Human body state detection method and device, computer device and computer storage medium

Abstract

The application relates to a human body state detection method and device, computer equipment and a computer storage medium. The method comprises the following steps: based on a plurality of frames of human body point cloud data of a monitoring area collected by a radar device, the number of clustering centers of each frame of human body point cloud data is determined, wherein the human body point cloud data comprises motion point data and static point data; based on the number of clustering centers of each frame of human body point cloud data, the number of motion point data and the number of static point data, the human body state in the monitoring area is determined. By adopting the method, the problems of false positives caused by multipath reflection in an unmanned environment and false positives caused by long-time static in a manned environment can be solved, and the accuracy of human body state detection is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image detection, in particular to a human state detection method and device, computer equipment and computer storage medium. BACKGROUND

[0002] Developing smart medical care has become an important part of building a smart city. Millimeter wave radar is not affected by light, can monitor human bodies by emitting electromagnetic waves without contact, has low power consumption, and can effectively protect the privacy of monitored personnel. Therefore, applying wireless radio frequency technology to smart medical care has become a trend of future social development and a direction for major high-tech companies to compete.

[0003] At present, millimeter wave radar can be applied to safety warning of indoor restricted areas for children, intelligent lighting based on human state, etc., but when detecting human state, there are still problems of false positives caused by multipath reflection interference in unoccupied environment and false positives caused by long-term inactivity in occupied environment. SUMMARY

[0004] Therefore, it is necessary to provide a human state detection method, device, computer equipment and computer readable storage medium capable of solving the problems of false positives caused by multipath interference in unoccupied environment and false positives caused by long-term inactivity in occupied environment.

[0005] In a first aspect, the present application provides a human state detection method, which comprises:

[0006] Based on the multi-frame human point cloud data of the monitoring area collected by the radar device, the number of clustering centers of each frame of human point cloud data is determined; the human point cloud data includes motion point data and stationary point data;

[0007] Based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data, the human state in the monitoring area is determined.

[0008] In one embodiment, the determination of the human state in the monitoring area based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data comprises:

[0009] Based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data in turn, the number of frames continuously satisfying the corresponding human state judgment condition is determined.

[0010] Based on the number of frames, human state flag information is determined.

[0011] Determine a human state based on the human state flag information.

[0012] In one of the embodiments, the human state includes a no human state, a human moving state, and a human static state; and the determining of the frame number continuously satisfying the corresponding human state determination condition based on the number of the clustering centers of the human point cloud data of each frame, the number of the moving point data, and the number of the static point data includes:

[0013] In the case that the number of the clustering centers is zero, if the number of the moving point data and the number of the static point data satisfy a no human state determination condition, update a first frame number continuously satisfying the no human state determination condition; if the number of the moving point data satisfies a first human moving state determination condition, or the number of the moving point data and the number of the static point data satisfy a second human moving state determination condition, update a second frame number continuously satisfying the first human moving state determination condition or the second human moving state determination condition; and if the number of the moving point data and the number of the static point data satisfy a first human static state determination condition, update a third frame number continuously satisfying the first human static state determination condition.

[0014] In the case that the number of the clustering centers is not zero, if the number of the moving point data and the number of the static point data satisfy a third human moving state determination condition, update the second frame number continuously satisfying the third human moving state determination condition; and if the number of the moving point data satisfies a second human static state determination condition, update a fourth frame number continuously satisfying the second human static state determination condition.

[0015] In one of the embodiments, the determining of the human state flag information based on the frame number includes:

[0016] If the frame number satisfies a corresponding threshold condition, determine the corresponding human state flag information.

[0017] In one of the embodiments, the method further includes:

[0018] Determine the distance information from the radar device to the human based on the number of the clustering centers of the current frame human point cloud data and the moving point data or the static point data in the current frame human point cloud data.

[0019] In one of the embodiments, the distance information from the radar device to the human is first distance information, and the determining of the distance information from the radar device to the human based on the number of the clustering centers of the current frame human point cloud data and the moving point data or the static point data in the current frame human point cloud data includes:

[0020] In a case where the number of cluster centers of the current frame human body point cloud data is zero, the first distance information is determined based on motion point data in the current frame human body point cloud data;

[0021] In a case where the number of cluster centers of the current frame human body point cloud data is not zero, the first distance information is determined based on stationary point data in the current frame human body point cloud data.

[0022] In one of the embodiments, in a case where the number of cluster centers of the current frame human body point cloud data is not zero, the determination of the first distance information based on the stationary point data in the current frame human body point cloud data comprises:

[0023] based on the associated point cloud data corresponding to each cluster center;

[0024] performing bubble sort on the number of the associated point cloud data to determine a cluster center point corresponding to a maximum value of the number of the associated point cloud data;

[0025] performing bubble sort on the position of the associated point cloud data to update coordinate information corresponding to the maximum value of the number of the associated point cloud data.

[0026] In one of the embodiments, the determination of the first distance information based on the stationary point data in the current frame human body point cloud data comprises:

[0027] based on the number of each cluster center, obtaining a single frame coordinate of the stationary point data corresponding to each cluster center on a single frame;

[0028] based on each single frame coordinate, obtaining an instantaneous distance from the radar device to the human body;

[0029] determining whether the instantaneous distance satisfies a distance validity determination condition to determine whether the instantaneous distance is valid.

[0030] In a second aspect, the present application further provides a human body state detection device, which comprises:

[0031] a data acquisition module configured to determine the number of cluster centers of each frame of human body point cloud data based on a plurality of frames of human body point cloud data of a monitoring area collected by a radar device; the human body point cloud data comprises motion point data and stationary point data;

[0032] a state determination module configured to determine the state of a human body in the monitoring area based on the number of cluster centers of each frame of human body point cloud data, the number of motion point data and the number of stationary point data.

[0033] In a third aspect, the present application also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to realize the content of the first aspect.

[0034] In a fourth aspect, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the content of the first aspect.

[0035] The human body state detection method, device, computer device and computer readable storage medium solve the problems of false positives caused by multipath reflection in an unmanned environment and false positives caused by long-time stillness in a manned environment, and improve the accuracy of detecting human body states based on millimeter wave radar. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 An application environment diagram of the human body state detection method in an embodiment;

[0037] Figure 2 A flowchart of the human body state detection method in an embodiment;

[0038] Figure 3 A flowchart of S204 in the human body state detection method in an embodiment;

[0039] Figure 4 A flowchart of the human body state detection method in another embodiment;

[0040] Figure 5 A flowchart before determining the distance information from the radar device to the human body in an embodiment;

[0041] Figure 6 A flowchart of determining the distance information from the radar device to the human body in an embodiment;

[0042] Figure 7 A structural block diagram of the human body state detection device in an embodiment;

[0043] Figure 8 An internal structure diagram of the computer device in an embodiment. DETAILED DESCRIPTION

[0044] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application and not to limit the present application.

[0045] Unless otherwise defined, technical terms or scientific terms used in the present application should be understood as their common meanings to those skilled in the art to which the present application pertains. The terms "one", "a", "an", "the", and similar terms used in the present application do not represent quantity limitation, but can represent singular or plural. The terms "include", "contain", "have", and any variations thereof used in the present application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device including a series of steps or modules (units) is not limited to the listed steps or units, but can further include steps or units not listed or can further include other steps or units inherent to the process, method, product or device. The terms "connect", "connected", "couple", and similar terms used in the present application are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The term "multiple" in the present application refers to two or more. The term "and / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, "A and / or B" can represent the following three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects. The terms "first", "second", "third", and the like in the present application are only to distinguish similar objects, and do not represent a specific order of the objects.

[0046] The human body state detection method provided by the embodiments of the present application can be applied in an application environment as shown in Figure 1 The terminal 102 communicates with the server 104 through a network. The data storage system 106 can store human point cloud data required to be processed by the server 104. The data storage system 106 can be integrated on the server 104, or placed on a cloud or other network server. The terminal 102 collects multiple frames of human point cloud data of a monitoring area based on a radar device, and sends the human point cloud data to the server 104. The number of clustering centers of each frame of human point cloud data is determined, wherein the human point cloud data includes motion point data and static point data. The human state in the monitoring area is determined based on the number of clustering centers of each frame of human point cloud data, the number of motion point data, and the number of static point data. The terminal 102 can be, but is not limited to, a millimeter wave radar and a personal computer, a notebook computer, a smart phone, a tablet computer, etc. associated with the millimeter wave radar. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0047] In one embodiment, as shown in Figure 2 Fig. 1, a human body state detection method is provided, which is applied to an application environment in Figure 1 Fig. 1, and includes the following steps:

[0048] S202, based on the multi-frame human body point cloud data of the monitoring area collected by the radar device, determining the number of cluster centers of each frame of human body point cloud data.

[0049] Wherein, the radar device can be a millimeter wave radar. The human body point cloud data includes moving point data and static point data. The maximum value of the human body point cloud data is determined by the detection range of the radar device and the hardware memory.

[0050] Specifically, the radar device acquires multi-frame human body point cloud data in the monitoring area, and the data processing module of the radar device divides each frame of human body point cloud data into moving point data and static point data, and stores the X-axis coordinate and Y-axis coordinate of each point into the corresponding X coordinate array and Y coordinate array respectively, and the threshold value of the array size is equal to the maximum value of the frame of human body point cloud data.

[0051] Specifically, the target classification method is used to cluster the point targets of each frame of human body point cloud data, and the number of cluster centers of each frame of human body point cloud data is determined. Based on the cluster centers of each frame of human body point cloud data, the point classification index flag array corresponding to each cluster center and the number storage array are obtained, wherein the point classification index flag array is a two-dimensional array, the rows represent the cluster centers, and the columns represent the cluster points corresponding to each cluster center. The number storage array is a one-dimensional array, which represents the number of cluster points corresponding to each cluster center. The threshold value of the array size is equal to the maximum value of the frame of human body point cloud data. The target classification method includes K-means classification, mean shift classification, density-based classification method and agglomerative hierarchical classification.

[0052] Specifically, when using a millimeter wave radar to cluster the human body point cloud data, one person usually corresponds to one cluster center.

[0053] S204, based on the number of cluster centers of each frame of human body point cloud data, the number of moving point data and the number of static point data, determining the human body state in the monitoring area.

[0054] Wherein, when clustering the human body point cloud data, only the static point data will be clustered, so according to the clustering situation of the static point data, two kinds of decisions are made respectively, that is, the cluster center number is zero and the cluster center number is not zero.

[0055] Specifically, according to the number of clustering centers of each frame of human point cloud data, the human state discrimination is divided into two categories. One category is that when the number of clustering centers is zero, the threshold value of the human state judgment condition is set, and the human state in the monitoring area is determined according to the number of motion points and the number of stationary points of each frame. The other category is that when the number of clustering centers is not zero, the threshold value of the human state judgment condition is set, and the human state in the monitoring area is determined according to the number of motion points and the number of stationary points of each frame.

[0056] Specifically, the human state includes no one state, someone moving state and someone stationary state.

[0057] Optionally, when multiple human states are detected in a frame of human point cloud data, if there is someone moving, the human state is output as someone moving state; if there is only someone stationary state, the human state is output as someone stationary.

[0058] In the above human state detection method, the number of clustering centers of each frame of human point cloud data is determined based on multiple frames of human point cloud data of the monitoring area collected by the radar device, wherein the human point cloud data includes motion point data and stationary point data; and the human state in the monitoring area is determined based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data, which solves the problem of false positives caused by multipath reflection in no one environment and the problem of false positives caused by long time stationary in someone condition, and improves the accuracy of human state detection.

[0059] In one embodiment, as shown in Figure 3 determining the human state in the monitoring area based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data includes:

[0060] S302, the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data are sequentially determined based on the number of clustering centers of each frame of human point cloud data, the number of motion point data and the number of stationary point data.

[0061] In one embodiment, when the number of cluster centers is zero, if the number of moving point data and the number of stationary point data satisfy a no-person condition, a first frame number of continuous satisfaction of the no-person condition is updated; if the number of moving point data satisfies a first person motion condition, or the number of moving point data and the number of stationary point data satisfy a second person motion condition, a second frame number of continuous satisfaction of the first person motion condition or the second person motion condition is updated; and if the number of moving point data and the number of stationary point data satisfy a first person stationary condition, a third frame number of continuous satisfaction of the first person stationary condition is updated.

[0062] The no-person condition is that the number of moving point data is less than a first threshold and the number of stationary point data is less than a second threshold; the first person motion condition is that the number of moving point data is greater than a third threshold; the second person motion condition is that the number of moving point data is greater than N times the number of stationary point data, N being greater than or equal to 3; and the first person stationary condition is that the value of the third frame number is greater than a corresponding third threshold.

[0063] When the number of cluster centers is not zero, if the number of moving point data and the number of stationary point data satisfy a third person motion condition, a second frame number of continuous satisfaction of the third person motion condition is updated; and if the number of moving point data satisfies a second person stationary condition, a fourth frame number of continuous satisfaction of the second person stationary condition is updated.

[0064] The third person motion condition is that the number of moving point data is greater than the third threshold; and the second person stationary condition is that the number of stationary point data is greater than M times the number of moving point data, M being greater than or equal to 3.

[0065] Optionally, when the first frame number is updated, the second frame number and the fourth frame number are set to zero; when the second frame number is updated, the first frame number and the fourth frame number are set to zero; and when the fourth frame number is updated, the first frame number and the second frame number are set to zero.

[0066] S304, determining human state flag information based on the frame number.

[0067] In one embodiment, if the frame number satisfies a corresponding threshold condition, corresponding human state flag information is determined.

[0068] Specifically, in the case that the number of clustering centers is zero, if the value of the first frame number is greater than a first threshold value, the human body state flag information is set to 0, and the third frame number in the step S302 is set to zero; if the value of the second frame number is greater than a second threshold value, the human body state flag information is set to 1, and the third frame number in the step S302 is set to zero; if the value of the third frame number is greater than a third threshold value, the human body state flag information is set to 2.

[0069] Specifically, in the case that the number of clustering centers is not zero, if the value of the second frame number is greater than a second threshold value, the human body state flag information is set to 1; if the value of the fourth frame number is greater than a third threshold value, the human body state flag information is set to 2.

[0070] The first, second, third and fourth threshold values are determined by the frame rate and response time of the radar device.

[0071] S306, determining the human body state in the monitoring area based on the human body state flag information.

[0072] Specifically, when the human body state flag information is 0, the human body state in the monitoring area is no one state; when the human body state flag information is 1, the human body state in the monitoring area is someone moving state; when the human body state flag information is 2, the human body state in the monitoring area is someone static state.

[0073] In the embodiment, by sequentially determining the number of clustering centers of each frame of human body point cloud data, the number of motion point data and the number of static point data, two detection paths of clustering zero and clustering non-zero are divided for each frame of image, the frame number continuously satisfying the corresponding human body state judgment condition is determined, the threshold value corresponding to each frame number is set based on the frame number, thereby determining the human body state flag information, and finally determining the human body state in the monitoring area. On the premise of ensuring timeliness, the decision delay is carried out, the stability of the state value is effectively maintained, the influence of the short-time point cloud distance value change caused by the environmental change or multipath reflection interference on the human body state decision is reduced or maximally inhibited, and the accuracy of the human body state detection is improved.

[0074] In one embodiment, the method further comprises the following steps:

[0075] Based on the number of clustering centers of the current frame of human body point cloud data and the motion point data or the static point data in the current frame of human body point cloud data, the distance information from the radar device to the human body is determined.

[0076] In one of the embodiments, as Figure 4As shown, the determining of the distance information from the radar device to the human body based on the number of cluster centers of the current frame human body point cloud data and the moving point data or the stationary point data in the current frame human body point cloud data includes the following steps:

[0077] S402 : When the number of cluster centers of the human body point cloud data of the current frame is zero, determine the first distance information based on the motion point data in the human body point cloud data of the current frame.

[0078] The first distance information indicates the distance from the radar device to a human body within the monitoring area.

[0079] Specifically, when the number of cluster centers of the human point cloud data in the current frame is zero, if the human state flag information is 1 and the human state in the monitoring area is a moving state, the serial number corresponding to the moving point data is obtained, and bubble sorting is performed from large to small. The X-axis coordinate and Y-axis coordinate of the moving point data corresponding to the middle serial number after the serial number arrangement are selected and recorded, and the distance formula between the two points is used to obtain the first distance information.

[0080] Optionally, when the number of cluster centers of the human body point cloud data of the current frame is zero, if the human body state flag information is 0 and the human body state in the monitoring area is no one, the X coordinate array and the Y coordinate array are all set to 0, and the value corresponding to the first distance information is 0.

[0081] Optionally, when the number of cluster centers of the human body point cloud data in the current frame is zero, if the human body state flag information is 2, and the human body state in the monitoring area is a stationary state, the first distance information of the previous frame is directly used as the first distance information of the current frame.

[0082] S404 : When the number of cluster centers of the human body point cloud data of the current frame is not zero, determine the first distance information based on the static point data in the human body point cloud data of the current frame.

[0083] Specifically, in a specific embodiment, Figure 5 As shown, when the number of cluster centers of the current frame human body point cloud data is not zero, the following steps are included before determining the first distance information based on the stationary point data in the current frame human body point cloud data:

[0084] S502, based on the associated point cloud data corresponding to each cluster center.

[0085] Specifically, in the case that the number of clustering centers of the current frame human body point cloud data is not zero, the stationary point data in the current frame human body point cloud data is classified according to each clustering center, a current clustering position storage two-dimensional array is generated, and the number of each position corresponding to the target reachable point condition is generated current clustering number storage two-dimensional array, wherein the rows of the current clustering position storage two-dimensional array represent each clustering center, and the columns represent the associated point cloud data corresponding to each clustering center; the rows of the current clustering number storage two-dimensional array represent each clustering center, and the columns represent the number of associated point cloud data corresponding to each clustering center. The condition that the position of each point meets the target reachable point includes whether the target reachable point is within the maximum number range of points that can be clustered around the human body point cloud data and whether the distance between the two points is within the maximum distance range of points that can be clustered, wherein the maximum number range and the maximum distance range are determined by the hardware conditions of the radar device.

[0086] Wherein, the row maximum value of the array is equal to the upper limit of the number of clusters, and the column maximum value is equal to the upper limit of the point cloud data of a frame. The upper limit of the number of clusters and the upper limit of the point cloud data of a frame are determined by the system memory of the radar device.

[0087] S504, bubble sorting the number of associated point cloud data to determine the clustering center point corresponding to the maximum number of associated point cloud data.

[0088] Specifically, based on the number of associated point cloud data corresponding to each clustering center, the current clustering number storage two-dimensional array is first sorted in descending order, and the points with the same maximum number of associated point cloud data in the sorting result are obtained. The corresponding current clustering position storage two-dimensional array is also adjusted.

[0089] S506, bubble sorting the position of the associated point cloud data to update the coordinate information corresponding to the maximum number of associated point cloud data.

[0090] Specifically, based on the clustering center points with the same maximum number of associated point cloud data, the position sequence number of the corresponding associated point cloud data is obtained, and the second bubble sorting is performed in descending order. The associated point cloud data corresponding to the median of the position sequence number is selected, and the coordinate information corresponding to the maximum number of associated point cloud data is updated.

[0091] Wherein, according to the frame rate and response time of the radar device, the associated point cloud data of each clustering center is updated in real time, and the corresponding coordinate information is also updated in real time.

[0092] In S502-S506 of the above embodiment, in the case that the number of clustering centers of the current frame human body point cloud data is not zero, the distance value is twice filtered by twice bubble sorting, the number of clustering centers with similar distances is reduced, the invalid target points are deleted, and the efficiency of distance calculation and the accuracy of distance detection are improved.

[0093] Optionally, the initial value of the clustering operation count value is 0, and the clustering operation count value is increased by 1 after each clustering operation when the number of clustering centers of the human body point cloud data is not zero.

[0094] In one embodiment, as shown in Figure 6 the number of clustering centers of the human body point cloud data of the current frame is not zero, the determination of the first distance information based on the static point data in the human body point cloud data of the current frame comprises the following steps:

[0095] S602, based on the number of each clustering center, obtaining the single-frame coordinates of the corresponding static point data on the single frame.

[0096] Specifically, based on the two bubble sorts in S502-S506, the coordinate information of each associated point cloud data corresponding to each clustering center is obtained, and based on the coordinate information of each associated point cloud data, the single-frame coordinates of the corresponding static point data on the single frame are obtained.

[0097] S604, based on each single-frame coordinate, obtaining the instantaneous distance from the radar device to the human body.

[0098] Specifically, based on each single-frame coordinate, the instantaneous distance from the radar device to the human body is calculated using the distance formula between two points, and an instantaneous distance two-dimensional array is generated, the number of rows of the array is the number of clustering centers, the row is used to store the corresponding instantaneous distance from the radar device to the human body, and the column is used to record the value of the number of associated point cloud data corresponding to each clustering center.

[0099] Specifically, the instantaneous distance corresponding to each clustering center calculated for the first time is taken as a standard judgment distance and stored in the first row of the instantaneous distance two-dimensional array, and compared with the subsequent instantaneous distance, and the distance difference between the subsequent instantaneous distance and the standard judgment distance is calculated. If the human state flag information is 1 at this time, the monitoring area is in a state of human motion, then the subsequent instantaneous distance is directly replaced by the original standard judgment distance; if the human state flag information is 2 at this time, the monitoring area is in a state of human static, and the distance difference value is greater than the corresponding distance difference threshold value and the clustering operation count value is greater than the corresponding clustering operation threshold value, then the clustering operation count value is set to zero and the subsequent instantaneous distance is replaced by the original standard judgment distance.

[0100] S606, judging whether the instantaneous distance satisfies the distance effective judgment condition to determine whether the instantaneous distance is effective.

[0101] Specifically, if the value of the first row of the instantaneous distance two-dimensional array is not equal to 0 and greater than a set standard distance threshold, it is determined whether the standard judgment distance meets the distance effective judgment condition. It should be understood that only when the instantaneous distance is determined to be the standard judgment distance, can the effective distance be judged.

[0102] The distance effective judgment condition is that, in the case that the effective distance frame number is greater than or equal to 1, after traversing the standard judgment distances corresponding to all cluster center points of the current frame, if the distance judgment flag bit is 1, the standard judgment distance is an effective distance, and the effective distance is stored in the single-frame distance storage array. When the distance judgment flag bit is 0, the standard judgment distance corresponding to the cluster center point of the current frame is invalid, and the effective distance judgment of the current frame is ended.

[0103] Exemplarily, after storing the standard judgment distance in the first row of the instantaneous distance two-dimensional array, the value of the effective distance frame number corresponding to the standard judgment distance is obtained, wherein the initial value of the effective distance frame number is 0. If the value of the effective distance frame number is less than 1, the standard judgment distance is directly assigned to the single-frame distance storage array for storage, and the value of the effective distance frame number is increased by 1. The single-frame distance storage array stores the standard judgment distances that are determined to be effective distances.

[0104] When the value of the effective distance frame number is greater than or equal to 1, it is sequentially judged whether the standard judgment distance of any associated point cloud data of each cluster center in the single-frame distance storage array of the previous frame is greater than 0. If the standard judgment distances of any associated point cloud data of each cluster center in the single-frame distance storage array of the previous frame are all greater than 0, and the difference between the current frame standard judgment distance and the previous frame standard judgment distance is greater than a set threshold, the distance judgment flag bit is 1, the current frame standard judgment distance is assigned to the single-frame distance storage array, the distance judgment flag bit is 0, and the judgment is exited. If the standard judgment distances of the first n-1 cluster centers in the single-frame distance storage array of the previous frame are all greater than 0, and the difference between the current frame standard judgment distance and the previous frame standard judgment distance is greater than a set threshold, the distance judgment flag bit is 1, the standard judgment distance of the nth cluster center is greater than 0, but the difference between the current frame standard judgment distance and the previous frame standard judgment distance is less than a set threshold, and it is necessary to continue to judge whether the number of associated point cloud data of the current frame standard judgment distance is greater than the number of associated point cloud data of the previous frame standard judgment distance. If it is greater, the current frame standard judgment distance is assigned to the single-frame distance storage array, the distance judgment flag bit is 0, and the judgment is exited.

[0105] Specifically, in the case that the number of cluster centers of the current frame human body point cloud data is not zero, the first distance information is the current frame standard judgment distance stored in the single-frame distance storage array after the effective distance screening of the above steps S602-S606.

[0106] In the above embodiments S602-S606, the distance values after two bubble sorts are further screened by the effective distance judgment, further improving the detection accuracy of the distance from the radar device to the human body.

[0107] In one example embodiment, a human body state detection method is introduced, which specifically includes the following contents:

[0108] Step 1: Based on the millimeter wave radar device, multiple frames of human body point cloud data are collected in the monitoring area, and the maximum value of each frame of human body point cloud data is Matrix_Size. After the data processing module of the millimeter wave radar device, the number of static point data Stop_Num of each frame of human body point cloud data and the X coordinate array x_stop and Y coordinate array y_stop corresponding to each point, the number of moving point data Move_Num and the X coordinate array x_move and Y coordinate array y_move corresponding to each point are output.

[0109] Step 2: Use the target classification method to cluster the point targets of each frame of human body point cloud data, and obtain the number of cluster centers Num_Class, the point classification index flag array Class_Flag_Array and the number storage array Num_Satisfy of each frame of human body point cloud data. The upper limit of the number of cluster centers is PRENUM, which is determined by the hardware equipment of the millimeter wave radar device.

[0110] Step 3: Set the initial value of the cluster operation count value Count_Func as 0, and the value of the cluster operation count value Count_Func is increased by 1 after each cluster operation. The initial values of the first frame number Count_Exist, the second frame number Count_Move, the third frame number Count_Func_Class and the fourth frame number Count_Stop are all 0.

[0111] Step 4: According to whether each frame of human body point cloud data is successfully clustered, the processing process is divided into two categories.

[0112] S41: The first category is when the number of human body point cloud data is zero, small or high degree of dispersion, at this time the number of cluster centers Num_Class is zero, and the following two steps are executed:

[0113] S411: Determine the quantity relationship of the moving point data and the static point data.

[0114] If the number of static point data Stop_Num is less than 2 and the number of moving point data Move_Num is also less than 2, the value of the first frame number Count_Exist is added by 1, and the values of the second frame number Count_Move and the fourth frame number Count_Stop are set to zero;

[0115] If the number of moving point data Move_Num is greater than 5 or the number of moving point data is greater than 3 times the number of static point data Stop_Num, the value of the second frame number Count_Move is added by 1, and the values of the first frame number Count_Exist and the fourth frame number Count_Stop are set to zero;

[0116] Except for the above two cases, the value of the fourth frame number Count_Stop is added by 1, and the values of the first frame number Count_Exist and the second frame number Count_Move are set to zero.

[0117] S412: According to the relationship between the first frame number Count_Exist, the second frame number Count_Move, the third frame number Count_Func_Class, the fourth frame number Count_Stop and the corresponding threshold value, the human body state in the monitoring area is judged.

[0118] If the value of the first frame number Count_Exist is greater than 20, the human body state flag information Stage_Flag is set to 0, 0 represents that the current monitoring area is in an unoccupied state, and the value of the third frame number Count_Func_Class is set to zero. At this time, the X coordinate array and the Y coordinate array are all set to zero, and the value of the first distance information Distance is also all set to zero;

[0119] If the value of the second frame number Count_Move is greater than 5, the human body state flag information Stage_Flag is set to 1, 1 represents that the current monitoring area is in a human motion state, and the value of the third frame number Count_Func_Class is set to zero. At this time, the serial number corresponding to the moving point data is obtained, bubble sorting is performed from large to small, the moving point data corresponding to the middle serial number after the serial number arrangement is selected and recorded, recorded as pmove, the X axis coordinate is x_move[pmove], the Y axis coordinate is y_move[pmove], and the calculation formula of the distance value R_move is:

[0120]

[0121] The calculated distance value R_move is assigned to the first distance information Distance;

[0122] In addition to the above two cases, every frame there is a human point cloud data clustering to zero, the third frame number Count_Func_Class value plus 1, when there are more than 5 consecutive frame human point cloud data clustering to zero, that is, when the third frame number Count_Func_Class value is greater than 5, the human state flag information Stage_Flag is set to 2, 2 represents the current monitoring area is a person static state. The first distance information of the previous frame is assigned to the first distance information Distance of the current frame.

[0123] S42: the second type is in the case of human point cloud data associated, at this time the number of clustering centers Num_Class is not zero, the following steps are executed:

[0124] S421: set the clustering index as L X , the change range is [0, Num_Class); the associated point count value corresponding to the current clustering center is count_y, and the initial value is 0. A current clustering position storage two-dimensional array Class_Position[PRENUM][Matrix_Size] and a current clustering number storage two-dimensional array Num_Position[PRENUM][Matrix_Size] are constructed respectively.

[0125] S422: classify the current frame static point data belonging to different clustering centers. Set the array coordinate index as L y , the initial value is 0, the change range is [0, Num_Input_data), Num_Input_data is the number of actual human point cloud data of the current frame, and the point classification index flag array Class_Flag_Array obtained in step 2 is traversed. If the following equation is satisfied:

[0126] Class_Flag_Array[ y ]= X +1

[0127] , the value of the array coordinate index L y is stored: Class_Position[ X ][count_y]=L y , and the number of points corresponding to the array coordinate index L y satisfying the condition is stored: Class_Position[ X ][count_y]=Num_Satisfy[ y ], and the associated point count value count_y corresponding to the current clustering center is incremented by 1.

[0128] S423: Allocate the cluster center reachable point array Count_Cy with the memory size of PRENUM and the final position storage array Final_P, and make the following judgment, wherein the cluster center reachable point array Count_Cy is used to store the count_y corresponding to the current cluster center, and the final position storage array Final_P is used to store the point corresponding to the maximum value of the associated point cloud data after two bubble sorts.

[0129] If the value of the cluster center reachable point array Count_Cy[ X ] is 0, the final position storage array Final_P[ X ] is equal to 0.

[0130] If the value of the cluster center reachable point array Count_Cy[ X ] is not 0, the data in different columns of each row of the current cluster number storage two-dimensional array Num_Position is sorted in descending order for the first time, and the corresponding current cluster position storage two-dimensional array Class_Position is adjusted.

[0131] The first bubble sort includes the following steps:

[0132] S4231: Set count_max as the column vector count flag bit, the initial value is 0, which represents the statistical value of the points with the same maximum value of the associated point cloud data in the current cluster number storage two-dimensional array Num_Position after the first bubble sort. Allocate the quantity information two-dimensional array Max_Vlaue with the upper limit of the row number PRENUM and the upper limit of the column number Matrix_Size, and the position information two-dimensional array Max_Position, which are used to store the numerical value of the points with the same maximum value of the associated point cloud data and the position information of the corresponding points in the current frame.

[0133] S4232: Screen the current cluster number storage two-dimensional array Num_Position according to the column direction, compare the data Num_Position[ X ][Index] in the same row and different columns with the first data Num_Position[ X ][0] in the row, wherein Index is the index value, which changes in the range of [0, L y ], representing a cluster index L XThe value of the number of associated point cloud data existing in the surroundings. If equal, the index value Index is stored in the number information two-dimensional array Max_Value and the corresponding position information is stored in the position information two-dimensional array Max_Position. The value of the column vector count flag bit count_max is incremented by 1 for each stored point meeting the condition.

[0134] S4233: Determine whether the index value Index is greater than or equal to the cluster center reachable point array Count_Cy[Index]. If less, continue to perform S4232; if greater than or equal to, perform the following steps to complete the second bubble sort.

[0135] S424: Perform a second bubble sort on the data in each row of the position information two-dimensional array Max_Position in descending order, and adjust the corresponding number information two-dimensional array Max_Value.

[0136] The second bubble sort includes the following steps:

[0137] S4241: Allocate a maximum value number storage array Max_Count with a memory size of PRENUM, which is used to store the same points as the maximum value of the associated point cloud data corresponding to the final value of the column vector count flag bit count_max after the first bubble sort.

[0138] S4242: Let P be the position selected in the maximum value number storage array Max_Count, i.e. is the floor symbol. Through the following equation, the position of each cluster center P point in the position information two-dimensional array Max_Position is obtained:

[0139] Final_P[L X ]=Max_Position[L X ][P]

[0140] S4243: Let CX and CY be two-dimensional coordinate arrays with PRENUM columns, where row is used to store the coordinate information of each cluster center P point, row number 0 represents the first row, row number 1 represents the second row, and column number represents each cluster center. According to the values of the X coordinate array x_stop and the Y coordinate array y_stop of the stationary point data obtained in step 1, the coordinate information of each cluster center P point is stored in the second row of the two-dimensional coordinate arrays CX and CY:

[0141] CX[1][L X ]=x_stop[Final_P[L X ]]

[0142] CY[1][L X ]=y_stop[Final_P[L x ]]

[0143] S425: Determine the number relationship between the moving point data and the stationary point data.

[0144] If the number of moving point data Move_Num is greater than 5, then the value of the second frame number Count_Move is increased by 1, and the values of the first frame number Count_Exist and the fourth frame number Count_Stop are set to zero.

[0145] If the number of stationary point data is greater than 3 times the number of moving point data Stop_Num, then the value of the fourth frame number Count_Stop is increased by 1, and the values of the first frame number Count_Exist and the second frame number Count_Move are set to zero.

[0146] Except for the above two cases, the value of the first frame number Count_Exist is set to zero.

[0147] S426: Set the instantaneous distance two-dimensional array Target_Range as a two-row PRENUM column array, where the row is used to store the coordinate information of each cluster center P point corresponding to the instantaneous distance, the row number 0 represents the first row, the row number 1 represents the second row, and the column is used to record the index value of each cluster center. The instantaneous distance of different cluster centers is calculated through the values of the second row of the two-dimensional coordinate array CX and CY, and is stored in the second row of the instantaneous distance two-dimensional array Target_Range:

[0148]

[0149] The first row of the instantaneous distance two-dimensional array Target_Range is used to store the standard judgment distance.

[0150] S427: If the result of the cluster operation count value Count_Func in step 3 is equal to 1, then the values of the second row of the instantaneous distance two-dimensional array Target_Range and the second row of the two-dimensional coordinate array CX and CY are assigned to the first row respectively:

[0151] Target_Range[0][Index]=Target_Range[1][Index]

[0152] CX[0][Index]=CX[1][Index]

[0153] CY[0][Index]=CY[1][Index]

[0154] and a distance difference array Diff_R with a memory size of PRENUM is allocated, and the distance difference value at the corresponding position of the first and second rows of the instantaneous distance two-dimensional array Target_Range is calculated by the following equation:

[0155] Diff_R[Index] = Target_Range[1][Index] - Target_Range[0][Index]

[0156] S428: In the presence of a person, according to the relationship between the second frame number Count_Move and the fourth frame number Count_Stop and their respective threshold values, the human body state in the monitoring area is determined.

[0157] If the value of the second frame number Count_Move is greater than 5, the human body state flag information Stage_Flag is set to 1, 1 indicating that the current monitoring area is in a person moving state, and the values of the second row of the instantaneous distance two-dimensional array Target_Range and the two-dimensional coordinate array CX and CY are assigned to the first row respectively.

[0158] If the value of the fourth frame number Count_Stop is greater than 5, the human body state flag information Stage_Flag is set to 2, 2 indicating that the current detection area is in a person stationary state, according to the result of the distance difference array Diff_R calculated in the above S427, if the distance difference array Diff_R[Index] is greater than 0.2 and the clustering operation count value Count_Func is greater than 10, the values of the second row of the instantaneous distance two-dimensional array Target_Range and the two-dimensional coordinate array CX and CY are assigned to the first row respectively, and the value of the clustering operation count value Count_Func is set to zero.

[0159] In addition to the above two cases, the human body state flag information of the previous frame is assigned to the human body state flag information Stage_Flag of the current frame. Let Num_Class_Last be the previous frame clustering storage flag, indicating the number of storage of the previous frame clustering center point, if the index value Index of the current frame is less than the previous frame clustering storage flag Num_Class_Last of the previous frame, the values of the instantaneous distance two-dimensional array Target_Range and the two-dimensional coordinate array CX and CY are not updated; if the index value Index of the current frame is greater than the previous frame clustering storage flag Num_Class_Last of the previous frame, the values of the second row of the instantaneous distance two-dimensional array Target_Range and the two-dimensional coordinate array CX and CY are assigned to the first row respectively.

[0160] S429: Assign the number of clustering centers of the current frame after clustering Num_Class to the front frame clustering storage flag Num_Class_Last, update the value of the front frame clustering storage flag Num_Class_Last.

[0161] S4210: Allocate memory size PRENUM for the distance record array Record_Range, the coordinate record array Record_CX and Record_CY, and clear the storage unit of the instantaneous distance two-dimensional array Target_Range, the two-dimensional coordinate array CX and CY, the distance record array Record_Range, and the coordinate record array Record_CX and Record_CY in the range of [Num_Class, PRENUM) of the index value Index:

[0162] Target_Range[][Index] = 0

[0163] CX[][Index] = 0; CY[][Index] = 0

[0164] Record_Range[Index] = 0

[0165] Record_CX[Index] = 0; Record_CY[Index] = 0

[0166] S4211: Set Value_Last as a single frame distance storage array for storing the effective distance after filtering of each clustering center, Sum_Last as a single frame statistical value storage array for storing the number of associated point cloud data corresponding to each clustering center after filtering, CX_Last and CY_Last as single frame coordinate storage arrays for storing the coordinate information corresponding to each clustering center after filtering, the above four arrays are one-dimensional arrays, and the memory size is PRENUM. Set the initial value of the effective distance frame number Flag_count as 0, and increase the value of the effective distance frame number Flag_count by 1 after each effective distance judgment. Perform effective distance filtering on the instantaneous distance recorded in the range of the index value Index in [0, Num_Class):

[0167] Specifically, effective distance filtering is only performed when the instantaneous distance two-dimensional array Target_Range[0][Index] is not equal to 0 and the instantaneous distance two-dimensional array Target_Range[0][Index] > 0.1. Including the following two cases:

[0168] The first case, if the value of the effective distance frame number Flag_count is less than 1, the value of the first row of the instantaneous distance two-dimensional array Target_Range[0][Index] (i.e. the standard judgment distance) is directly assigned to the single-frame distance storage array Value_Last, the value of the cluster center reachable point array Count_Cy is assigned to the single-frame statistical value storage array Sum_Last, and the following assignment process is executed:

[0169] Value_Last[Flag_count]=Target_Range[0][Index]

[0170] Sum_Last[Flag_count]=Count_Cy[Index]

[0171] Then perform coordinate assignment:

[0172] CX_Last[Flag_count]=CX[0][Index]

[0173] CY_Last[Flag_count]=CY[0][Index]

[0174] Coordinate storage:

[0175] Record_CX[Flag_count]=CX_Last[Flag_count]

[0176] Record_CY[Flag_count]=CY_Last[Flag_count]

[0177] Distance storage:

[0178] Record_Range[Flag_count]=Value_Last[Flag_count]

[0179] After the assignment is completed, the value of the effective distance frame number Flag_count is incremented by 1.

[0180] The second case, if the value of the effective distance frame number Flag_count is greater than or equal to 1, let flag_result be the index value of the effective distance frame number Flag_count, and its change interval is [0, max), where the value of max is determined by the hardware memory of the millimeter wave radar device.

[0181] Traverse the change interval [0, max), according to the value of the single frame distance storage array Value_Last[lag_result] whether greater than 0, effective distance decision. Including the following two states, the first state: in Value_Last[lag_result]>0 and |Target_Range[0][Index]-Value_Last[lag_result]| value greater than 0.2, the distance judgment flag position 1, until there is Value_Last[lag_result]>0 and |Target_Range[0][Index]-Value_Last[lag_result]| value less than or equal to 0.2, if the current frame clustering center reachable point array Count_Cy[Index] greater than single frame statistics value storage array Sum_Last[lag_result] at this time, then proceed as above the first kind of case of assignment process, assignment after the end, the distance judgment flag position for 0, end the current frame of effective distance decision. The second state: in Value_Last[lag_result]>0 and |Target_Range[0][Index]-Value_Last[lag_result]| value greater than 0.2, the distance judgment flag position for 1, until Value_Last[lag_result]<=0, the distance judgment flag bit is still 1, then proceed as above the first kind of case of assignment process, assignment after the end, the value of the effective distance frame number Flag_count is added 1, the distance judgment flag position for 0, end the current frame of effective distance decision.

[0182] S4212: the value of the single frame distance storage array Value_Last after effective distance screening is assigned to the first distance information Distance.

[0183] Step 5: output the human state and the corresponding first distance information Distance according to the value of the human state flag information Stage_Flag.

[0184] In the above example embodiment, the multi-frame human point cloud data is acquired by the millimeter wave radar device, the human state decision is divided into two categories based on whether the clustering center of the human point cloud data is zero, the array and the set threshold value are used for adjustment, which can effectively distinguish the human state in the monitoring area, not only can judge whether there is a person and distinguish whether the person in the human state is stationary or moving, but also solve the problem of human state decision failure caused by environmental change or multipath interference; When calculating the distance from the millimeter wave radar device to the human body, the two bubble sorting is used to delete the invalid point interference and select the appropriate point as the distance calculation point, which improves the operation efficiency and the accuracy of distance calculation.

[0185] It should be understood that although the steps in the flowcharts involved in the embodiments described above are shown in sequence according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of the steps or stages is not necessarily sequential, but can be alternately or alternately executed with at least part of other steps or steps or stages in other steps.

[0186] Based on the same inventive concept, the embodiments of the present application also provide a human state detection device for implementing the human state detection method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more human state detection device embodiments provided below can refer to the limitations of the human state detection method described above, which will not be repeated here.

[0187] In one embodiment, as shown in Figure 7 A human state detection device is provided, comprising: a data acquisition module 72 and a state judgment module 74, wherein:

[0188] The data acquisition module 72 is configured to determine the number of cluster centers of each frame of human point cloud data based on the multiple frames of human point cloud data of the monitoring area collected by the radar device; the human point cloud data includes motion point data and static point data;

[0189] The state judgment module 74 is configured to determine the human state in the monitoring area based on the number of cluster centers of each frame of human point cloud data, the number of motion point data, and the number of static point data.

[0190] In one embodiment, the data acquisition module 74 comprises:

[0191] sequentially based on the number of cluster centers of each frame of human point cloud data, the number of motion point data, and the number of static point data, determine the number of frames that continuously satisfy the corresponding human state judgment condition;

[0192] Based on the number of frames, determine the human state flag information;

[0193] Based on the human state flag information, determine the human state in the monitoring area.

[0194] In one embodiment, the data acquisition module 74 further comprises:

[0195] In the case that the number of cluster centers is zero, if the number of motion point data and the number of static point data satisfy the no one state judgment condition, the first frame number continuously satisfying the no one state judgment condition is updated; if the number of motion point data satisfies the first person motion state judgment condition, or the number of motion point data and the number of static point data satisfy the second person motion state judgment condition, the second frame number continuously satisfying the first person motion state judgment condition or the second person motion state judgment condition is updated; if the number of motion point data and the number of static point data satisfy the first person static state judgment condition, the third frame number continuously satisfying the first person static state judgment condition is updated.

[0196] In the case that the number of cluster centers is not zero, if the number of motion point data and the number of static point data satisfy the third person motion state judgment condition, the second frame number continuously satisfying the third person motion state judgment condition is updated; if the number of motion point data satisfies the second person static state judgment condition, the fourth frame number continuously satisfying the second person static state judgment condition is updated.

[0197] In one of the embodiments, the data acquisition module 74 further comprises: determining the corresponding human state flag information if the frame number satisfies the corresponding threshold condition.

[0198] In one of the embodiments, the data acquisition module 74 further comprises: determining the distance information from the radar device to the human body based on the number of cluster centers of the current frame human point cloud data and the motion point data or the static point data in the current frame human point cloud data.

[0199] In one of the embodiments, the data acquisition module 74 further comprises:

[0200] In the case that the number of cluster centers of the current frame human point cloud data is zero, the first distance information is determined based on the motion point data in the current frame human point cloud data;

[0201] In the case that the number of cluster centers of the current frame human point cloud data is not zero, the first distance information is determined based on the static point data in the current frame human point cloud data.

[0202] In one of the embodiments, the data acquisition module 74 further comprises:

[0203] Based on the associated point cloud data corresponding to each cluster center;

[0204] The number of associated point cloud data is bubble sorted to determine the cluster center point corresponding to the maximum value of the number of associated point cloud data.

[0205] Bubble sort the positions of the associated point cloud data, and update the coordinate information corresponding to the maximum number of the associated point cloud data.

[0206] In one of the embodiments, the data acquisition module 74 further comprises:

[0207] Based on the number of each cluster center, a single-frame coordinate of the corresponding stationary point data on a single frame is acquired.

[0208] Based on each single-frame coordinate, an instantaneous distance from the radar device to the human body is acquired.

[0209] It is determined whether the instantaneous distance satisfies a distance validity judgment condition, to determine whether the instantaneous distance is valid.

[0210] Each module in the human body state detection device can be realized by software, hardware, and combinations thereof, in whole or in part. Each module described above can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform the operations corresponding to each module.

[0211] In one embodiment, a computer device, which can be a terminal, can have an internal structure as shown in Figure 8 The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to perform wired or wireless communication with external terminals. The wireless communication can be achieved through WIFI, mobile cellular network, NFC (Near Field Communication), or other technologies. The computer program is executed by the processor to implement a human body state detection method. The display unit of the computer device is configured to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, a trackball, or a touchpad arranged on the shell of the computer device. The input device can also be an external keyboard, a touchpad, or a mouse, etc.

[0212] Those skilled in the art can understand, Figure 8 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0213] In one embodiment, a computer device is provided, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0214] In one embodiment, a computer readable storage medium is also provided, which stores a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0215] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0216] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0217] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0218] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for detecting a human body state, characterized in that: The method comprises: Determining the number of cluster centers of each frame of human point cloud data based on multiple frames of human point cloud data collected by the radar device in the monitoring area, wherein the human point cloud data includes moving point data and stationary point data; Determining the state of the human body in the monitoring area based on the number of cluster centers of the human body point cloud data of each frame, the number of the moving point data, and the number of the static point data; The determining of the state of the human body in the monitoring area based on the number of cluster centers of the human body point cloud data of each frame, the number of the moving point data, and the number of the stationary point data includes: Determining the number of frames that continuously meet the corresponding human body state judgment condition based on the number of cluster centers of the human body point cloud data, the number of the moving point data, and the number of the stationary point data in each frame; Based on the number of frames, determining human body status flag information; Based on the human body state flag information, determining the human body state in the monitoring area; the human body state includes an unmanned state, a moving state, and a stationary state; The determining, based on the number of cluster centers of the human body point cloud data of each frame, the number of the moving point data, and the number of the stationary point data, the number of frames that continuously meet the corresponding human body state judgment condition comprises: In the case where the number of cluster centers is zero, if the number of the moving point data and the number of the stationary point data meet the unmanned state judgment condition, then a first number of frames that continuously meet the unmanned state judgment condition is updated; if the number of the moving point data meets the first manned motion state judgment condition, or the number of the moving point data and the number of the stationary point data meet the second manned motion state judgment condition, then a second number of frames that continuously meet the first manned motion state judgment condition or continuously meet the second manned motion state judgment condition is updated; if the number of the moving point data and the number of the stationary point data meet the first manned stationary state judgment condition, then a third number of frames that continuously meet the first manned stationary state judgment condition is updated; When the number of cluster centers is not zero, if the number of motion point data and the number of stationary point data meet the third condition for judging whether a person is in motion, the second number of frames that continuously meet the third condition for judging whether a person is in motion is updated; if the number of motion point data meets the second condition for judging whether a person is in stationary, the fourth number of frames that continuously meet the second condition for judging whether a person is in stationary is updated.

2. The human body state detection method according to claim 1, characterized in that: The determining of human body status flag information based on the number of frames includes: If the number of frames meets the corresponding threshold condition, the corresponding human body state flag information is determined.

3. The human body state detection method according to claim 1, characterized in that: The method further comprises: The distance information from the radar device to the human body is determined based on the number of cluster centers of the current frame human body point cloud data and the moving point data or the stationary point data in the current frame human body point cloud data.

4. The human body state detection method according to claim 3, characterized in that: The distance information from the radar device to the human body is the first distance information, and determining the distance information from the radar device to the human body based on the number of cluster centers of the human body point cloud data of the current frame and the moving point data or the stationary point data in the human body point cloud data of the current frame includes: When the number of cluster centers of the current frame human body point cloud data is zero, determining the first distance information based on the motion point data in the current frame human body point cloud data; When the number of cluster centers of the current frame human body point cloud data is not zero, the first distance information is determined based on the stationary point data in the current frame human body point cloud data.

5. The human body state detection method according to claim 4, characterized in that: In a case where the number of cluster centers of the current frame human body point cloud data is not zero, determining the first distance information based on the stationary point data in the current frame human body point cloud data includes: Based on the associated point cloud data corresponding to each cluster center; performing bubble sorting on the number of associated point cloud data to determine a cluster center point corresponding to a maximum number of associated point cloud data; Bubble sort is performed on the positions of the associated point cloud data, and coordinate information corresponding to the maximum number of the associated point cloud data is updated.

6. The human body state detection method according to claim 4, characterized in that: The determining the first distance information based on the stationary point data in the current frame of human body point cloud data includes: Based on the number of cluster centers, obtaining the single-frame coordinates of the stationary point data corresponding to the single frame; Based on the single-frame coordinates, obtaining the instantaneous distance from the radar device to the human body; It is determined whether the instantaneous distance satisfies a distance validity determination condition, and whether the instantaneous distance is valid.

7. A human body state detection device, characterized in that: The device comprises: A data acquisition module is configured to determine the number of cluster centers of each frame of human point cloud data based on multiple frames of human point cloud data collected by the radar device in the monitoring area; the human point cloud data includes moving point data and stationary point data; A state judgment module is used to determine the state of the human body in the monitoring area based on the number of cluster centers of each frame of human body point cloud data, the number of moving point data, and the number of stationary point data; the human body state includes an unmanned state, a moving state with a person, and a stationary state with a person; The state judgment module includes: when the number of cluster centers is zero, if the number of the moving point data and the number of the stationary point data meet the unmanned state judgment condition, then the first number of frames that continuously meet the unmanned state judgment condition is updated; if the number of the moving point data meets the first manned motion state judgment condition, or the number of the moving point data and the number of the stationary point data meet the second manned motion state judgment condition, then the second number of frames that continuously meet the first manned motion state judgment condition or continuously meet the second manned motion state judgment condition is updated; if the number of the moving point data and the number of the stationary point data meet the first manned motion state judgment condition, or the number of the moving point data and the number of the stationary point data meet the second manned motion state judgment condition, then the second number of frames that continuously meet the first manned motion state judgment condition or continuously meet the second manned motion state judgment condition is updated. If the number of the cluster centers is not zero, and the number of the motion point data and the number of the static point data meet the third condition for judging a person in motion, the second number of frames that continuously meet the third condition for judging a person in motion is updated; if the number of the motion point data meets the second condition for judging a person in static state, the fourth number of frames that continuously meet the second condition for judging a person in static state is updated; based on the number of frames, human body state flag information is determined; based on the human body state flag information, the human body state in the monitoring area is determined.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images