Mobile robot map point location automatic labeling method and related equipment

Abstract

The invention discloses a mobile robot map point location automatic labeling method and related equipment, relates to the field of robot path planning, and mainly aims to solve the problem that mobile robot map point location automatic labeling is low in precision and excessively depends on manpower. The method comprises the following steps: identifying first key node information based on an identification graph of a target area to construct a first topological graph; collecting environment data to identify second key node information so as to construct a second topological graph; and determining annotation information of the target area based on a matching result of the first topological graph and the second topological graph.

Description

Technical Field

[0001] The present invention relates to the field of robot path planning, and in particular, to a method and related device for automatically marking map points of a mobile robot. Background Art

[0002] Generally, after purchasing a robot, in order to enable the robot to perform tasks in a specific scenario, a series of preparatory and configuration work such as on-site deployment and debugging of the robot is required. One of the most important tasks is the marking of map points of the robot. Map point marking refers to the process of marking and recording specific positions or features in the robot map. These points usually include obstacles, target points, path nodes, etc., which are crucial for the navigation and positioning of the robot. Through point marking, the robot can more accurately understand its surrounding environment and thus execute tasks more efficiently. Existing map point marking methods include manual marking and automatic marking.

[0003] However, manual marking requires a large amount of manpower. Especially in a complex environment, the workload of the operator will increase significantly; while the accuracy of automatic marking is often not as good as that of manual marking. Especially in an environment with complex conditions, large light changes or dynamic obstacles, the robot may misidentify or miss key points, resulting in inaccurate marking. Summary of the Invention

[0004] In view of the above problems, the present invention provides a method and related device for automatically marking map points of a mobile robot, mainly aiming to solve the problems of low accuracy of automatic marking of map points of a mobile robot and over-reliance on manual work.

[0005] To solve the above at least one technical problem, in a first aspect, the present invention provides a method for automatically marking map points of a mobile robot, the method comprising:

[0006] Identifying first key node information based on an identification map of a target area to construct a first topological graph;

[0007] Collecting environmental data to identify second key node information to construct a second topological graph;

[0008] Determining the marking information of the target area based on the matching result of the first topological graph and the second topological graph.

[0009] Optionally, the first key node information includes theoretical point legend information, theoretical room information, and theoretical corridor information. The identifying first key node information based on an identification map of a target area to construct a first topological graph includes:

[0010] Identifying the identification map of the target area through a target detection algorithm to determine the point legend information;

[0011] Identify the identification map of the target area through the OCR algorithm to determine the room information;

[0012] Identify the identification map of the target area through the skeleton line extraction algorithm to determine the corridor information;

[0013] Construct a first topological map based on the point legend information, the room information, and the corridor information.

[0014] Optionally, the second key node information includes actual point legend information, actual room information, and actual corridor information. The acquisition of environmental data to identify the second key node information and construct the second topological map includes:

[0015] Collect environmental data based on sensors;

[0016] Identify the environmental data to obtain the actual point legend information and the actual room information;

[0017] Project the actual point legend information and the actual room information to form a 2D map;

[0018] Perform skeleton line extraction on the 2D map to obtain the actual corridor information;

[0019] Construct the second topological map based on the actual point legend information, the actual room information, and the actual corridor information.

[0020] Optionally, the determination of the annotation information of the target area based on the matching result of the first topological map and the second topological map includes:

[0021] Analyze the node attributes and edge attributes of the first topological map and the second topological map based on the graph isomorphism algorithm to obtain the matching result;

[0022] Map the matching result as the annotation information of the target area to the robot map.

[0023] Optionally, the determination of the annotation information of the target area based on the matching result of the first topological map and the second topological map includes:

[0024] During the matching process of the first topological map and the second topological map, preferentially match the uniquely-attribute nodes;

[0025] After the uniquely-attribute nodes are matched, match the similar-attribute nodes based on the adjacent nodes.

[0026] Optionally, the above method further includes:

[0027] Perform a rough match on the first topological map and the second topological map first, and

[0028] Perform local matching on the first topological map and the second topological map preferentially to determine subgraphs with higher similarity.

[0029] Optionally, the above method further includes:

[0030] Overlay the first topological map, the second topological map, and the robot map to determine abnormal nodes;

[0031] Modify the abnormal nodes based on the manual intervention interface.

[0032] In a second aspect, an embodiment of the present invention further provides a device for automatically annotating map points of a mobile robot, including:

[0033] A first construction unit, configured to identify first key node information based on an identification map of a target area to construct a first topological map;

[0034] A second construction unit, configured to collect environmental data to identify second key node information to construct a second topological map;

[0035] A determination unit, configured to determine annotation information of the target area based on a matching result of the first topological map and the second topological map.

[0036] To achieve the above object, according to a third aspect of the present invention, there is provided a computer-readable storage medium, where the computer-readable storage medium includes a stored program, and when the above program is executed by a processor, the steps of the above method for automatically annotating map points of a mobile robot are implemented.

[0037] To achieve the above object, according to a fourth aspect of the present invention, there is provided an electronic device, including at least one processor and at least one memory connected to the processor; wherein, the above processor is configured to call program instructions in the above memory to execute the steps of the above method for automatically annotating map points of a mobile robot.

[0038] With the above technical solutions, for the problems of low accuracy and excessive dependence on manual work in the automatic annotation of mobile robot map points, the automatic annotation method and related devices for mobile robot map points provided by the present invention construct a first topological map by identifying first key node information based on an identification map of a target area; collect environmental data to identify second key node information to construct a second topological map; and determine the annotation information of the target area based on the matching result of the first topological map and the second topological map. In the above solution, high-precision automatic annotation without manual intervention is achieved through algorithms. Compared with traditional manual annotation methods, the annotation efficiency is greatly improved, and human errors are avoided; compared with pure automatic annotation technologies, the point recognition accuracy is greatly improved by the standardization of the identification map of the target area and the topological structure constraint. This method is particularly suitable for complex scenarios (such as large hotels and shopping malls), can quickly complete the point annotation of an environment of tens of thousands of square meters, and provide reliable semantic information support for robot navigation and path planning.

[0039] Correspondingly, the automatic annotation device, equipment, and computer-readable storage medium for mobile robot map points provided by the embodiments of the present invention also have the above technical effects.

[0040] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention are specifically given below. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0042] Figure 1 A flowchart showing the process of an automatic annotation method for mobile robot map points provided by an embodiment of the present invention is shown;

[0043] Figure 2 A schematic block diagram showing the composition of an automatic annotation device for mobile robot map points provided by an embodiment of the present invention is shown;

[0044] Figure 3 A schematic block diagram showing the composition of an automatic annotation electronic device for mobile robot map points provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present invention can be more thoroughly understood and the scope of the present invention can be completely conveyed to those skilled in the art.

[0046] To solve the problem that the automatic annotation accuracy of map points of a mobile robot is relatively low and overly dependent on manual work, an embodiment of the present invention provides a method for automatically annotating map points of a mobile robot, as Figure 1 shown, the method includes:

[0047] S101. Identify first key node information based on an identification map of a target area to construct a first topological map;

[0048] Exemplarily, the above-mentioned target area can be a complex area such as a large hotel or a shopping mall, and the above-mentioned identification map can be a map with area indication properties such as a fire evacuation map.

[0049] Specifically, use a high-resolution camera or scanner to capture the identification map to ensure that the image is clear and distortion-free. Perform preprocessing on the image, including operations such as denoising, binarization, and perspective correction, to improve the recognition accuracy of subsequent algorithms.

[0050] In one embodiment, the first key node information includes theoretical point legend information, theoretical room information, and theoretical corridor information. The identifying first key node information based on the identification map of the target area to construct a first topological map includes:

[0051] Identify the point legend information of the target area by a target detection algorithm to identify the identification map of the target area;

[0052] Identify the room information by an OCR algorithm to identify the identification map of the target area;

[0053] Identify the corridor information by a skeleton line extraction algorithm to identify the identification map of the target area;

[0054] Construct a first topological map based on the point legend information, the room information, and the corridor information.

[0055] First, the present application uses a deep learning-based target detection algorithm (such as the YOLO series, etc.) to analyze the identification map and identify the legends of key points such as doors and elevators. Convert the detection result into machine-readable coordinate information, denoted as node A, that is, the theoretical point legend information and the theoretical room information (including the positions and attribute information of doors and elevators).

[0056] Secondly, use the OCR (Optical Character Recognition) algorithm (such as a Transformer-based model) to recognize the room numbers in the fire evacuation map. Associate the recognized room numbers with the door position information extracted by the object detection algorithm to form node A with semantic information (for example, "Room 101" corresponds to the position of a certain door).

[0057] Thirdly, use a skeleton line extraction algorithm (such as the Zhang-Suen algorithm) to process the corridor area in the fire evacuation map and extract the skeleton line of the corridor. Extract key points from the skeleton line, including corridor inflection points (denoted as B) and corridor center points (denoted as C), which are the above-mentioned theoretical corridor information. These key points will be used to construct the edges of the topological graph to describe the connection relationship between points.

[0058] Finally, this application uses node A (doors, elevators), node B (corridor inflection points), and node C (corridor center points) as the nodes of the graph. According to the path information in the identification map, connect nodes A-A, A-B, and A-C to form edges. Finally, construct the topological graph M1 of the identification map, which is the above-mentioned first topological graph, and this graph describes the overall structure of the target area.

[0059] Exemplarily, the identification map of the fire evacuation Figure 1 class is usually presented in the form of a 2D floor plan, which contains legends such as doors and elevators. The object detection algorithm can quickly locate the positions of these legends in the identification map and convert them into machine-readable coordinate information. This step provides the basic data for subsequent graph isomorphism matching.

[0060] Exemplarily, the OCR algorithm is mainly used in this solution to extract theoretical room information such as room numbers from the identification map. The room numbers in the identification map are usually marked in text form, and the OCR algorithm can convert this text information into structured data, providing semantic support for the annotation of map points and improving the annotation quality and practicality of map points.

[0061] Based on the above solution, the OCR algorithm extracts room number information from the identification map through steps such as image preprocessing, text detection, and character recognition. For example, the algorithm can recognize room numbers such as "101" and "202" and associate them with the door position information extracted by the object detection algorithm. This process not only improves the accuracy of automatic annotation of map points but also provides semantic map information for the robot, enabling it to better understand the environment.

[0062] It is understandable that modern OCR algorithms (such as Transformer-based models) have powerful multi-language and multi-font recognition capabilities and can adapt to fire evacuation map standards in different countries and regions. In addition, the OCR algorithm can further improve the recognition accuracy through post-processing techniques (such as error correction algorithms).

[0063] S102. Collect environmental data to identify the second key node information and construct a second topological map;

[0064] In one embodiment, the second key node information includes actual point legend information, actual room information, and actual corridor information. The process of collecting environmental data to identify the second key node information and construct a second topological map includes:

[0065] Collect environmental data based on sensors;

[0066] Identify the environmental data to obtain the actual point legend information and the actual room information;

[0067] Project the actual point legend information and the actual room information to form a 2D map;

[0068] Extract the skeleton lines from the 2D map to obtain the actual corridor information;

[0069] Construct the second topological map based on the actual point legend information, the actual room information, and the actual corridor information.

[0070] First, during the mapping process, the robot uses sensors (such as cameras, lidar, etc.) carried on it to collect environmental data in real time. The object detection algorithm is used to identify doors and elevators in the scene, and the detection results are projected onto the robot's 2D map to form node A', that is, the actual point legend information and the actual room information (including the positions and attribute information of doors and elevators).

[0071] Secondly, perform skeleton line extraction on the 2D map constructed by the robot to find the skeleton lines in the corridor area. Extract the corridor inflection points (denoted as B') and the corridor center points (denoted as C'), that is, the actual corridor information, as the key nodes of the robot map.

[0072] Finally, use node A' (doors, elevators), node B' (corridor inflection points), and node C' (corridor center points) as the nodes of the graph. According to the path information in the robot map, connect nodes A'-A', A'-B', and A'-C' to form edges. Finally, construct the topological map M2 of the robot map, that is, the above-mentioned second topological map, which describes the environmental structure perceived by the robot.

[0073] Exemplarily, the above-mentioned object detection algorithm is mainly used to extract key point information from the environmental data and identification map sensed by the robot. The object detection algorithm is based on deep learning models (such as YOLO series, etc.) and can efficiently and accurately identify target objects in the scene.

[0074] Exemplarily, the robot collects environmental data in real time through sensors such as cameras or lidar carried on it. The object detection algorithm processes these environmental data, identifies the key points in the scene, that is, the above-mentioned actual point legend information, actual room information, and actual corridor information, and projects them onto the self-built map of the robot to form a semantic map. Key points usually include, for example, doors, elevators, etc., which are important reference information for map construction and navigation. Through the training of the deep learning model, the algorithm can adapt to the forms of doors and elevators in different environments (such as single-leaf doors, double-leaf doors, etc.) and has strong anti-interference capabilities (such as light changes, occlusions, etc.).

[0075] The algorithms for extracting skeleton lines involved in the above first topology map and second topology map are specifically as follows:

[0076] First, perform image preprocessing, including: binaryzation, converting the map image into a binary image to ensure that the corridor area is white and other areas are black; denoising, using morphological operations (such as opening operation, closing operation) to remove noise and small areas in the image; filling holes, using a filling algorithm to ensure that the corridor area is connected.

[0077] Secondly, perform skeleton line extraction. The core idea is to gradually remove the pixels at the edges of the image through iterative erosion operations until only a one-pixel-wide line remains. Specifically, it can be the Zhang-Suen algorithm: This algorithm extracts the skeleton line through two iterative erosion operations (processing boundary pixels in different directions respectively). In each iteration, the algorithm checks the 8-neighborhood of each pixel to determine whether the deletion condition is met. If it is met, the pixel is marked as to be deleted. Repeat the iteration until no pixel can be deleted.

[0078] Thirdly, perform key point extraction, including: corridor inflection point extraction, performing corner detection (such as Harris corner detection or Shi-Tomasi corner detection) on the skeleton line image to extract the inflection points of the corridor. Inflection points are usually located at the turns or intersections of the corridor; corridor center point extraction, performing distance transformation (Distance Transform) on the skeleton line image, calculating the distance from each pixel to the nearest background pixel. Find the local maximum points in the distance transformation graph. These points are usually located on the center line of the corridor. Cluster the local maximum points to extract the center points of the corridor.

[0079] Finally, perform skeleton line smoothing: Use morphological operations or curve fitting algorithms to smooth the skeleton line and remove unnecessary jaggedness, and key point screening: According to actual requirements, screen the extracted inflection points and center points to remove redundant points.

[0080] S103. Determine the annotation information of the target area based on the matching result of the first topological map and the second topological map.

[0081] In one embodiment, the determining the annotation information of the target area based on the matching result of the first topological map and the second topological map includes:

[0082] Analyze the node attributes and edge attributes of the first topological map and the second topological map based on the graph isomorphism algorithm to obtain the matching result;

[0083] Map the matching result as the annotation information of the target area to the robot map.

[0084] It can be understood that the identification map usually includes key points such as rooms, doors, elevators, and safety passages. These points are connected by paths to form a clear topological structure. The graph isomorphism matching algorithm finds the corresponding relationship between the two by analyzing the topological structures of the identification map and the 2D map built by the robot. For example, the algorithm can match the "Room 101" in the fire evacuation map with a specific location in the robot map, thereby realizing automatic annotation of points.

[0085] Specifically, the present application uses a graph isomorphism matching algorithm (such as the VF2 algorithm) to match the first topological map M1 and the second topological map M2. The algorithm finds the corresponding relationship between the two by analyzing the node attributes (such as room numbers, door types, etc.) and edge attributes (such as path lengths, directions, etc.) of the two graphs. During the matching process, the algorithm can handle local errors and incomplete information to ensure the robustness of the matching result.

[0086] Exemplarily, the graph isomorphism matching algorithm is mainly used to match the topological structure in the identification map with the map built by the robot. The advantage of the algorithm lies in its strong structure analysis ability, which can effectively solve the map matching problem in complex environments and provide high-precision map point annotation for the robot.

[0087] According to the result of the graph isomorphism matching in the present application, the house number and elevator position information in the identification map are mapped onto the robot map. For example, the "Room 101" in the identification map is marked at the corresponding position on the robot map, and at the same time, key points such as elevators and safety passages are marked. Finally, a robot map with semantic information is generated to provide support for subsequent navigation, path planning, and other tasks.

[0088] Furthermore, the above graph isomorphism matching algorithm can significantly improve the matching accuracy by introducing node attributes (such as room numbers, door types, etc.) and edge attributes (such as path lengths, directions, etc.). In addition, the algorithm also supports local matching and dynamic adjustment, and can handle errors or incomplete information that may occur during the map construction process.

[0089] Based on the above solution, by integrating object detection, OCR, and graph isomorphism algorithms, high-precision automatic annotation without manual intervention is achieved. Compared with traditional manual annotation methods, the annotation efficiency is greatly improved, and human errors are avoided; compared with pure automatic annotation technologies, the point recognition accuracy is improved by the normativity and topological structure constraints of the fire evacuation map. This method is especially suitable for complex scenarios (such as large hotels, shopping malls), can quickly complete the point annotation of an environment of tens of thousands of square meters, and provide reliable semantic information support for robot navigation and path planning.

[0090] In one embodiment, determining the annotation information of the target area based on the matching result of the first topological graph and the second topological graph includes:

[0091] During the matching process of the first topological graph and the second topological graph, nodes with unique attributes are preferentially matched;

[0092] After the matching of the nodes with unique attributes is completed, nodes with similar attributes are matched based on adjacent nodes.

[0093] Specifically, during the matching process, nodes with unique attributes (such as room numbers, elevator numbers, etc.) are preferentially matched. For nodes with similar attributes, context information (such as the attributes of adjacent nodes) can be combined for matching.

[0094] Based on the above solution, through attribute matching, the search space for matching can be significantly reduced, and the matching efficiency can be improved. By the attribute matching priority mechanism (such as room numbers with high uniqueness are preferentially matched), the search space is reduced; the combination of local matching (such as subgraph matching) and global optimization (such as minimum cost matching) can handle local deformations (such as wall offsets) between the fire evacuation map and the robot map. Experimental data shows that in a scenario with a 10% map deviation, the matching accuracy still reaches 92%, which is significantly better than traditional shape-based matching methods.

[0095] In one embodiment, the above method further includes:

[0096] Performing a rough match on the first topological graph and the second topological graph first, and

[0097] Performing a local match on the first topological graph and the second topological graph first to determine a subgraph with a relatively high similarity.

[0098] Specifically, for the multi-scale analysis of the topological map, rough matching is first performed at a large scale, and then gradually refined to a small scale.

[0099] Based on the above solution, multi-scale matching can improve the robustness of the algorithm and avoid matching failures caused by local errors.

[0100] It can be understood that local matching is first performed to find subgraphs with high similarity in the two graphs. On the basis of local matching, it is gradually extended to global matching.

[0101] Based on the above solution, local matching can reduce the computational complexity, while global optimization can ensure the accuracy of the overall matching.

[0102] In one embodiment, the above method further includes:

[0103] Overlay the first topological map, the second topological map, and the robot map to determine abnormal nodes;

[0104] Based on the manual intervention interface, correct the abnormal nodes.

[0105] Based on the above solution, visualize the matching result and overlay it on the fire evacuation map and the robot map for manual inspection. For nodes that cannot be automatically matched, a manual intervention interface is provided for operators to perform manual matching. The above visual verification can intuitively detect matching errors and facilitate timely adjustment of algorithm parameters. Manual intervention can solve complex situations that the algorithm cannot handle.

[0106] The visual verification module of the above solution supports cross-modal comparison between 3D point cloud data and 2D maps, and provides a semi-automatic correction function that allows users to drag and drop annotation points for fine-tuning. By overlaying the robot point cloud data and the rendering result of the identification map, the annotation deviation (such as wall offset) is intuitively displayed. The semi-automatic correction function improves the efficiency of manual intervention by 3 times, and only needs to click to correct the point position deviation within 5m. This design significantly reduces the annotation difficulty in complex scenarios (such as glass curtain walls, reflective surfaces) and ensures the robustness of the system in extreme environments.

[0107] With the above technical solution, for the problem of low accuracy and excessive dependence on manual labor in the automatic annotation of mobile robot map points, the automatic annotation method of mobile robot map points provided by the present invention constructs a first topological map by identifying first key node information based on the identification map of the target area; collects environmental data to identify second key node information to construct a second topological map; and determines the annotation information of the target area based on the matching result of the first topological map and the second topological map. In the above solution, high-precision automatic annotation without manual intervention is achieved through algorithms. Compared with the traditional manual annotation method, the annotation efficiency is greatly improved, and human errors are avoided; compared with the pure automatic annotation technology, the point recognition accuracy is greatly improved through the standardization of the identification map of the target area and the topological structure constraint. This method is especially suitable for complex scenarios (such as large hotels and shopping malls), can quickly complete the point annotation of an environment of tens of thousands of square meters, and provides reliable semantic information support for robot navigation and path planning.

[0108] This application proposes a high-precision and high-efficiency automatic annotation method for robot map points by integrating object detection algorithms, OCR algorithms, and graph isomorphism matching algorithms. This method not only solves the problems of high cost and low efficiency of existing manual annotation, but also significantly improves the accuracy and practicality of automatic annotation, providing strong technical support for the autonomous map construction and navigation of robots in complex environments, and is applicable to the following scenarios:

[0109] Hotel scenario: Utilize the hotel fire evacuation map to achieve automatic annotation of points such as rooms, elevators, and safety channels.

[0110] Shopping mall scenario: Automatically annotate key points such as stores, entrances and exits, and stairs through the floor plan of the shopping mall.

[0111] Office building scenario: Achieve rapid annotation of points such as offices, meeting rooms, and fire protection facilities.

[0112] Further, as an implementation of the above Figure 1 shown method, the embodiment of the present invention also provides an automatic annotation device for mobile robot map points for implementing the above Figure 1 shown method. The device embodiment corresponds to the foregoing method embodiment. For the convenience of reading, the details of the foregoing method embodiment will not be repeated one by one in this device embodiment, but it should be clear that the device in this embodiment can correspondingly implement all the contents of the foregoing method embodiment. As Figure 2 shown, the device includes: a first construction unit 21, a second construction unit 22, and a determination unit 23, where

[0113] The first construction unit 21 is used to identify first key node information based on the identification map of the target area to construct a first topological map;

[0114] A second construction unit 22, configured to collect environmental data to identify second key node information and construct a second topology map;

[0115] A determination unit 23, configured to determine annotation information of the target area based on a matching result between the first topology map and the second topology map.

[0116] The processor includes a kernel, and the kernel retrieves corresponding program units from the memory. One or more kernels can be set, and by adjusting kernel parameters, an automatic annotation method for map points of a mobile robot can be implemented, which can solve the problems of low accuracy of automatic annotation of map points of a mobile robot and over - reliance on manual work.

[0117] An embodiment of the present invention provides a computer - readable storage medium. The computer - readable storage medium includes a stored program, and when the program is executed by a processor, the automatic annotation method for map points of the mobile robot is implemented.

[0118] An embodiment of the present invention provides a processor. The processor is used to run a program, and when the program runs, the automatic annotation method for map points of the mobile robot is executed.

[0119] An embodiment of the present invention provides an electronic device. The electronic device includes at least one processor and at least one memory connected to the processor; wherein, the processor is used to call program instructions in the memory to execute the automatic annotation method for map points of the mobile robot as described above

[0120] An embodiment of the present invention provides an electronic device 30, as Figure 3 shown. The electronic device includes at least one processor 301, at least one memory 302 connected to the processor, and a bus 303; wherein, the processor 301 and the memory 302 communicate with each other through the bus 303; the processor 301 is used to call program instructions in the memory to execute the automatic annotation method for map points of the mobile robot as described above.

[0121] The intelligent electronic device herein can be a PC, a PAD, a mobile phone, etc.

[0122] The present application also provides a computer program product, which is suitable for executing a program initialized with the steps of the automatic annotation method for map points of the mobile robot when executed on a process management electronic device.

[0123] It should be noted that in the above - mentioned embodiments, the descriptions of each embodiment have their own emphases. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0124] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0125] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0126] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0128] The embodiments of the present application also provide a computer program product, which includes computer software instructions. When the computer software instructions run on a processing device, the processing device is caused to execute the process of controlling the memory in the corresponding Figure 1 embodiment.

[0129] A computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or a wireless manner (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be stored by a computer or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium may be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)).

[0130] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein again.

[0131] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other may be indirect couplings or communication connections through some interfaces, devices, or units, and may be in electrical, mechanical, or other forms.

[0132] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0133] In addition, in each embodiment of the present application, each functional unit may be integrated into one processing unit, may exist physically alone for each unit, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0134] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, may be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0135] The above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A method for automatically marking map points of a mobile robot, characterized in that: include: Identify first key node information based on the identification graph of the target area to construct a first topological graph; Collecting environmental data to identify second key node information to construct a second topological graph; The labeling information of the target area is determined based on the matching result of the first topological map and the second topological map.

2. The method according to claim 1, characterized in that The first key node information includes theoretical point legend information, theoretical room information, and theoretical corridor information. The identification of the first key node information based on the identification map of the target area to construct a first topological map includes: Identify the identification map of the target area by a target detection algorithm to determine the point legend information; Recognize the identification image of the target area by using an OCR algorithm to determine the room information; Identify the identification map of the target area by using a skeleton line extraction algorithm to determine the corridor information; A first topological map is constructed based on the point legend information, the room information and the corridor information.

3. The method according to claim 1, characterized in that: The second key node information includes actual point legend information, actual room information, and actual corridor information. The collected environment data identifies the second key node information to construct a second topological map, including: Collect environmental data based on sensors; Identify the environmental data to obtain the actual point legend information and the actual room information; Projecting the actual point legend information and the actual room information to form a 2D map; Extracting skeleton lines from the 2D map to obtain the actual corridor information; The second topological map is constructed based on the actual point legend information, the actual room information and the actual corridor information.

4. The method according to claim 1, characterized in that The determining the labeling information of the target area based on the matching result of the first topological map and the second topological map includes: Analyzing the node attributes and edge attributes of the first topological graph and the second topological graph based on a graph isomorphism algorithm to obtain the matching result; The matching result is used as the annotation information of the target area and mapped into the robot map.

5. The method according to claim 1, characterized in that: The determining the labeling information of the target area based on the matching result of the first topological map and the second topological map includes: In the matching process of the first topological graph and the second topological graph, the unique attribute nodes are matched preferentially; After the unique attribute nodes are matched, similar attribute nodes are matched based on adjacent nodes.

6. The method according to claim 5, characterized in that Also includes: Prioritizing coarse matching of the first topology map and the second topology map, and The first topological graph and the second topological graph are preferentially locally matched to determine a subgraph with a higher similarity.

7. The method according to claim 4, characterized in that Also includes: Overlaying the first topological map, the second topological map, and the robot map to determine abnormal nodes; The abnormal node is corrected based on a manual intervention interface.

8. A mobile robot map point automatic marking device, characterized in that: Also includes: A first construction unit, configured to identify first key node information based on an identification map of a target area to construct a first topological map; A second construction unit is used to collect environmental data to identify second key node information to construct a second topological graph; A determining unit is used to determine the labeling information of the target area based on the matching result of the first topological map and the second topological map.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the steps of the method for automatic labeling of map points of a mobile robot as claimed in any one of claims 1 to 7 are implemented.

10. An electronic device, characterized in that: The electronic device includes at least one processor and at least one memory connected to the processor; wherein the processor is used to call program instructions in the memory to execute the steps of the mobile robot map point automatic labeling method as described in any one of claims 1 to 7.

Images