid map constructed by a laser sensing device, and the specific construction method is as follows: dividing the working environment of the mobile equipment into a series of grids, and giving a gray value to each grid according to the laser scanning data obtained by the laser scanning device, wherein the gray value represents the probability of the grid being occupied. The specific process of giving a gray value to each grid according to the laser scanning data obtained by the laser scanning device is as follows: pixel values of an object in a working environment of the mobile device are obtained according to the laser scanning data, and a gray value is given to a grid according to the pixel values.
There may be many possible schemes for the specific process of generating a global grid map of the current environment and determining the final positions of the nodes in the current environment on the grid map, and fig. 2 shows one preferred embodiment; as described in detail below in conjunction with fig. 2.
Step S101-1: and taking the node where the mobile equipment is located as the current node.
Step S101-2: laser scan data obtained from a mobile device at a current node is obtained.
Step S101-3: constructing a local grid map of a current node, comprising: constructing a local grid map of the current node according to laser scanning data acquired by the mobile equipment at the current node and laser scanning numbers acquired by the mobile equipment at all nodes before the current node;
step S101-4: adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node, comprising: judging whether all nodes before the current node have similar nodes of the current node; if so, performing closed-loop optimization on the local grid map of the current node, and adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node;
step S101-5: and generating a global grid map of the current environment at the final node, wherein the global grid map comprises the following steps: and (4) repeating and sequentially executing the steps S101-1 to S101-4 until the current node is the final node, finishing the closed loop optimization of the global grid map, and determining the final positions of all the nodes on the global grid map.
The above steps S101-1 to S101-5 are a specific implementation method of the step S101; the following describes the implementation of the method in a practical task.
Assuming that the operation of step S101 is currently required to be performed on a working scene by a self-moving device equipped with a vision device and a laser scanning device, the specific implementation procedure is as follows.
The initial node is taken as the current node (i.e. equivalent to step S101-1), and when the initial node is the current node, no other node exists before the current node. After obtaining the laser scanning data at the initial node, it is necessary to construct a local grid map of the current node according to the laser scanning data at the initial node (i.e., equivalent to step S101-3).
After the self-moving device constructs the local grid of the current node at the initial node according to the laser scanning data, the self-moving device will continue to move to the next node of the initial node, and at this time, the node is the new current node (i.e., step S101-1). The self-moving device will continue to obtain the laser scanning data of the node at the current node, i.e. the laser scanning data obtained from the mobile device at the current node (i.e. equivalent to step S101-2), and construct a local grid map of the current node according to the laser scanning data obtained from the mobile device at the current node and the laser scanning numbers obtained from the mobile device at all nodes before the current node (i.e. equivalent to step S101-3). Because the current node is not the initial node, after the local grid map of the current node is generated, whether similar nodes of the current node exist in all nodes before the current node needs to be judged; if yes, performing closed-loop optimization on the local grid map of the current node, and adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node (namely, step S101-4). Specifically, the step of judging whether all nodes before the current node have similar nodes of the current node includes: judging whether similar nodes of the current node exist in all nodes before the current node or not according to whether the similarity of laser scanning data or visual semantic information of the self-mobile equipment at the current node and any node before the current node is equal to or larger than a similarity threshold value or not; if the similarity is equal to or greater than the similarity threshold, the similar node of the current node exists; otherwise, the similar node of the current node does not exist.
After the local grid map of the current node is generated at the node next to the initial node, the self-moving device will continue to move to the node next to the current node, and at this time, the node next to the current node is the new current node. And repeatedly executing the step of obtaining the laser scanning data acquired from the mobile equipment at the current node; constructing a local grid map of the current node according to laser scanning data acquired by the mobile equipment at the current node and laser scanning numbers acquired by the mobile equipment at all nodes before the current node; judging whether all nodes before the current node have similar nodes of the current node; if yes, performing closed-loop optimization on the local grid map of the current node, adjusting the current node and positions of all nodes before the current node on the local grid map of the current node (namely, equivalent to step S101-4), and generating a global grid map of the current environment at the final node until the current node is the final node, and determining final positions of all nodes in the current environment on the global grid map (namely, equivalent to step S101-5).
The mode of generating the local grid map at the appointed node according to the laser scanning data and the mode of generating the global grid map according to the laser scanning data are as follows: the method comprises the steps of obtaining world coordinates of a current environment according to laser scanning data, converting the world coordinates of the current environment into coordinates in a grid map through coordinate conversion, determining a corresponding grid of the current environment on a local grid map or a global grid map (namely determining a corresponding position of the current environment on the local grid map or the global grid map), obtaining pixel values of objects corresponding to the grid according to the laser scanning data, and giving corresponding gray values to different grids according to the pixel values, so that the local grid map or the global grid map capable of indicating the positions of the objects in the current environment is generated. Since the laser scanning device is located at any position in the current environment, a reference coordinate system is selected in the current environment to describe the position of the laser scanning device, and coordinates in the reference coordinate system are used to describe the position of any object in the current environment, which is called a world coordinate system.
The process of determining the coordinates of the current environment in the world coordinate system is: obtaining the coordinates of the current environment in a laser coordinate system according to the laser scanning data; obtaining the coordinates of the laser scanning device in a world coordinate system according to the laser scanning data; and obtaining the coordinates of the current environment in the world coordinate system according to the coordinates of the current environment in the laser coordinate system and the coordinates of the laser scanning device in the world coordinate system. Specifically, firstly, a laser coordinate system with the laser scanning device as a coordinate origin is established, and coordinates of the current environment in the laser coordinate system are determined, wherein the laser coordinate system is used for representing the position relationship between the current environment and the laser scanning device. Then, a world coordinate system is established, and the coordinates of the laser scanning device in the world coordinate system are determined. And finally, obtaining the coordinates of the current environment in the world coordinate system through the rotation matrix and the translation vector. The process of obtaining the pixel value corresponding to the current environment according to the laser scanning data is the prior art, and is not described in detail herein.
As shown in fig. 1, in step S102, in the process of generating the global grid map, the relative position of the visual semantic information of the current environment and the current node is obtained.
Obtaining the relative position of the visual semantic information of the current environment and the current node, including: taking a node where the mobile equipment is located at present as a current node; obtaining visual semantic information in the process that the mobile equipment moves from a current node to a next node of the current node, and determining the relative position of the visual semantic information and the current node; and meanwhile, establishing a mapping relation between the visual semantic information and the current node. Before obtaining the relative position of the visual semantic information of the current environment and the current node, the method further comprises the following steps: visual information of a current environment is obtained through a visual recognition device, wherein the visual information is image information of an object in the current environment; identifying and processing the image information to obtain visual semantic information of the current environment; correspondingly, the obtaining of the relative position of the visual semantic information of the current environment and the current node comprises; obtaining the relative position of the visual recognition device and the current node; obtaining the relative position of the visual identification device and the visual semantic information of the current environment; and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the current node and the relative position of the visual recognition device and the semantic information of the current environment. Specifically, the process of obtaining the relative position of the visual recognition device and the node may be: firstly, determining and obtaining the relative position of a laser scanning device and a visual recognition device; then, determining the relative position of the laser scanning device and the node; and finally, indirectly obtaining the relative position of the visual recognition device and the node according to the relative position of the laser scanning device and the visual recognition device and the relative position of the laser scanning device and the node.
It should be noted that the visual information is image information of an object in the current environment. The visual semantic information of the current environment is semantic information which can be identified by a machine and corresponds to the visual information of the current environment, and image information is identified and processed to obtain the visual semantic information of the current environment as follows: visual information is identified and processed through a deep neural network algorithm, so that semantic information which can be read and identified by a machine and corresponds to the visual information is obtained, for example: name information of the object in the image of the object (for example, the object is a table, an elevator, etc.). The relative positions of the visual semantic information of the current environment and the nodes are as follows: the visual semantic information describes the relative position of the current environment and the node.
As shown in fig. 1, in step S103, visual semantic information is embedded in the global grid map according to the relative position, and the global grid semantic map is generated.
According to the relative position, embedding the visual semantic information into a global grid map to generate the global grid semantic map, wherein the method comprises the following steps: and finding the visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map. The grid map is used to reflect the position of the object existing in the current environment. Specifically, if there is an object at a certain position in the current environment, the grid map will reflect that there is an obstacle at the position. However, some objects in the current environment may not be obstacles, such as: straight ladders, movable objects, and the like. The visual information of the current environment is obtained, the image information of the object in the current environment can be obtained, the visual information is converted into the visual semantic information which can be recognized by a machine and corresponds to the global grid map, the generated grid semantic map can determine the position of the object in the current environment in the map, and the image information of the object, so that the map can more accurately reflect the condition of the environment corresponding to the map.
The method for generating a grid semantic map provided in the first embodiment of the present application further includes: and generating a navigation map of the self-mobile device according to the global grid semantic map.
The process of generating a navigation map from a mobile device based on a global grid semantic map is shown in fig. 3:
as shown in FIG. 3, in step S301, a start point 301A-1 and a target point 301A-2 are marked from the mobile device in the global grid semantic map 301A.
As shown in FIG. 3, in step S302, a path plan from a start point 301A-1 to a target point 301A-2 is generated in the global grid semantic map based on the start point 301A-1 to the target point 301A-2.
As shown in fig. 3, in step S303, a navigation map 301-B is constructed according to the route plan.
When a navigation map is generated according to the existing two-dimensional map, all scanned objects are considered as obstacles, and automatic obstacle avoidance is carried out during path planning. However, in the navigation map generated according to the global two-dimensional grid semantic map provided in the first embodiment of the present application, during path planning, it can be determined whether an object in the current environment is a real obstacle according to the visual semantic information of the current environment, some obstacles are actually movable, or may be pedestrians, and after moving to the area from the mobile device, the self-moving device can wait for the obstacle to move and then freely pass through the area, so as to reach the destination point from the starting point at the fastest speed; or the barrier is an elevator, and the robot can reach the area through the elevator when needing to reach different floors, so that navigation planning among different floors is realized. Therefore, in the grid semantic map, if a fixed obstacle is displayed according to the visual semantic information, the obstacle is avoided in the route planning; if the visual semantic information shows that the object is a movable object or a person, the object does not belong to the barrier and belongs to a passable area during path planning; in addition, when navigation to different floors is needed, the position information of the straight ladder can be obtained through semantic information, and therefore an optimal path is planned.
Application scenarios
In the application scene, the current environment is selected as a market, and the self-mobile equipment is selected as a self-mobile navigation robot for navigating customers in the market. Firstly, the self-moving navigation robot builds a grid semantic map of a market in advance. Then, when the customer interacts with the navigation robot through the voice interaction system, the self-moving navigation robot can mark a starting point and a target point on the grid semantic map through information obtained during interaction. And finally, planning a path from the starting point to the target point on the basis of the grid semantic map to generate a navigation map. After the navigation map is generated, the self-moving navigation robot displays the navigation map, a plurality of routes exist in the navigation map so that a customer can independently select the navigation route, and the customer is led to arrive at a target point from a starting point according to the navigation route selected by the customer.
When the navigation route is selected, the visual semantic information in the grid semantic map can reflect the types of various obstacles in the shopping mall, so that the navigation route is suitable to be determined; for example, in a grid semantic map, if an obstacle at a certain position is found to be a person through visual semantic information, it is considered that the person is active, and it is considered that the person can pass through the position after moving when selecting a route; or, if some obstacle is found to be the elevator room actually through the visual semantic information, the elevator room can be planned to pass through to other floors.
Second embodiment
In the foregoing first embodiment, a grid semantic map generating method is provided, and correspondingly, a second embodiment of the present application provides a grid semantic map generating apparatus. Since the apparatus embodiment is substantially similar to the method first embodiment, it is relatively simple to describe, and reference may be made to some descriptions of the method embodiments for relevant points. The device embodiments described below are merely illustrative.
Referring to fig. 4, a schematic diagram of a grid semantic map generating apparatus according to a second embodiment of the present application is shown.
The grid semantic map generating device comprises:
a final position determining unit 401, configured to generate a global grid map of a current environment, and determine a final position of each node in the current environment on the global grid map;
a relative position obtaining unit 402, configured to obtain a relative position between the visual semantic information of the current environment and a current node in a process of generating a global grid map;
a grid semantic map generating unit 403, configured to embed the visual semantic information into the global grid map according to the relative position, and generate a global grid semantic map.
Optionally, the method further includes:
and the navigation map generating unit is used for generating the navigation map of the self-moving equipment according to the global grid semantic map.
Optionally, the generating a global grid map of the current environment and determining a final position of a node in the current environment on the grid map include:
step 1: taking a node where the mobile equipment is located at present as a current node;
step 2: obtaining laser scanning data acquired by the self-mobile equipment at the current node;
and step 3: constructing a local grid map of the current node according to the laser scanning data acquired by the self-moving equipment at the current node and the laser scanning number acquired by the self-moving equipment at all nodes before the current node;
and 4, step 4: judging whether all nodes before the current node have similar nodes of the current node; if so, performing closed-loop optimization on the local grid map of the current node, and adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node;
and 5: and (4) repeating and sequentially executing the steps 1-4 until the current node is the final node, finishing the closed-loop optimization of the global grid map, and determining the final positions of all the nodes on the global grid map.
Optionally, the determining whether similar nodes of the current node exist in all nodes before the current node includes:
judging whether all nodes before the current node have similar nodes of the current node or not according to whether the similarity of the laser scanning data or the visual semantic information of the current node and any node before is equal to or greater than a similarity threshold value or not;
if the similarity is equal to or greater than the similarity threshold, the similar node of the current node exists; otherwise, the similar node of the current node does not exist.
Optionally, the obtaining the relative position between the visual semantic information of the current environment and the current node includes:
taking a node where the mobile equipment is located at present as a current node;
visual semantic information in the process that a mobile device moves from a current node to a next node of the current node is obtained, and the relative position of the visual semantic information and the current node is determined.
Optionally, the method further includes: establishing a mapping relation between the visual semantic information and the current node;
the embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map, including:
and finding the visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map.
Optionally, before obtaining the relative position between the visual semantic information of the current environment and the current node, the method further includes:
obtaining visual information of the current environment through a visual recognition device, wherein the visual information is image information of an object in the current environment;
and identifying the image information to obtain the visual semantic information of the current environment.
Correspondingly, the obtaining of the relative position of the visual semantic information of the current environment and the current node further includes;
obtaining the relative position of the visual recognition device and the current node;
obtaining the relative position of the visual identification device and the visual semantic information of the current link;
and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the node and the relative position of the visual semantic information of the current environment and the visual semantic information of the current environment.
Optionally, marking a starting point and a target point of the self-moving device in the global grid semantic map;
generating a path plan from the starting point to the target point in the global grid semantic map according to the starting point to the target point;
and constructing the navigation map according to the path plan.
Third embodiment
In correspondence with the method for generating a grid semantic map provided in the first embodiment of the present application, a third embodiment of the present application provides a storage device storing a program of the method for generating a grid semantic map, the program being executed by a processor; as a typical example, fig. 5 shows the storage device 502, and the processor 501 connected to the storage device 502; by reading the program stored in the storage device 502, the following steps can be performed:
generating a global grid map of the current environment, and determining the final position of each node in the current environment on the global grid map;
in the process of generating the global grid map, obtaining the relative position of the visual semantic information of the current environment and the current node;
and embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map.
It should be noted that, for the detailed description of the storage device provided in the third embodiment of the present application, reference may be made to the relevant description of the first embodiment of the present application, and details are not repeated here.
Although the present invention has been described with reference to the preferred embodiments, it should be understood that the scope of the present invention is not limited to the embodiments described above, and that various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the present invention.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
1. Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.
2. As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Claims (9)
1. A method for generating a grid semantic map, comprising:
generating a global grid map of the current environment, and determining the final position of each node in the current environment on the global grid map;
obtaining the relative position of the visual semantic information of the current environment and a current node in the process of generating a global grid map;
embedding the visual semantic information into the global grid map according to the relative position to generate a global grid semantic map;
obtaining the relative position of the visual semantic information of the current environment and the current node comprises: establishing a mapping relation between the visual semantic information and the current node;
finding out visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map;
the obtaining of the relative position of the visual semantic information of the current environment and the current node further includes:
obtaining a relative position of a visual recognition device and the current node;
obtaining the relative position of the visual identification device and the visual semantic information of the current environment;
and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the current node and the relative position of the visual recognition device and the semantic information of the current environment.
2. The grid semantic map generating method according to claim 1, further comprising:
and generating a navigation map of the mobile equipment according to the global grid semantic map.
3. The grid semantic map generating method according to claim 1, wherein the generating a global grid map of the current environment and determining final positions of nodes in the current environment on the grid map comprises:
step 1: taking a node where the mobile equipment is located at present as a current node;
step 2: obtaining laser scanning data acquired by the self-mobile equipment at the current node;
and 3, step 3: constructing a local grid map of the current node according to the laser scanning data acquired by the self-moving equipment at the current node and the laser scanning number acquired by the self-moving equipment at all nodes before the current node;
and 4, step 4: judging whether all nodes before the current node have similar nodes of the current node; if so, performing closed-loop optimization on the local grid map of the current node, and adjusting the positions of the current node and all nodes before the current node on the local grid map of the current node;
and 5: and (4) repeating and sequentially executing the steps 1-4 until the current node is the final node, finishing the closed-loop optimization of the global grid map, and determining the final positions of all the nodes on the global grid map.
4. The grid semantic map generating method according to claim 3, wherein the determining whether similar nodes of the current node exist in all nodes before the current node comprises:
judging whether all nodes before the current node have similar nodes of the current node according to whether the similarity of the laser scanning data or the visual semantic information between the current node and any node before is equal to or greater than a similarity threshold;
if the similarity is equal to or greater than the similarity threshold, the similar node of the current node exists; otherwise, the similar node of the current node does not exist.
5. The grid semantic map generating method according to claim 1, wherein the obtaining of the relative position of the visual semantic information of the current environment and the current node comprises:
taking a node where the mobile equipment is located at present as a current node;
visual semantic information in the process that a mobile device moves from a current node to a next node of the current node is obtained, and the relative position of the visual semantic information and the current node is determined.
6. The grid semantic map generating method according to claim 1, further comprising, before obtaining the relative position of the visual semantic information of the current environment and the current node:
obtaining visual information of the current environment through a visual recognition device, wherein the visual information is image information of an object in the current environment;
and identifying the image information to obtain the visual semantic information of the current environment.
7. The grid semantic map generating method according to claim 2, characterized in that:
marking a starting point and a target point of the self-moving equipment in the global grid semantic map;
generating a path plan from the starting point to the target point in the global grid semantic map according to the starting point to the target point;
and constructing the navigation map according to the path plan.
8. A grid semantic map generating apparatus, comprising:
the final position determining unit is used for generating a global grid map of the current environment and determining the final position of each node in the current environment on the global grid map;
the relative position obtaining unit is used for obtaining the relative position of the visual semantic information of the current environment and the current node in the process of generating the global grid map;
the grid semantic map generating unit is used for embedding the visual semantic information into the global grid map according to the relative position to generate the global grid semantic map;
obtaining the relative position of the visual semantic information of the current environment and the current node comprises: establishing a mapping relation between the visual semantic information and the current node;
finding visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map;
the obtaining of the relative position of the visual semantic information of the current environment and the current node further includes:
obtaining a relative position of a visual recognition device and the current node;
obtaining the relative position of the visual identification device and the visual semantic information of the current environment;
and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the current node and the relative position of the visual recognition device and the semantic information of the current environment.
9. A storage device storing a program of a grid semantic map generating method, the program being executed by a processor and executing the steps of:
generating a global grid map of the current environment, and determining the final position of each node in the current environment on the global grid map;
in the process of generating the global grid map, obtaining the relative position of the visual semantic information of the current environment and the current node;
embedding the visual semantic information into the global grid map according to the relative position to generate a global grid semantic map;
obtaining the relative position of the visual semantic information of the current environment and the current node comprises: establishing a mapping relation between the visual semantic information and the current node;
finding visual semantic information corresponding to the node according to the mapping relation between the visual semantic information and the node, and embedding the visual semantic information into the global grid map according to the relative position between the visual semantic information and the node to generate the global grid semantic map;
the obtaining of the relative position of the visual semantic information of the current environment and the current node further includes:
obtaining the relative position of a visual recognition device and the current node;
obtaining the relative position of the visual identification device and the visual semantic information of the current environment;
and obtaining the relative position of the visual semantic information of the current environment and the current node according to the relative position of the visual recognition device and the current node and the relative position of the visual recognition device and the semantic information of the current environment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910739998.XA CN112393719B (en) | 2019-08-12 | 2019-08-12 | Grid semantic map generation method and device and storage equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910739998.XA CN112393719B (en) | 2019-08-12 | 2019-08-12 | Grid semantic map generation method and device and storage equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112393719A CN112393719A (en) | 2021-02-23 |
| CN112393719B true CN112393719B (en) | 2023-04-14 |
Family
ID=74602253
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910739998.XA Active CN112393719B (en) | 2019-08-12 | 2019-08-12 | Grid semantic map generation method and device and storage equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112393719B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114310908B (en) * | 2022-01-25 | 2023-10-24 | 深圳市优必选科技股份有限公司 | Robot control method, robot control device and robot |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780735A (en) * | 2016-12-29 | 2017-05-31 | 深圳先进技术研究院 | A kind of semantic map constructing method, device and a kind of robot |
| CN109084749A (en) * | 2018-08-21 | 2018-12-25 | 北京云迹科技有限公司 | The method and device of semantic positioning is carried out by object in environment |
| CN109736236A (en) * | 2018-12-31 | 2019-05-10 | 广州赛特智能科技有限公司 | A kind of the semantic information system and construction method of the unmanned sweeper Driving Scene in garden |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101242252B1 (en) * | 2011-02-10 | 2013-03-11 | 고려대학교 산학협력단 | Method for building simantic grid map and method for exploring an environment using simantic grid map |
| CN105628034B (en) * | 2016-02-04 | 2019-04-23 | 合肥杰发科技有限公司 | Navigation map update method and equipment |
| CN105843223B (en) * | 2016-03-23 | 2018-11-20 | 东南大学 | A kind of mobile robot three-dimensional based on space bag of words builds figure and barrier-avoiding method |
| CN107145578B (en) * | 2017-05-08 | 2020-04-10 | 深圳地平线机器人科技有限公司 | Map construction method, device, equipment and system |
| CN107677279B (en) * | 2017-09-26 | 2020-04-24 | 上海思岚科技有限公司 | Method and system for positioning and establishing image |
| CN108256574B (en) * | 2018-01-16 | 2020-08-11 | 广东省智能制造研究所 | Robot positioning method and device |
| CN108920584B (en) * | 2018-06-25 | 2020-10-30 | 广州视源电子科技股份有限公司 | Semantic grid map generation method and device |
| CN109163731A (en) * | 2018-09-18 | 2019-01-08 | 北京云迹科技有限公司 | A kind of semanteme map constructing method and system |
| CN109557928A (en) * | 2019-01-17 | 2019-04-02 | 湖北亿咖通科技有限公司 | Automatic driving vehicle paths planning method based on map vector and grating map |
-
2019
- 2019-08-12 CN CN201910739998.XA patent/CN112393719B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780735A (en) * | 2016-12-29 | 2017-05-31 | 深圳先进技术研究院 | A kind of semantic map constructing method, device and a kind of robot |
| CN109084749A (en) * | 2018-08-21 | 2018-12-25 | 北京云迹科技有限公司 | The method and device of semantic positioning is carried out by object in environment |
| CN109736236A (en) * | 2018-12-31 | 2019-05-10 | 广州赛特智能科技有限公司 | A kind of the semantic information system and construction method of the unmanned sweeper Driving Scene in garden |
Non-Patent Citations (1)
| Title |
|---|
| 秦玉鑫 ; 张高峰 ; 王裕清 ; .针对复杂环境的模块化栅格地图构建算法.控制工程.2016,(10),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112393719A (en) | 2021-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11846511B2 (en) | Sampling method and system for path planning of mobile robot in man-machine environment | |
| CN112859859B (en) | Dynamic grid map updating method based on three-dimensional obstacle object pixel object mapping | |
| JP5685380B2 (en) | Route generation apparatus using grid map and operation method thereof | |
| KR101427186B1 (en) | Device for searching area and mapping for path of intelligent robot in unknown environments | |
| Ivanov et al. | Effective informational entropy reduction in multi-robot systems based on real-time TVS | |
| CN112015187B (en) | Semantic map construction method and system for intelligent mobile robot | |
| CN109737980A (en) | A kind of air navigation aid and its corresponding robot | |
| CN113052761B (en) | Laser point cloud map fusion method, device and computer readable storage medium | |
| Xu et al. | An efficient algorithm for environmental coverage with multiple robots | |
| CN112393719B (en) | Grid semantic map generation method and device and storage equipment | |
| CN114911228A (en) | Robot path planning method and device and robot | |
| CN113848893A (en) | Robot navigation method, device, equipment and storage medium | |
| US20220019227A1 (en) | Route planning apparatus, route planning method, and computer-readable recording medium | |
| Mantha et al. | Investigating the fiducial marker network characteristics for autonomous mobile indoor robot navigation using ROS and Gazebo | |
| Parikh et al. | Autonomous mobile robot for inventory management in retail industry | |
| KR102009479B1 (en) | Apparatus and method for controlling mobile robot | |
| ELzaiady et al. | Next-best-view planning for environment exploration and 3D model construction | |
| Saigol et al. | Efficient search for known objects in unknown environments using autonomous indoor robots | |
| Ali et al. | Global mobile robot path planning using laser simulator | |
| EP4332715A1 (en) | Obstacle avoidance method and apparatus for self-walking device, and medium and electronic device | |
| Portugal et al. | Retrieving topological information for mobile robots provided with grid maps | |
| CN108789420A (en) | It is a kind of when causing danger for evacuating the robot and storage medium that guide | |
| WO2024033801A1 (en) | A method and system for mapping a real-world environment | |
| EP4328546A1 (en) | Determining motion in indoor space using probabilistic representation of motion transition | |
| CN116300843A (en) | Obstacle detouring target point determining method and device, intelligent equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |