Map construction method, map construction device, delivery robot, and storage medium

Abstract

The application discloses a map construction method, a map construction device, a delivery robot and a computer readable storage medium. The method comprises the following steps: constructing an initial map of an indoor environment according to a plurality of detection data of the indoor environment, wherein the plurality of detection data is obtained by detecting the indoor environment through at least two types of detectors carried by the delivery robot; determining a risk position according to the plurality of detection data; and optimizing the initial map according to the risk position to obtain a target map of the indoor environment. Through the application, a more optimized map can be obtained for the indoor environment, and the delivery robot can be relatively more accurately positioned.

Description

Technical Field

[0001] This application belongs to the field of robotics technology, and in particular relates to a map building method, a map building device, a delivery robot, and a computer-readable storage medium. Background Technology

[0002] When a robot is deployed to a new environment, it needs to first map that environment before it can navigate autonomously based on the map to perform various tasks. However, for delivery robots working in special indoor environments (such as restaurants or hotels), the environmental characteristics of these environments limit their ability to accurately locate themselves using the map. Summary of the Invention

[0003] This application provides a map building method, a map building device, a delivery robot, and a computer-readable storage medium, which can obtain more optimized maps for indoor environments and help the delivery robot achieve relatively more accurate positioning.

[0004] Firstly, this application provides a map building method applied to a delivery robot, the map building method including:

[0005] An initial map of the indoor environment is constructed based on multi-frame detection data of the indoor environment. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors that are already mounted on the delivery robot.

[0006] Based on multi-frame detection data, the location of the risk is determined;

[0007] The initial map is optimized based on the risk location to obtain the target map of the indoor environment.

[0008] Secondly, this application provides a map building device applied to a delivery robot, the map building device comprising:

[0009] The construction module is used to build an initial map of the indoor environment based on multi-frame detection data of the indoor environment. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors that are already mounted on the delivery robot.

[0010] The first determination module is used to determine the risk location based on multi-frame detection data;

[0011] The optimization module is used to optimize the initial map based on the risk locations to obtain the target map of the indoor environment.

[0012] Thirdly, this application provides a delivery robot, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method described in the first aspect.

[0013] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described in the first aspect above.

[0014] Fifthly, this application provides a computer program product comprising a computer program that, when executed by one or more processors, implements the steps of the method described in the first aspect.

[0015] The advantages of this application compared to existing technologies are as follows: First, the delivery robot constructs an initial map of the indoor environment based on multi-frame detection data. In addition, it identifies risk locations based on the multi-frame detection data. Finally, it optimizes the initial map based on the risk locations to obtain a target map of the indoor environment. In this process, on the one hand, the multi-frame detection data is obtained from at least two types of detectors already mounted on the delivery robot, providing strong data support and making the initial map more accurate. On the other hand, the initial map is not directly applied; instead, risk locations in the indoor environment are identified based on the detection data, and the initial map is optimized before the optimized target map is applied. This allows the delivery robot to more accurately locate itself based on the target map during subsequent task execution.

[0016] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram illustrating the implementation process of the map construction method provided in this application embodiment;

[0019] Figure 2 This is a schematic diagram of the structure of the delivery robot provided in the embodiments of this application;

[0020] Figure 3 This is a structural block diagram of the map building apparatus provided in the embodiments of this application;

[0021] Figure 4 This is a structural block diagram of the delivery robot provided in the embodiments of this application. Detailed Implementation

[0022] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0023] The following describes a map construction method provided by an embodiment of this application. This map construction method can be applied to delivery robots; of course, it can also be applied to other devices that have established a communication connection with the delivery robot, such as servers, etc., and this embodiment of the application does not limit this application. For ease of description, the following explanation and illustration will use the application of this map construction method to a delivery robot as an example. Please refer to... Figure 1 The map construction method in this application embodiment includes:

[0024] Step 101: Construct an initial map of the indoor environment based on multi-frame detection data of the indoor environment.

[0025] When a delivery robot is moved into a new indoor environment, or when it is necessary to remap the indoor environment in which the delivery robot is located, it is advisable to first determine whether the indoor environment is a specific scenario. For example, this specific scenario can be set up by the deployment personnel, including but not limited to hotel and restaurant scenarios, which are prone to navigation problems. Once it is determined that the indoor environment is an indoor environment of any specific scenario, it can be considered that there is a need to optimize the subsequent mapping process.

[0026] As part of the mapping process, the delivery robot can autonomously control its movement within the indoor environment, or it can be manually pushed or remotely controlled by personnel to scan the environment. Specifically, the delivery robot is equipped with at least two types of detectors, which provide strong data support, resulting in more accurate detection data.

[0027] As an example, delivery robots in indoor environments can be equipped with single-line lidar based on the time-of-flight (TOF) principle. This type of single-line lidar emits a modulated laser signal via a laser transmitter. This modulated laser signal is reflected by the object being measured and received by a laser detector. The distance to the object can be calculated by measuring the phase difference between the emitted and received lasers. However, considering that single-line lidar typically only scans a cross-section of space and has limited detection range and accuracy, the delivery robot can also be equipped with a depth camera based on structured light or binocular structured light principles. The detection range of this depth camera is mainly determined by the detection range and the field of view (FOV). That is, the delivery robot can be equipped with at least two types of detectors: single-line lidar and a depth camera. Of course, delivery robots can also be equipped with other types of detectors, such as multi-line lidar and binocular cameras, etc., which are not limited here.

[0028] Furthermore, in order to enable the delivery robot to better identify obstacles, steps, and cliffs in front of it, the deployers may consider setting up 1 to 2 depth cameras on the bottom of the front of the delivery robot, tilted slightly upwards; and placing 1 depth camera on the top of the front, tilted slightly downwards.

[0029] Based on the above deployment, each time the delivery robot performs a detection operation at a location, it can obtain data collected by different types of detectors at that location. Taking a delivery robot equipped with a single-line LiDAR and a depth camera as an example, the robot can collect both LiDAR data and depth data. By fusing the data collected by different types of detectors at the same location, the detection data for that location can be obtained. This detection data can be represented in the form of a point cloud. Through the movement of the delivery robot in the indoor environment, it can obtain detection data at multiple different locations (i.e., multi-frame detection data), and thereby construct an initial map of the current indoor environment.

[0030] Step 102: Determine the risk location based on multi-frame detection data.

[0031] After acquiring multi-frame detection data from the indoor environment, the delivery robot can analyze this data. The analysis results can identify locations where potential risks exist, including but not limited to navigational risks and fall risks. Furthermore, the analysis results can indicate the risk level of these locations. For ease of description, these potentially risky locations can be referred to as risky locations.

[0032] Among these risks, getting lost is more likely to occur in deep corridors and lobbies of hotel settings because single-line LiDAR can only scan a cross-section of space and has a relatively limited detection range. Additionally, getting lost is also more likely to occur in restaurant settings due to the high density of tables. Falling risks are more likely to occur in the upper and lower floors of hotel and restaurant settings because depth cameras have weaker recognition capabilities for dark objects, but these areas are often covered with dark gray carpets or have concrete floors.

[0033] Of course, deployment personnel can also inform the delivery robot of potential risks within the indoor environment through interactive methods indicated by the robot, based on their observations of the indoor environment; this will not be elaborated upon here. In other words, this application embodiment does not limit the method for determining risk locations.

[0034] Step 103: Optimize the initial map based on the risk location to obtain the target map of the indoor environment.

[0035] Once at least one risky location is identified, the delivery robot can mark each risky location on the initial map to record relevant information. As an example only, the delivery robot could use color-coding or icon-based marking methods; no limitation is made here.

[0036] Furthermore, the delivery robot can also mark each risk location on the initial map based on the risk level. As an example, if the delivery robot uses color-coding and divides the risk into low, medium, and high risk levels, the color for high-risk locations can be set to red, medium-risk locations to orange, and low-risk locations to yellow. Of course, other colors can be used to mark risk locations at each risk level, as long as they serve a distinguishing purpose; this is not limited here.

[0037] In some embodiments, the delivery robot can identify potentially disorienting locations by:

[0038] A1. For any two frames of detection data obtained at different detection locations, calculate the similarity between the two frames of detection data.

[0039] The delivery robot can traverse the obtained detection data. For the currently traversed detection data, the delivery robot can compare the detection data with the detection data obtained at each other detection location, thereby obtaining the similarity between two frames of detection data obtained at different detection locations. In this embodiment, no limitation is made on the method of calculating similarity.

[0040] A2. If the similarity is greater than the preset similarity threshold, then the detection locations corresponding to the two frames of detection data will be determined as risk locations.

[0041] Deployers can pre-set a similarity threshold in the delivery robot. For example, this threshold could be 0.8 or other values; no limitation is made here. After calculating the similarity between two frames of detection data, the delivery robot can compare this similarity to the similarity threshold. If the comparison finds that the similarity is greater than the threshold, the two frames of detection data are considered quite close and may confuse the delivery robot. In this case, the detection locations corresponding to these two frames of detection data can be designated as risk locations.

[0042] Furthermore, after identifying the risk location through the above methods, the delivery robot can also determine the risk level based on the calculated similarity. Specifically, the range of values ​​greater than the similarity threshold can be divided into multiple non-overlapping similarity intervals, with different similarity intervals corresponding to different risk levels. When the similarity between two frames of detection data is greater than the similarity threshold, the risk level corresponding to the target similarity interval can be determined as the risk level of the detection location corresponding to these two frames of detection data, where the target similarity interval refers to the similarity interval into which the similarity falls.

[0043] In some embodiments, the delivery robot can identify potentially risky locations where there is a risk of falling by means of:

[0044] B1. For each frame of probe data, obtain the hole rate of the probe data.

[0045] When the delivery robot's detector includes a depth camera, the porosity corresponding to the depth data can also be obtained when collecting depth data through the depth camera. Therefore, for the detection data obtained at any detection location, the porosity of the detection data is equal to the porosity of the depth data at that detection location.

[0046] B2. If the void ratio is greater than the preset void ratio threshold, the detection location corresponding to the detection data will be determined as a risk location.

[0047] Generally, when there is a risk of fall, the void ratio of depth data collected by depth cameras will be relatively high. Based on this, deployment personnel can pre-set a void ratio threshold in the delivery robot. Once the delivery robot finds that the void ratio of the detected data exceeds the void ratio threshold, the detection location corresponding to that data can be identified as a risk location.

[0048] Furthermore, after identifying the risk location through the above methods, the delivery robot can also determine the risk level based on the porosity. Specifically, the range of values ​​greater than the porosity threshold can be divided into multiple non-overlapping porosity intervals, with different porosity intervals corresponding to different risk levels. When the porosity of a frame of detection data is greater than the porosity threshold, the risk level corresponding to the target porosity interval can be determined as the risk level of the detection location corresponding to that frame of detection data. Here, the target porosity interval refers to the porosity interval into which the porosity falls.

[0049] In some embodiments, delivery robots can also identify potentially risky locations where there is a risk of falling by means of:

[0050] C1. For each frame of detection data, analyze whether the scene corresponding to the detection data contains a risky scene.

[0051] When the delivery robot is equipped with a color camera as its detector, it can acquire color images of the environment at each detection location. In this case, for any given frame of detection data, the delivery robot can obtain the corresponding color image of the environment and perform image processing on it, including target detection and recognition. Through image processing, the delivery robot can identify the targets contained in the color image. Based on the targets in the color image, the delivery robot can determine whether the scene corresponding to the detection data contains a risky scenario, which refers to a scenario where the height difference is greater than a preset height difference.

[0052] C2. If the scenario includes a risky scenario, then the detection location corresponding to the detection data will be determined as the risky location.

[0053] If the scenario corresponding to the detection data includes a risky scenario, then a fall risk can be considered to exist. Based on this, the delivery robot can determine the detection location corresponding to the detection data as a risky location.

[0054] Furthermore, after identifying the risk location through the above methods, the delivery robot can also determine the risk level based on the area (or number) of the risk scenario. Specifically, it can be divided into multiple non-overlapping area intervals (or number intervals) for risk scenarios, with different area intervals (or number intervals) corresponding to different risk levels. When the scenario corresponding to a frame of detection data contains a risk scenario, the risk level corresponding to the target area interval (or target number interval) can be determined as the risk level of the detection location corresponding to that frame of detection data. Here, the target area interval refers to the area interval into which the area of ​​the risk scenario contained in the scenario corresponding to the detection data falls, and the target number interval refers to the number interval into which the number of risk scenarios contained in the scenario corresponding to the detection data falls.

[0055] It is understandable that delivery robots can choose to execute one or more of the three risk location determination methods mentioned above, depending on their own needs; no restrictions are imposed here.

[0056] In some embodiments, to further improve the positioning accuracy of the delivery robot in indoor environments, the delivery robot can output a reminder message after identifying a risky location. This reminder message is used to remind the user (i.e., the deployment personnel) to set up a positioning assistance tool at the risky location. As an example only, this positioning assistance tool includes, but is not limited to, Near Field Communication (NFC) tags and QR codes used for positioning. Furthermore, the positioning assistance tool (NFC tag or QR code) corresponding to each risky location is unique throughout the indoor environment. For ease of description, the NFC tag can be referred to as a positioning NFC tag, and the QR code as a positioning QR code. The deployment personnel can choose an appropriate positioning assistance tool based on cost and accuracy requirements; no limitations are imposed here.

[0057] To enhance the usability of location assistance tools, each tool can contain a variety of information. For example, a location QR code can consist of multiple smaller QR codes, each representing different dimensions of location information, such as orientation and floor information; this is not limited here. It is understandable that delivery robots can fully utilize these location assistance tools in subsequent processes to further improve their positioning accuracy.

[0058] In some embodiments, when the deployment personnel select the location QR code as a location assistance tool, to prevent wear and tear on the QR code, they can specifically place it on the ceiling at a high-risk location. In this case, the delivery robot can be equipped with a top-view camera, specifically facing upwards to capture images. Please refer to [link to relevant documentation]. Figure 2 , Figure 2 A schematic diagram of the delivery robot is provided. Figure 2 The solid cubes in the image represent the top-view camera, single-line LiDAR, depth camera 1, and depth camera 2 mounted on the delivery robot. Figure 2 The text has been annotated, so it will not be repeated here. Also, Figure 2 In the diagram, angle 1 is the field of view of depth camera 1, angle 2 is the field of view of depth camera 2, and angle 3 is the field of view of the top-view camera. Based on the top-view camera already mounted on the delivery robot, after step 103, the map construction method may further include:

[0059] D1. During the execution of mobile navigation tasks, top-view cameras are used to collect real-time images of the indoor environment from the top.

[0060] To prevent inaccurate positioning of the delivery robot after it moves to a risky location, the delivery robot can trigger the operation of the top-view camera, which can then collect real-time images of the indoor environment from the top.

[0061] In one application scenario, the top-view camera can be triggered to run when the delivery robot starts performing a mobile navigation task, and can be triggered to pause operation when the delivery robot stops performing a mobile navigation task; that is, as long as the delivery robot is performing a mobile navigation task, its top-view camera remains running.

[0062] In another application scenario, the operation of the top-view camera will consume some system resources of the delivery robot; moreover, the delivery robot is not usually in a risky position all the time. Based on this, the top-view camera can be triggered to operate when the original positioning result indicates that the delivery robot has entered a preset range of any risky position, and triggered to stop operating when the original positioning result indicates that the delivery robot has left the preset range of the risky position; that is, the top-view camera is only in operation when the delivery robot is close to the risky position.

[0063] The raw positioning result refers to the positioning result obtained based on the detectors (such as single-line LiDAR and depth cameras) carried by the delivery robot. Under normal circumstances, without positioning risks, this raw positioning result is already a relatively accurate one.

[0064] In some examples, the preset range can be adaptively set according to the risk level of the corresponding risk location. Specifically, the higher the risk level of the risk location, the larger the preset range of that risk location; conversely, the lower the risk level of the risk location, the smaller the preset range of that risk location.

[0065] D2. When a positioning QR code exists in the top image, determine the auxiliary positioning result based on the positioning QR code.

[0066] After the top-view camera captures an image of the top, the delivery robot checks if a location QR code is present in that image. If a location QR code is detected, it's assumed the robot has moved into a high-risk area, and the initial location result may be inaccurate. In this case, the robot can parse the location QR code and determine the corresponding location result based on the parsed information. For clarity, this location result can be referred to as the auxiliary location result.

[0067] D3. Combining the original positioning results and the auxiliary positioning results, the final positioning result is obtained.

[0068] Given the assisted positioning result, the delivery robot can fuse this result with the original positioning result obtained at the same time to obtain the final positioning result. This fusion process can be considered as optimizing the original positioning result using the assisted positioning result, which can improve the accuracy of the final positioning result compared to the original positioning result.

[0069] In some embodiments, to enable the delivery robot to perform well in low-light environments, the top-view camera mounted on the delivery robot may specifically be an infrared camera. It is understood that, compared to ordinary color cameras, infrared cameras have a longer detection range and are almost unaffected by light, helping the delivery robot to recognize positioning QR codes that are farther away and / or smaller in size.

[0070] In addition, the delivery robot can also be equipped with an infrared fill light near the top-view camera to provide supplementary lighting for the top-view camera in low-light environments.

[0071] In some examples, the delivery robot may turn on the fill light when entering a specific area of ​​the indoor environment (e.g., a deep corridor) while the top-view camera is running; and turn off the fill light when leaving a specific area of ​​the indoor environment, and / or when the top-view camera is paused.

[0072] In other examples, the delivery robot may turn on its fill light when the current time is within a preset nighttime period and the top-view camera is running; and turn off its fill light when the current time is not within a preset nighttime period and / or when the top-view camera is paused.

[0073] In some other examples, the delivery robot may also be equipped with a brightness sensor. When the ambient brightness value detected by the brightness sensor is less than a preset brightness threshold and the top-view camera is running, the delivery robot can turn on the supplementary light; when the ambient brightness value detected by the brightness sensor is greater than or equal to the brightness threshold, and / or when the top-view camera is not running, the delivery robot can turn off the supplementary light.

[0074] It is understood that the conditions for starting / stopping the supplementary light are not limited in the embodiments of this application.

[0075] As can be seen from the above, in this embodiment, on the one hand, the multi-frame detection data is obtained by the delivery robot itself using at least two types of detectors, which form a strong data support, making the initial map relatively more accurate; on the other hand, the initial map is not directly applied, but the risk locations in the indoor environment are determined based on the detection data and the initial map is optimized before the optimized target map is applied, so that the delivery robot can more accurately locate itself based on the target map when performing tasks in the subsequent application of the target map.

[0076] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0077] Corresponding to the map building method provided above, this application also provides a map building apparatus, which can be integrated into a delivery robot or other devices that establish a communication connection with the delivery robot. Please refer to... Figure 3 The map building device 3 in this embodiment includes:

[0078] The construction module 301 is used to construct an initial map of the indoor environment based on multi-frame detection data of the indoor environment. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors that are already mounted on the delivery robot.

[0079] The first determining module 302 is used to determine the risk location based on multi-frame detection data;

[0080] The optimization module 303 is used to optimize the initial map based on the risk location to obtain the target map of the indoor environment.

[0081] In some embodiments, the first determining module 302 includes:

[0082] The calculation unit is used to calculate the similarity between any two frames of detection data obtained at different detection locations.

[0083] The first determining unit is used to determine the detection locations corresponding to the two frames of detection data as risk locations if the similarity is greater than a preset similarity threshold.

[0084] In some embodiments, the first determining module 302 includes:

[0085] The acquisition unit is used to acquire the hole rate of the probe data for each frame of probe data;

[0086] The second determining unit is used to determine the detection location corresponding to the detection data as a risk location if the void ratio is greater than a preset void ratio threshold.

[0087] In some embodiments, the first determining module 302 includes:

[0088] The analysis unit is used to analyze whether the scene corresponding to each frame of detection data contains a risky scene. The risky scene is a scene where the height difference is greater than a preset height difference.

[0089] The third determining unit is used to determine the detection location corresponding to the detection data as a risk location if the scenario includes a risk scenario.

[0090] In some embodiments, the map building device 3 further includes:

[0091] The reminder module is used to output reminder messages based on the risk location. The reminder messages are used to remind users to set up a location QR code at the risk location.

[0092] In some embodiments, the location QR code is placed on the ceiling at a high-risk location; the delivery robot is also equipped with a top-view camera that faces upwards to capture images; the map building device 3 further includes:

[0093] The acquisition module is used to acquire real-time images of the indoor environment from the top using a top-view camera during the execution of mobile navigation tasks;

[0094] The second determining module is used to determine the auxiliary positioning result based on the positioning QR code when a positioning QR code exists in the top image of the indoor environment.

[0095] The module combines the original positioning results and the auxiliary positioning results to obtain the final positioning result, wherein the original positioning results are obtained based on at least two types of detectors.

[0096] In some embodiments, during the execution of a mobile navigation task, the top-view camera is triggered to run when the original positioning results indicate that the delivery robot has entered a preset range of any risky location, and is paused to run when the original positioning results indicate that the delivery robot has left the preset range.

[0097] As can be seen from the above, in this embodiment, on the one hand, the multi-frame detection data is obtained by the delivery robot itself using at least two types of detectors, which form a strong data support, making the initial map relatively more accurate; on the other hand, the initial map is not directly applied, but the risk locations in the indoor environment are determined based on the detection data and the initial map is optimized before the optimized target map is applied, so that the delivery robot can more accurately locate itself based on the target map when performing tasks in the subsequent application of the target map.

[0098] Corresponding to the map construction method provided above, this application also provides a delivery robot. Please refer to... Figure 4 The delivery robot 4 in this embodiment includes: a memory 401, and one or more processors 402. Figure 4 (Only one is shown) and a computer program stored in memory 401 and executable on the processor. Memory 401 stores software programs and modules. The processor 402 executes various functional applications and data processing by running the software programs and units stored in memory 401 to obtain resources corresponding to the aforementioned preset events. Specifically, the processor 402 performs the following steps by running the aforementioned computer program stored in memory 401:

[0099] An initial map of the indoor environment is constructed based on multi-frame detection data of the indoor environment. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors already mounted on the delivery robot 4.

[0100] Based on multi-frame detection data, the location of the risk is determined;

[0101] The initial map is optimized based on the risk location to obtain the target map of the indoor environment.

[0102] Assuming the above is the first possible implementation, then in the second possible implementation provided based on the first possible implementation, determining the risk location based on multi-frame detection data includes:

[0103] For any two frames of detection data obtained at different detection locations, calculate the similarity between the two frames of detection data;

[0104] If the similarity is greater than the preset similarity threshold, then the detection locations corresponding to the two frames of detection data will be determined as risk locations.

[0105] In a third possible implementation provided based on the first possible implementation described above, determining the risk location based on multi-frame detection data includes:

[0106] For each frame of probe data, obtain the hole rate of the probe data;

[0107] If the void ratio is greater than the preset void ratio threshold, the detection location corresponding to the detection data will be identified as a risk location.

[0108] In the fourth possible implementation provided based on the first possible implementation described above, determining the risk location based on multi-frame detection data includes:

[0109] For each frame of detection data, analyze whether the scene corresponding to the detection data contains a risky scene. A risky scene is a scene where the height difference is greater than the preset height difference.

[0110] If the scenario includes a risky scenario, then the detection location corresponding to the detection data will be determined as the risky location.

[0111] In a fifth possible implementation provided based on the first, second, third, or fourth possible implementations described above, after determining the risk location based on multi-frame detection data, the processor 402 further performs the following steps when running the computer program stored in the memory 401:

[0112] Based on the risk location, an alert message is output, which reminds the user to set a location QR code at the risk location.

[0113] In the sixth possible implementation provided based on the fifth possible implementation described above, the location QR code is placed on the ceiling at the risk location; the delivery robot 4 is also equipped with a top-view camera, which faces upwards to acquire images; after obtaining the target map of the indoor environment, the map construction method further includes:

[0114] During the mobile navigation mission, top-view cameras are used to capture real-time images of the indoor environment from the top.

[0115] When a location QR code is present in the top image, the auxiliary location result is determined based on the location QR code.

[0116] The final positioning result is obtained by combining the original positioning result and the auxiliary positioning result, wherein the original positioning result is obtained based on at least two types of detectors.

[0117] In the seventh possible implementation based on the sixth possible implementation described above, during the execution of the mobile navigation task, the top-view camera is triggered to run when the original positioning result indicates that the delivery robot 4 has entered a preset range of any risky location, and is suspended to run when the original positioning result indicates that the delivery robot 4 has left the preset range.

[0118] It should be understood that, in the embodiments of this application, the processor 402 may be a central processing unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0119] Memory 401 may include read-only memory and random access memory, and provides instructions and data to processor 402. Some or all of memory 401 may also include non-volatile random access memory. For example, memory 401 may also store device type information.

[0120] As can be seen from the above, in this embodiment, on the one hand, the multi-frame detection data is obtained by the delivery robot itself using at least two types of detectors, which form a strong data support, making the initial map relatively more accurate; on the other hand, the initial map is not directly applied, but the risk locations in the indoor environment are determined based on the detection data and the initial map is optimized before the optimized target map is applied, so that the delivery robot can more accurately locate itself based on the target map when performing tasks in the subsequent application of the target map.

[0121] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the above device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0122] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0123] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of external device software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0124] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules or units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units through some interfaces, and may be electrical, mechanical, or other forms.

[0125] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0126] If the integrated units described above are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing associated hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer-readable storage device, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the contents of the aforementioned computer-readable storage media may be appropriately added to or subtracted from the contents according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media may not include electrical carrier signals and telecommunication signals.

[0127] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A map construction method, characterized in that, The map building method is applied to a delivery robot, and the map building method includes: Based on multi-frame detection data of the indoor environment, an initial map of the indoor environment is constructed. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors already mounted on the delivery robot. The at least two types of detectors include a single-line lidar and a depth camera. The depth camera is respectively set at the bottom and top of the front of the delivery robot, with the depth camera at the bottom tilted upward and the depth camera at the top tilted downward. The multi-frame detection data is point cloud data formed by fusing the data collected by the at least two types of detectors at the same detection location. Determining risk locations based on the multi-frame detection data includes: calculating the similarity between any two frames of detection data obtained at different detection locations; if the similarity is greater than a preset similarity threshold, then determining the detection locations corresponding to the two frames of detection data as the risk locations; dividing the range of values ​​greater than the similarity threshold into multiple non-overlapping similarity intervals, wherein different similarity intervals correspond to different risk levels, and determining the risk level of the risk location based on the target similarity interval into which the similarity falls; The initial map is optimized based on the risk locations to obtain a target map of the indoor environment. The optimization includes: using differentiated color markings based on the risk level of the risk locations. Based on the risk location, an alert message is output. The alert message is used to remind the user to set a positioning QR code on the ceiling at the risk location. The delivery robot is also equipped with a top-view camera, which captures images upwards from the delivery robot. After obtaining the target map, when the delivery robot performs a mobile navigation task, the original positioning result of the delivery robot is obtained based on the single-line lidar and the depth camera; The preset range of the risk location is adaptively set according to the risk level of the risk location, wherein the higher the risk level, the larger the preset range; When the original positioning result indicates that the delivery robot enters a preset range of any of the risk locations, the top-view camera is triggered to capture a top image of the indoor environment, and when the original positioning result indicates that the delivery robot leaves the preset range, the top-view camera is triggered to pause operation. When a positioning QR code exists in the top image, the auxiliary positioning result is determined based on the positioning QR code; The auxiliary positioning result is fused with the original positioning result obtained at the same time to obtain the final positioning result.

2. The map construction method as described in claim 1, characterized in that, The step of determining the risk location based on the multi-frame detection data also includes: For each frame of probe data, the hole rate of the probe data is obtained; If the void ratio is greater than a preset void ratio threshold, the detection location corresponding to the detection data is determined as the risk location.

3. The map construction method as described in claim 1, characterized in that, The step of determining the risk location based on the multi-frame detection data also includes: For each frame of detection data, analyze whether the scene corresponding to the detection data contains a risky scene, wherein the risky scene is a scene where the height difference is greater than a preset height difference; If the scenario includes a risky scenario, then the detection location corresponding to the detection data is determined as the risky location.

4. A map building device, characterized in that, The map building device is used in a delivery robot, and the map building device includes: The module is used to construct an initial map of the indoor environment based on multi-frame detection data of the indoor environment. The multi-frame detection data is obtained by detecting the indoor environment through at least two types of detectors already mounted on the delivery robot. The at least two types of detectors include a single-line lidar and a depth camera. The depth camera is respectively set at the bottom and top of the front of the delivery robot, with the depth camera at the bottom tilted upward and the depth camera at the top tilted downward. The multi-frame detection data is point cloud data formed by fusing the data collected by the at least two types of detectors at the same detection location. The first determining module is used to determine risk locations based on the multi-frame detection data, including: calculating the similarity between any two frames of detection data obtained at different detection locations; if the similarity is greater than a preset similarity threshold, determining the detection locations corresponding to the two frames of detection data as the risk locations; dividing the range of values ​​greater than the similarity threshold into multiple non-overlapping similarity intervals, wherein different similarity intervals correspond to different risk levels, and determining the risk level of the risk location based on the target similarity interval into which the similarity falls; An optimization module is used to optimize the initial map based on the risk location to obtain a target map of the indoor environment. The optimization includes: performing differentiated color labeling based on the color corresponding to the risk level of the risk location. The risk level is determined based on similarity, void ratio, or area / number of risk scenes. The reminder module is used to output a reminder message based on the risk location. The reminder message is used to remind the user to set a positioning QR code on the ceiling at the risk location. The delivery robot is also equipped with a top-view camera, which is directed upwards to capture images of the delivery robot. The map building device is further configured to, after obtaining the target map, when the delivery robot performs a mobile navigation task, obtain the original positioning result of the delivery robot based on the single-line lidar and the depth camera; adaptively set a preset range of the risk location according to the risk level of the risk location, wherein the higher the risk level, the larger the preset range; when the original positioning result indicates that the delivery robot enters the preset range of any of the risk locations, trigger the top-view camera to acquire a top image of the indoor environment, and when the original positioning result indicates that the delivery robot leaves the preset range, trigger the top-view camera to pause operation; when a positioning QR code exists in the top image, determine an auxiliary positioning result based on the positioning QR code; and fuse the auxiliary positioning result with the original positioning result obtained at the same time to obtain the final positioning result.

5. A delivery robot, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 3.

Images