A robot people recognition method, device and electronic equipment

Abstract

This application provides a robot people recognition method, device, and electronic device. The robot people recognition method includes: acquiring target point cloud data within a space collected by a single-line lidar; extracting boundary features from the target point cloud data within the space based on the target point cloud data to determine the corresponding boundary point cloud data; and extracting foreground target features from the internal spatial point cloud data within the space based on the internal spatial point cloud data to determine the number of foreground target features within the space and to determine the target number of people within the space. This application enables mobile robots to simultaneously determine obstacles and the target number of people within the elevator when riding in passenger elevators where humans and robots coexist, thereby improving the service efficiency and accuracy of mobile robots.

Description

Technical Field

[0001] This application relates to the field of lidar technology, and in particular to a method, device and electronic device for robot people recognition. Background Technology

[0002] With social development and technological progress, more and more service industries are starting to use intelligent mobile robots, and the application of mobile robots is becoming more and more widespread. For example, mobile robots can provide services to users on different floors by operating in different public areas.

[0003] However, in traditional service industries, mobile robots often need to use dedicated robotic freight elevators to move between floors to serve users and meet their needs. But as people's living standards improve, there are more and more scenarios where mobile robots are needed. A single freight elevator can no longer meet the needs of mobile robots to provide services to users. Therefore, it is necessary for mobile robots to use passenger elevators where humans and robots coexist. However, when mobile robots use passenger elevators where humans and robots coexist, they can only detect whether there are obstacles in a certain space within the elevator, but they cannot accurately identify the number of people in the elevator. This makes it difficult for mobile robots to move between floors, resulting in low service efficiency and reduced accuracy of mobile robot services. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a robot people recognition method, device and electronic device, which enables the simultaneous determination of the target number of people in obstacles and target space when the mobile robot is taking a passenger elevator where people and robots coexist, thereby improving the service efficiency of the mobile robot and improving the accuracy of the mobile robot service.

[0005] This application provides a robot crowd recognition method, which includes:

[0006] Acquire target point cloud data within the space collected by a single-line lidar;

[0007] Based on the target point cloud data, boundary features are extracted from the target point cloud data in the space to determine the boundary point cloud data corresponding to the space.

[0008] Based on the boundary point cloud data and the target point cloud data, determine the internal space point cloud data corresponding to the space;

[0009] Based on the internal spatial point cloud data, foreground target features are extracted from the internal spatial point cloud data in the space to determine the number of foreground target features in the space.

[0010] Based on the number of foreground target features, the number of target people in the space is determined.

[0011] Furthermore, target point cloud data within the space collected by single-line lidar is obtained through the following methods:

[0012] Acquire initial point cloud data within the space collected by a single-line lidar;

[0013] Perform at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space.

[0014] Furthermore, the step of performing at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space includes:

[0015] The initial point cloud data is subjected to pass-through filtering, voxel filtering, and outlier filtering in sequence to determine the target point cloud data in space.

[0016] Furthermore, determining the internal spatial point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data includes:

[0017] The boundary point cloud data and the target point cloud data are deduplicated to determine the internal spatial point cloud data corresponding to the space.

[0018] Furthermore, the foreground target features include limb features. The step of extracting foreground target features from the internal spatial point cloud data to determine the number of foreground target features within the space includes:

[0019] Based on the internal spatial point cloud data and the preset Hough circle transform rule, the limb features are extracted from the internal spatial point cloud data within the space to determine the number of limb features within the space.

[0020] Furthermore, the step of extracting boundary features from the target point cloud data within the space based on the target point cloud data to determine the boundary point cloud data corresponding to the space includes:

[0021] Based on the target point cloud data and the preset Hough line transform rule, the boundary features of the target point cloud data in the space are extracted to determine the boundary point cloud data corresponding to the space.

[0022] Furthermore, determining the number of target individuals within the space based on the number of foreground target features includes:

[0023] The number of limb features is reduced by half, and the reduced number of limb features is determined as the target number of people in the space.

[0024] This application embodiment also provides a robot people recognition device, the robot people recognition device comprising:

[0025] The acquisition module is used to acquire target point cloud data within the space collected by a single-line lidar.

[0026] The first determining module is used to extract boundary features from the target point cloud data in the space based on the target point cloud data, and determine the boundary point cloud data corresponding to the space.

[0027] The second determining module is used to determine the internal space point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data.

[0028] The third determining module is used to extract foreground target features from the internal spatial point cloud data based on the internal spatial point cloud data, and determine the number of foreground target features in the space.

[0029] The fourth determining module is used to determine the number of targets in the space based on the number of foreground target features.

[0030] This application also provides an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the robot people recognition method described above are performed.

[0031] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the robot people recognition method described above.

[0032] Compared with existing methods for recognizing people, the robot people recognition method, device, and electronic equipment provided in this application use a single-line lidar to collect target point cloud data within a space. Based on this target point cloud data, boundary features are extracted to determine the corresponding boundary point cloud data. Then, based on the boundary point cloud data and the target point cloud data, the corresponding internal space point cloud data is determined. Finally, the internal space point cloud data is extracted from the target point cloud data, and foreground target features are extracted to determine the number of foreground target features within the space. Based on the number of target features, the number of target people within the space is determined. This enables the simultaneous identification of obstacles and the number of target people within the space when a mobile robot is using a passenger elevator where humans and robots coexist, thereby improving the service efficiency and accuracy of the mobile robot.

[0033] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 One of the flowcharts of a robot people recognition method provided in this application embodiment is shown;

[0036] Figure 2 A second flowchart of a robot people recognition method provided in an embodiment of this application is shown;

[0037] Figure 3 This paper shows a schematic diagram of the structure of a robot people recognition device provided in an embodiment of this application;

[0038] Figure 4 A schematic diagram of the structure of an electronic device is shown.

[0039] In the picture:

[0040] 300 - Robot personnel recognition device; 310 - Acquisition module; 320 - First determination module; 330 - Second determination module; 340 - Third determination module; 350 - Fourth determination module; 400 - Electronic device; 410 - Processor; 420 - Memory; 430 - Bus. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. Based on the embodiments of this application, every other embodiment obtained by those skilled in the art without inventive effort falls within the scope of protection of this application.

[0042] First, the applicable application scenarios of this application will be introduced. This application can be applied to the field of lidar technology.

[0043] The locations provided in the embodiments of this application can be passenger elevators used by mobile robots to serve users, or field robots operating in locations such as manholes, tunnels, and pump rooms. That is, this application is not limited to scenarios involving identifying the number of passengers in a passenger elevator.

[0044] Research has found that in traditional service industries, mobile robots often need to use dedicated robotic freight elevators to move between floors to serve users and meet their needs. However, as people's living standards improve, the scenarios requiring mobile robots are increasing, and single freight elevators are no longer sufficient to provide services. Therefore, mobile robots need to use passenger elevators where humans and robots coexist. However, when mobile robots use passenger elevators where humans and robots coexist, they can only detect the presence of obstacles within a certain space inside the elevator, but they cannot accurately identify the number of people inside. This makes it difficult for mobile robots to move between floors, resulting in low service efficiency and reduced accuracy of their services.

[0045] Based on this, embodiments of this application provide a robot people recognition method, device, and electronic device, which enables the simultaneous determination of obstacles and the number of people in the elevator when the mobile robot is riding in a passenger elevator where humans and robots coexist. This results in lower service efficiency of the mobile robot and reduced accuracy of the mobile robot service.

[0046] Please see Figure 1 , Figure 1 This is one of the flowcharts for a robot people recognition method provided in an embodiment of this application. Figure 1As shown in the figure, the robot people recognition method provided in this application includes the following steps:

[0047] S101. Acquire target point cloud data within the space collected by single-line lidar.

[0048] In this step, the space in the embodiments provided in this application can be customized according to different actual application scenarios, including but not limited to elevators.

[0049] In this context, single-line lidar refers to lidar where the laser source emits a single beam. It is categorized into triangulation lidar and Time-of-Flight (TOF) lidar, and is primarily used in robotics. It features fast scanning speed, high resolution, and high reliability. Compared to multi-line lidar, single-line lidar reacts more quickly in terms of angular frequency and sensitivity, resulting in more accurate distance and precision for obstacle measurement.

[0050] The following sub-steps are used to acquire target point cloud data within the space collected by a single-line lidar:

[0051] Sub-step 1: Acquire initial point cloud data within the space collected by single-line lidar.

[0052] Sub-step 2: Perform at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space.

[0053] In this step, the initial point cloud data collected by the single-line lidar consists of an array of thousands of distances and angles. Therefore, the initial point cloud data needs to be filtered at least once to remove unnecessary information and retain the necessary information, which is to determine the target point cloud data inside the passenger elevator.

[0054] Here, the step of performing at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space includes:

[0055] The initial point cloud data is subjected to pass-through filtering, voxel filtering, and outlier filtering in sequence to determine the target point cloud data in space.

[0056] Here, the initial point cloud data is first subjected to pass-through filtering. Pass-through filtering is to remove the initial point cloud data within a specified range (or outside) of a certain dimension of the initial point cloud data, and then save the initial point cloud data within the specified range.

[0057] The specific process of the through-pass filtering is as follows:

[0058] Read in the initial point cloud data → Create a filter object → Set the filter field range → Execute the filter → Save the filter result.

[0059] The purpose of this pass-through filter is to save the initial point cloud data inside the elevator and remove the initial point cloud data outside the elevator.

[0060] Next, voxel filtering is performed on the initial point cloud data after pass-through filtering.

[0061] Thus, the principle of voxel filtering is to create a three-dimensional voxel grid (a collection of tiny three-dimensional cubes) from the initial point cloud data, and use the centroid of each voxel to approximate other points in the voxel. Specifically, it is represented by using one approximate point to replace multiple points. The purpose is to optimize the large amount of initial point cloud data, thereby reducing the amount of initial point cloud data without affecting the result. This reduces the computational load of subsequent algorithms, thereby improving computational efficiency and reducing computational cost.

[0062] The voxel filtering process described above is as follows:

[0063] Read in the initial point cloud data → Create a filter object → Set the voxel size → Execute filtering → Save the filtered point cloud.

[0064] The purpose of performing this filter, as described above, is to reduce the amount of initial point cloud data within the elevator.

[0065] Next, outlier filtering is performed on the initial point cloud data after voxel filtering.

[0066] The goal of outlier filtering is to remove these noisy points, as laser point clouds are often accompanied by noise, which complicates the estimation of local point cloud features.

[0067] The principle of outlier filtering is to perform statistical analysis on the neighborhood of each point. Based on the distance distribution characteristics from the point to all neighboring points, some outliers that do not meet the requirements are filtered out. The average distance from each point to its surrounding points is calculated, and if it is higher than the set threshold, the corresponding point is identified as an outlier and will be filtered out.

[0068] The outlier filtering process described above is as follows:

[0069] Read in the initial point cloud data → Create a filter object → Set the outlier threshold → Perform statistical filtering → Save the filtered point cloud.

[0070] The purpose of performing this filter here is to remove noise points inside the elevator.

[0071] S102. Based on the target point cloud data, perform boundary feature extraction on the target point cloud data in the space to determine the boundary point cloud data corresponding to the space.

[0072] In this step, the Hough transform is used to transform the target point cloud data to determine the corresponding boundary point cloud data in space.

[0073] The Hough transform is a feature extraction method widely used in image analysis, computer vision, and digital image processing. It's used to identify features in objects, such as lines. The algorithm generally works as follows: given an object and the type of shape to be identified, the algorithm performs a voting process in the parameter space to determine the object's shape, which is determined by the local maximum in the accumulator space.

[0074] Optionally, step S102 includes the following sub-steps:

[0075] Sub-step 1021: Based on the target point cloud data and the preset Hough line transform rule, extract the boundary features of the target point cloud data in the space to determine the boundary point cloud data corresponding to the space.

[0076] In this step, the pre-set Hough linear transformation rule is first used to determine the boundary features of the passenger elevator in the target point cloud data, and then the boundary features are removed to determine the corresponding boundary point cloud data.

[0077] The goal of removing boundary features is to avoid interference from subsequent target number detection.

[0078] Thus, the principle of the pre-defined Hough line transformation rule is the transformation from the rectangular coordinate system to the polar coordinate system and the transformation from the image space to the parameter space. That is, a straight line in the image space is transformed into a point in the parameter space, and a point in the image space is transformed into a sine curve in the parameter space. Therefore, when points on the same straight line in the image space are transformed into the parameter space, their sine curves must intersect at a point.

[0079] According to the aforementioned pre-defined Hough line transformation rule, as long as the points where curves intersect frequently in the parameter space are determined, they are likely to be straight lines in the image space. Therefore, it is only necessary to transform all the pixels (coordinate values) in the image into curves in the parameter space and detect the intersection points of the curves in the parameter space to determine the straight line (the line composed of boundary features).

[0080] S103. Based on the boundary point cloud data and the target point cloud data, determine the internal space point cloud data corresponding to the space.

[0081] In this step, boundary point cloud data is removed from the target point cloud data, and the remaining target point cloud data is identified as the corresponding internal spatial point cloud data.

[0082] S104. Based on the internal spatial point cloud data, foreground target features are extracted from the internal spatial point cloud data in the space to determine the number of foreground target features in the space.

[0083] In this step, the foreground target feature in the embodiments provided in this application is the user's limb feature. Here, the purpose of distinguishing between foreground target features and background features is to remove impurity features in the internal spatial point cloud data, so as to facilitate the subsequent determination of the number of target people in the space.

[0084] Optionally, the foreground target features include limb features, and step S104 includes the following sub-steps:

[0085] Sub-step 1041: Based on the internal spatial point cloud data and the preset Hough circle transform rule, extract the features of the limb features from the internal spatial point cloud data in the space, and determine the number of the limb features in the space.

[0086] In this step, based on the internal spatial point cloud data and the preset Hough circle transform rule, background features are removed from the internal spatial point cloud data, limb features are extracted, and the number of limb features in the space is determined.

[0087] The principle of the preset Hough circle transformation rule is to confirm that every non-zero pixel point may be a point on a potential circle, then generate a cumulative coordinate plane, and set a cumulative weight to locate the circle.

[0088] The preset Hough circle transformation rule is the process of converting a circle in a two-dimensional image space into a point in a three-dimensional parameter space determined by the radius and the horizontal and vertical coordinates of the circle's center. Therefore, the circle determined by any three points on the circumference should correspond to a point in the three-dimensional parameter space after the Hough transformation. Since the circle equations of all points on the same circle are the same, they are mapped to the same point in the three-dimensional coordinate system. Therefore, this point should have N curves intersecting with the total number of pixels of the circle in the three-dimensional coordinate system. By judging the number of intersections (cumulative) of each point in the three-dimensional coordinate system, points that are greater than a certain threshold are considered to be circles.

[0089] S105. Based on the number of foreground target features, determine the number of target people in the space.

[0090] In this step, the foreground target feature in the embodiments provided in this application can be specifically human legs. At this time, the number of target users in the elevator can be determined based on the number of human legs.

[0091] Optionally, determining the number of target people in the space based on the number of foreground target features includes:

[0092] The number of limb features is reduced by half, and the reduced number of limb features is determined as the target number of people in the space.

[0093] In this scenario, assuming the limb features are the leg features of the users in the elevator, if we want to determine the number of target users in the elevator based on the leg features, we need to divide the number of leg features by 2, round up to get the number of target users in the elevator, and then determine this number of target users as the target number of people in the passenger elevator.

[0094] The robot people recognition method provided in this application, compared with the existing people recognition methods, uses target point cloud data collected by a single-line lidar in space. Based on the target point cloud data, boundary features are extracted from the target point cloud data in space to determine the corresponding boundary point cloud data. Based on the boundary point cloud data and the target point cloud data, the corresponding internal space point cloud data is determined. Then, the target point cloud data is used to determine the corresponding internal space point cloud data. Foreground target features are extracted from the internal space point cloud data in space to determine the number of foreground target features in the space. Based on the number of target features, the number of target people in the space is determined. This enables the mobile robot to simultaneously determine obstacles and the number of target people in the elevator when the mobile robot is riding in a passenger elevator where humans and robots coexist, thereby improving the service efficiency and accuracy of the mobile robot service.

[0095] Please see Figure 2 , Figure 2 This is a second flowchart of a robot people recognition method provided in an embodiment of this application. Figure 2 As shown in the figure, the robot people recognition method provided in this application includes the following steps:

[0096] S201. Acquire target point cloud data within the space collected by a single-line lidar.

[0097] S202. Based on the target point cloud data, perform boundary feature extraction on the target point cloud data in the space to determine the boundary point cloud data corresponding to the space.

[0098] S203. Perform deduplication processing on the boundary point cloud data and the target point cloud data to determine the internal space point cloud data corresponding to the space.

[0099] In this step, the target point cloud data and boundary point cloud data inside the elevator are overlapped, the boundary point cloud data is deleted, and the remaining target point cloud data is identified as the internal space point cloud data corresponding to the space.

[0100] S204. Based on the internal spatial point cloud data, foreground target features are extracted from the internal spatial point cloud data in the space to determine the number of foreground target features in the space.

[0101] S205. Based on the number of foreground target features, determine the number of target people in the space.

[0102] The descriptions of S201 to S202 and S204 to S205 can be referred to the descriptions of S101 to S102 and S104 to S105, and can achieve the same technical effect, so they will not be elaborated further.

[0103] The robot people recognition method provided in this application, compared with the existing people recognition methods, uses target point cloud data collected by a single-line lidar in space. Based on the target point cloud data, boundary features are extracted from the target point cloud data in space to determine the corresponding boundary point cloud data. Based on the boundary point cloud data and the target point cloud data, the corresponding internal space point cloud data is determined. Then, the target point cloud data is used to determine the corresponding internal space point cloud data. Foreground target features are extracted from the internal space point cloud data in space to determine the number of foreground target features in the space. Based on the number of target features, the number of target people in the space is determined. This enables the mobile robot to simultaneously determine the number of people in obstacles and the passenger elevator when the mobile robot is riding in a human-robot coexisting passenger elevator, thereby improving the service efficiency and accuracy of the mobile robot service.

[0104] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a robot people recognition device provided in an embodiment of this application. Figure 3 As shown, the robot people recognition device 300 includes:

[0105] The acquisition module 310 is used to acquire target point cloud data within the space collected by a single-line lidar.

[0106] Optionally, the acquisition module 310 acquires the target point cloud data within the space collected by the single-line lidar in the following manner:

[0107] Acquire initial point cloud data within the space collected by a single-line lidar.

[0108] Perform at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space.

[0109] Optionally, the step of performing at least one data filtering process on the initial point cloud data to determine the target point cloud data within the space includes:

[0110] The initial point cloud data is subjected to pass-through filtering, voxel filtering, and outlier filtering in sequence to determine the target point cloud data in space.

[0111] The first determining module 320 is used to extract boundary features from the target point cloud data in space based on the target point cloud data, and determine the boundary point cloud data corresponding to the space.

[0112] Optionally, the first determining module 320 is specifically used for:

[0113] Based on the target point cloud data and the preset Hough line transform rule, the boundary features of the target point cloud data in the space are extracted to determine the boundary point cloud data corresponding to the space.

[0114] The second determining module 330 is used to determine the internal spatial point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data.

[0115] Optionally, the second determining module 330 is specifically used for:

[0116] The boundary point cloud data and the target point cloud data are deduplicated to determine the internal spatial point cloud data corresponding to the space.

[0117] The third determining module 340 is used to extract foreground target features from the internal spatial point cloud data based on the internal spatial point cloud data, and determine the number of foreground target features in the space.

[0118] Optionally, the foreground target features include limb features, and the third determining module 340 is specifically used for:

[0119] Based on the internal spatial point cloud data and the preset Hough circle transform rule, the limb features are extracted from the internal spatial point cloud data within the space to determine the number of limb features within the space.

[0120] The fourth determining module 350 is used to determine the number of target people in the space based on the number of foreground target features.

[0121] Optionally, the fourth determining module 350 is specifically used for:

[0122] The number of limb features is reduced by half, and the reduced number of limb features is determined as the target number of people in the space.

[0123] The robot people recognition device 300 provided in this application embodiment, compared with the existing people recognition methods, uses target point cloud data collected by a single-line lidar in space, and extracts boundary features from the target point cloud data in space based on the target point cloud data to determine the corresponding boundary point cloud data. Based on the boundary point cloud data and the target point cloud data, it determines the corresponding internal space point cloud data. Then, based on the target point cloud data, it determines the corresponding internal space point cloud data. Foreground target features are extracted from the internal space point cloud data in space to determine the number of foreground target features in space, and the number of target people in space is determined based on the number of target features. This enables the mobile robot to simultaneously determine the number of people in the space (elevator) when taking a passenger elevator where humans and robots coexist, thereby improving the service efficiency and accuracy of the mobile robot service.

[0124] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 400 includes a processor 410, a memory 420, and a bus 430.

[0125] The memory 420 stores machine-readable instructions executable by the processor 410. When the electronic device 400 is running, the processor 410 communicates with the memory 420 via the bus 430. When the machine-readable instructions are executed by the processor 410, they can perform the operations described above. Figure 1 as well as Figure 2 The steps of the robot people recognition method in the method embodiment shown are specifically implemented in the method embodiment and will not be repeated here.

[0126] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can perform the above-described actions. Figure 1 as well as Figure 2 The steps of the robot people recognition method in the method embodiment shown are specifically implemented in the method embodiment and will not be repeated here.

[0127] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0128] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0129] The units described 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.

[0130] In addition, the functional units in the various embodiments of this application 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.

[0131] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0132] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, 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 covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A machine person number recognition method, characterized by, The method for recognizing the number of people by the robot comprises the following steps: acquiring target point cloud data in a space collected by a single-line laser radar; extracting boundary features of the target point cloud data in the space based on the target point cloud data, and determining boundary point cloud data corresponding to the space; determining internal space point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data; extracting foreground target features of the internal space point cloud data in the space based on the internal space point cloud data, and determining the number of foreground target features in the space; determining the number of target people in the space based on the number of foreground target features; the acquiring target point cloud data in a space collected by a single-line laser radar comprises the following steps: acquiring initial point cloud data in a space collected by a single-line laser radar; performing straight-through filtering, voxel filtering and outlier filtering on the initial point cloud data in sequence, and determining target point cloud data in the space; the straight-through filtering is used to remove the initial point cloud data outside the elevator, and comprises the following steps: for a certain dimension of the initial point cloud data, removing the initial point cloud data outside a specified range, and then saving the initial point cloud data in the specified range; the voxel filtering is used to reduce the number of the initial point cloud data in the elevator, and comprises the following steps: creating a three-dimensional voxel grid for the initial point cloud data in the specified range, and replacing other points in the voxel with the center of each voxel; the outlier filtering is used to remove noise points in the elevator, and comprises the following steps: statistically analyzing the neighborhood of each point in the three-dimensional voxel grid, and filtering out outlier points that do not meet the requirements based on the distance distribution characteristics of the point to all neighboring points.

2. The machine person number recognition method according to claim 1, characterized in that, the determining internal space point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data comprises the following steps: performing deduplication processing on the boundary point cloud data and the target point cloud data, and determining the internal space point cloud data corresponding to the space. 3.The machine person number recognition method of claim 1, wherein, The foreground target features comprise limb features, and the extracting foreground target features of the internal space point cloud data in the space based on the internal space point cloud data, and determining the number of foreground target features in the space comprises the following steps: extracting the limb features of the internal space point cloud data in the space according to the internal space point cloud data and a preset Hough circle transformation rule, and determining the number of limb features in the space. 4.The machine person number recognition method of claim 1, wherein, The extracting boundary features of the target point cloud data in the space based on the target point cloud data, and determining the boundary point cloud data corresponding to the space comprises the following steps: extracting boundary features of the target point cloud data in the space according to the target point cloud data and a preset Hough straight line transformation rule, and determining the boundary point cloud data corresponding to the space.

5. The machine person number recognition method according to claim 3, characterized in that, The determining the number of target people in the space based on the number of foreground target features comprises the following steps: halving the number of limb features, and determining the number of limb features after the halving as the number of target people in the space.

6. A machine person number recognition device characterized by, The device for recognizing the number of people by the robot comprises: an acquisition module, configured to acquire target point cloud data in a space collected by a single-line laser radar; The first determination module is configured to perform boundary feature extraction on the target point cloud data in the space based on the target point cloud data, and determine boundary point cloud data corresponding to the space. The second determination module is configured to determine internal space point cloud data corresponding to the space based on the boundary point cloud data and the target point cloud data. The third determination module is configured to perform foreground target feature extraction on the internal space point cloud data in the space based on the internal space point cloud data, and determine a foreground target feature quantity in the space. The fourth determination module is configured to determine a target person quantity in the space based on the foreground target feature quantity. When the acquisition module is used to acquire target point cloud data in a space collected by a single-line laser radar, the acquisition module is specifically configured to: acquire initial point cloud data in the space collected by the single-line laser radar; perform, in sequence, straight-through filtering processing, voxel filtering processing, and outlier filtering processing on the initial point cloud data, and determine target point cloud data in the space; wherein the straight-through filtering processing is used to remove initial point cloud data outside the elevator, and includes: for a certain dimension of the initial point cloud data, removing initial point cloud data outside a specified range, and then saving initial point cloud data in the specified range; the voxel filtering processing is used to reduce the quantity of initial point cloud data in the elevator, and includes: creating a three-dimensional voxel grid for the initial point cloud data in the specified range, and replacing other points in the voxel with a center of gravity of each voxel; and the outlier filtering processing is used to remove noise points in the elevator, and includes: performing statistical analysis on a neighborhood of each point in the three-dimensional voxel grid, and filtering out outlier points that do not meet the requirements based on distance distribution characteristics of the point to all neighboring points.

7. An electronic device, comprising: The processor, the memory, and the bus, the memory stores machine readable instructions executable by the processor, when the electronic device runs, the processor and the memory are communicated through the bus, the machine readable instructions are executed by the processor, and the steps of the machine person number identification method in any one of the above claims 1 to 5 are executed. The computer readable storage medium stores a computer program, and the computer program is executed by the processor, and the steps of the machine person number identification method in any one of the above claims 1 to 5 are executed.

8. A computer-readable storage medium, characterized in that, ​

Images