Living body information detection method and apparatus

Abstract

Embodiments of the present application provide a living body information detection method and apparatus. The detection method comprises: dividing a spatial region covered by a radar signal into a plurality of grid spaces; determining a radar signal at each spatial position in each grid space; for each grid space, calculating feature data of the radar signals in the grid space within a first time period, wherein the feature data is an average value of the variations of the radar signals at all spatial positions in the grid space at each moment within the first time period; and inputting the feature data corresponding to all grid spaces into a trained machine learning model so as to detect living body information. In the embodiments of the present application, by delineating grids on a radar coverage region, calculating the feature data for each grid, and outputting the feature data, good detection accuracy is ensured while a plurality of types of information of a living body are detected.

Description

Methods and devices for detecting life information Technical Field

[0001] This application relates to the field of living organism detection, and in particular to a method and apparatus for detecting living organism information. Background Technology

[0002] Currently, the number and proportion of the elderly population are continuously increasing, and the global population is entering an aging stage. Focusing on the physical health and quality of life of the elderly, and addressing the problems they face, has significant social and economic implications. To prevent accidents involving elderly people living alone, smart home devices or elderly care equipment with vital sign detection functions have been widely adopted.

[0003] Common life detection methods include radar-based or video-based life detection. Among them, radar-based life detection technology has a wide range of applications and strong market competitiveness. Radar is a non-contact detection technology, so users do not need to wear devices, making it more convenient to use. In addition, compared with cameras, radar can work normally in the absence of light and does not involve the user's image information, thus better protecting privacy.

[0004] It should be noted that the above introduction to the technical background is only for the purpose of providing a clear and complete explanation of the technical solutions of this application and facilitating understanding by those skilled in the art. It should not be assumed that these technical solutions are known to those skilled in the art simply because they have been described in the background section of this application. Summary of the Invention

[0005] The inventors discovered that in existing radar-based life detection technologies, the existing radar signal processing technology has limited functionality. For example, it can only detect and output information on whether a living being has fallen, and the detection accuracy is low. When the detection area is too large, the detection accuracy cannot be guaranteed.

[0006] To address at least one of the aforementioned problems or other similar issues, embodiments of this application provide a method and apparatus for detecting living organism information. This method divides the radar coverage area into grids, calculates and outputs feature data for each grid, and ensures good detection accuracy while detecting various types of information about living organisms.

[0007] According to one aspect of the embodiments of this application, a living organism information detection device is provided, the device comprising:

[0008] The division unit divides the spatial area covered by the radar signal into multiple grid spaces;

[0009] A determining unit, which determines the radar signal at each spatial location in each of the grid spaces;

[0010] The calculation unit calculates the feature data of the radar signal in the grid space within a first time period for each grid space, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period.

[0011] The detection unit uses the feature data corresponding to all the grid spaces as input to a trained machine learning model to detect living organism information.

[0012] According to another aspect of the embodiments of this application, a method for detecting living organism information is provided, the method comprising:

[0013] The spatial area covered by radar signals is divided into multiple grid spaces;

[0014] Determine the radar signal at each spatial location within each of the grid spaces;

[0015] For each grid space, the feature data of the radar signal in the grid space within the first time period is calculated, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period;

[0016] The feature data corresponding to the grid space is input into the trained machine learning model to detect living organism information.

[0017] According to another aspect of the embodiments of this application, a computer device is provided, the computer device including a memory and a processor, the memory storing a computer program, and the processor being configured to execute the computer program to implement the method as described above.

[0018] According to another aspect of the embodiments of this application, a storage medium storing a computer-readable program is provided, the computer-readable program causing a computer to perform the method described above.

[0019] One of the beneficial effects of this application embodiment is that, according to this application embodiment, the radar coverage area is gridded, feature data is calculated for each grid and output, which ensures good detection accuracy while detecting multiple information about living organisms.

[0020] Specific embodiments of this application are disclosed in detail with reference to the following description and accompanying drawings, indicating how the principles of this application can be adopted. It should be understood that the embodiments of this application are not limited in scope. Within the scope of the appended claims, embodiments of this application include many changes, modifications, and equivalents.

[0021] Features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, combined with features in other embodiments, or substituted for features in other embodiments.

[0022] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, whole, step, or component, but does not exclude the presence or addition of one or more other features, wholes, steps, or components. Attached Figure Description

[0023] The elements and features described in one drawing or embodiment of this application may be combined with elements and features shown in one or more other drawings or embodiments. Furthermore, in the drawings, similar reference numerals denote corresponding parts in several drawings and can be used to indicate corresponding parts used in more than one embodiment.

[0024] The accompanying drawings, which form part of the specification, are used to provide a further understanding of the embodiments of this application and illustrate the implementation methods of this application, together with the textual description, to explain the principles of this application. Obviously, the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any creative effort. In the drawings:

[0025] Figure 1 is a schematic diagram of a life form information detection method according to an embodiment of this application;

[0026] Figure 2 is a schematic diagram of multiple grid spaces according to an embodiment of this application;

[0027] Figure 3 is a schematic diagram of the center position of the grid space and the radar signal in an embodiment of this application, taking Figure 2 as an example;

[0028] Figure 4 is a schematic diagram of a life information detection device according to an embodiment of this application;

[0029] Figure 5 is a schematic diagram of a computer device according to an embodiment of this application. Detailed Implementation

[0030] Referring to the accompanying drawings, the foregoing and other features of this application will become apparent from the following description. Specific embodiments of this application are specifically disclosed in the description and drawings, illustrating partial implementations in which the principles of this application may be employed. It should be understood that this application is not limited to the described embodiments; rather, it includes all modifications, variations, and equivalents falling within the scope of the appended claims.

[0031] In the embodiments of this application, the terms "first," "second," etc., are used to distinguish different elements by name, but do not indicate the spatial arrangement or chronological order of these elements, and these elements should not be limited by these terms. The term "and / or" includes any one or more of the terms listed in association and all combinations thereof. The terms "comprising," "including," "having," etc., refer to the presence of the stated features, elements, components, or assemblies, but do not exclude the presence or addition of one or more other features, elements, components, or assemblies.

[0032] In the embodiments of this application, the singular forms "a," "the," etc., including the plural forms, should be broadly understood as "a kind" or "a class" rather than limited to the meaning of "an." Furthermore, the term "the" should be understood to include both the singular and plural forms, unless the context explicitly indicates otherwise. Additionally, the term "according to" should be understood as "at least partially based on…," and the term "based on" should be understood as "at least partially based on…," unless the context explicitly indicates otherwise.

[0033] The various embodiments of this application will now be described with reference to the accompanying drawings.

[0034] First aspect of the embodiments

[0035] This application provides a method for detecting living organism information, and more specifically, a method for detecting living organisms using radar.

[0036] Figure 1 is a schematic diagram of a life form information detection method according to an embodiment of this application. As shown in Figure 1, the method includes:

[0037] 110: Divide the spatial area covered by radar signals into multiple grid spaces;

[0038] 120: Determine the radar signal at each spatial location within each of the grid spaces;

[0039] 130: For each grid space, calculate the feature data of the radar signal in the grid space within a first time period, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period;

[0040] 140: Input the feature data corresponding to the grid space into the trained machine learning model to detect life information.

[0041] According to the embodiments of this application, the radar coverage area is gridded, and feature data is calculated and output for each grid, which ensures good detection accuracy while detecting a variety of information about living organisms.

[0042] In the embodiments of this application, the radar periodically transmits wireless signals and then receives and processes the wireless signals reflected by the space environment. Taking a frequency modulated continuous wave (FMCW) radar as an example, each frame of radar signal transmitted by an FMCW radar consists of multiple chirp signals, and each chirp signal is a continuous wave with a linearly increasing frequency.

[0043] In this embodiment, an intermediate frequency (IF) signal is obtained by calculating the frequency difference between the received and transmitted signals. The frequency of this signal is proportional to the distance between the target (i.e., the reflection point) and the radar. Therefore, performing a Fast Fourier Transform (FFT) operation on the IF signal of each chirp signal separates the radar signal in the range dimension, obtaining radar signals reflected from reflection points corresponding to each distance. Then, performing a Doppler FFT operation on multiple chirp signals on the same receiving and transmitting antennas at the same distance separates the radar signals in the Doppler velocity dimension, thus obtaining radar signals reflected from objects at different velocities at that distance. Finally, performing an angular FFT on the chirp signals on different receiving and transmitting antennas separates the radar signals in terms of angle, obtaining radar signals reflected in different horizontal and vertical angular directions. Optionally, the Doppler FFT operation can be omitted. After the above signal processing, signals reflected from different distances and angles, i.e., radar signals reflected from various three-dimensional spatial locations, can be obtained.

[0044] In this embodiment of the application, s can be used x,y,z This represents the radar signal reflected from spatial locations (x, y, z). These spatial locations are not continuous in the radar data but are discrete and finite in number. Therefore, (x, y, z) can be used as the reference point. i ,y i ,z i () represents the i-th spatial location in the radar signal coverage area, where 1≤i≤N c N c The total number of spatial locations in the radar coverage area, and the radar signal s x,y,z It can be represented using complex numbers, for example, using the following formula: s x,y,z =A x,y,z expiφ x,y,z

[0045] Among them, A x,y,z φ represents the amplitude of the radar signal. x,y,z This indicates the phase of the radar signal.

[0046] In this application embodiment, millimeter-wave radar is used as an example, such as FMCW radar. This application is not limited to this, and the radar can also be other types of radar, as long as it can generate multiple radar signals.

[0047] Figure 2 is a schematic diagram of multiple grid spaces in an embodiment of this application.

[0048] In this embodiment of the application, as shown in Figure 2, the spatial area covered by the radar signal is defined as a cuboid space, which is then divided into multiple grid spaces. This cuboid space can be as large as possible of the spatial area covered by the radar signal, and should ideally encompass that spatial area.

[0049] Taking Figure 2 as an example, the spatial area covered by the radar signal (not shown in the figure) is defined as a cuboid space 200, with a length of L, a width of W, and a height of H. The length L of the cuboid space 200 is divided into N equal parts. L The width W is divided into N equal parts. W Parts, high H is divided into N H Therefore, the rectangular prism space 200 is divided into multiple grid spaces with length ΔL, width ΔW, and height ΔH, where ΔL = L / N L ΔW=W / N W ΔH=H / N H Therefore, the cuboid space 200 in Figure 2 can be divided into N... L ×N W ×N H A grid space.

[0050] In the above embodiments, taking Figure 2 as an example, the multiple grid spaces are of the same size and are cubes. This application is not limited to this. The sizes of the multiple grid spaces may be different, and the shape of the grid spaces may be cuboids or other shapes, as long as they can completely cover or basically include the spatial area covered by the radar signal.

[0051] In this embodiment of the application, as shown in Figure 2, the position of each grid space is represented by the coordinates of its center position. For example, the position of the i-th grid space can be represented as:

[0052] In the above embodiments, the center position of the grid space is the geometric center of the grid space. This application is not limited to this, and other positions can also be used to represent the position of the grid space.

[0053] Figure 3 is a schematic diagram of the center position of the grid space and the radar signal in an embodiment of this application, taking Figure 2 as an example.

[0054] In some embodiments, for all spatial locations within the spatial area covered by radar signals, the distance from that spatial location to the center of all grid spaces is calculated; the grid space closest to that spatial location is taken as the grid space to which the radar signal reflected from that spatial location belongs, thereby enabling the acquisition of all radar signals contained in each grid space.

[0055] Taking Figure 3 as an example, g1 and g2 are the grid centers of grids G1 and G2, respectively. For the radar signal S1 reflected from one of the spatial locations, calculate the distance from that spatial location to the grid centers g1 and g2, as well as the centers of other grid spaces (not shown in the figure). That is, calculate s. x,y,z arrive The distance, where i is an integer and 1≤i≤N. L *N W *N H The calculation shows that the spatial location is closest to the grid center g1, therefore grid space G1 is the grid space to which radar signal S1 belongs. In other words, grid space G1 contains radar signal S1. Using this method, it is possible to obtain the radar signals contained in all grid spaces.

[0056] The above example uses the center position of the grid space as the grid space location. As mentioned earlier, this application is not limited to this. The grid space location can also be other than the center position. The calculation method is similar and will not be repeated here.

[0057] Therefore, the coverage area of ​​the radar signal is gridded, and the radar signal contained in each grid is calculated so that subsequent calculations can be performed on each grid, thereby improving the accuracy of life form information detection.

[0058] In the above embodiments, the distance between the spatial location of the reflected radar signal and the center of the grid space can be calculated using a spatial coordinate system or other methods. The method for calculating the distance between points in space can refer to the prior art, and this application does not impose any limitations.

[0059] In this embodiment of the application, after calculating the radar signal contained in each grid space, the characteristic data of the radar signal contained in each grid space within the first time period are calculated. The first time period can refer to the cycle of human respiratory characteristics and is at least longer than a normal respiratory cycle, such as 5 seconds or 10 seconds, so as to ensure that the information of the living body in one respiratory cycle can be completely detected and reduce the impact of external or other sudden factors on the detection results.

[0060] In the above embodiment, the characteristic data of the radar signal is the average value of the changes in all radar signals in the grid space at each time point in the first time period. The average value of the changes in the radar signal can be the average value of the amplitude fluctuation of the radar signal, the average value of the phase change of the radar signal, and so on.

[0061] For example, after the above operations, each grid space contains all radar signals reflected from the spatial locations closest to that grid space, denoted by s. G Represented. The radar periodically generates radar signals; the radar signal in the grid space at time t is represented by... This indicates that the radar signals in the grid space during time period T are used... express, By calculating the radar signals in each grid space over a period of time, feature data for subsequent detection models can be obtained.

[0062] For example, firstly, the radar signal in the grid space at time t is calculated to obtain the characteristic data of the grid space at the corresponding time. Then, the feature data of each moment in the grid space within the time period T are integrated to obtain the feature data of the radar signal in the grid space within that time period.

[0063] The following describes the characteristic data of radar signals within the grid space during the first time period.

[0064] In the above embodiments, to calculate the average value of the change in radar signal in the first time period, it is first necessary to calculate the average value of the change in radar signal at a certain moment. For example, when the average value of the amplitude fluctuation of radar signal is used as the average value of the change in radar signal, the average value of the amplitude fluctuation of radar signal at all spatial locations contained in grid space G at time t can be calculated using the following formula:

[0065] Where, N G This represents the number of radar signals contained in the grid space G. This represents the mean amplitude of the radar signal at spatial location (x, y, z). This represents the amplitude of the radar signal at spatial location (x, y, z) at time t. This refers to the change in the amplitude of the radar signal at spatial location (x, y, z) at time t. The mean value of the change in radar signal amplitude at all spatial locations in the grid space G is obtained by summing the changes or absolute values ​​of these changes and then dividing by the number of radar signals in the grid space G.

[0066] For example, when the mean of the phase change of the radar signal is taken as the mean of the change in the radar signal, the mean of the phase change of the radar signal at all spatial locations contained in the grid space G at time t can be calculated using the following formula:

[0067] Where, N G This represents the number of radar signals contained in the grid space G. This represents the mean value of the radar signal phase at spatial location (x, y, z). This represents the phase of the radar signal at spatial location (x, y, z) at time t. This refers to the phase change of the radar signal at spatial location (x, y, z) at time t. The mean value of the phase change of the radar signal at all spatial locations in the grid space G is obtained by summing the phase changes or absolute values ​​of these changes and then dividing by the number of radar signals in the grid space G.

[0068] Therefore, by using the average value of radar signal amplitude or phase changes as characteristic data of radar signals, it is possible to more accurately measure radar signal fluctuations and precisely detect relevant information about living organisms.

[0069] In the above embodiments, the average value of amplitude change and the average value of phase change were used as examples to illustrate the average value of radar signal change. This application is not limited to this, and the average value of radar signal change can also be calculated directly, for example, using the following formula:

[0070] Where, N G This represents the number of radar signals contained in the grid space G. Let G represent the mean of the radar signal at spatial location (x, y, z) at time t. This is obtained by summing the mean values ​​of all radar signals contained in the grid space G, and then dividing by the number of radar signals in the grid space G. at this time, It is a complex number, including both amplitude and phase.

[0071] In this embodiment, the above three schemes are used to calculate the mean value of the change in radar signal contained in the grid space G at time t. Alternatively, one can combine the above three or two of the above schemes, or combine the amplitude and phase with other characteristics of the radar signal to calculate the mean of the change in the radar signal.

[0072] Based on the above embodiment, the mean value of the change in radar signal contained in the grid space G at time t is calculated. Subsequently, the mean value of the changes in radar signals at all spatial locations contained in the grid space G within the first time period, i.e., the characteristic data of the radar signals, can be represented by the following formula:

[0073] Where T represents the first time period, It is the average of the changes in radar signals at all locations contained in the grid space G at each time point within the first time period. It is represented in set form.

[0074] According to the above embodiment, radar signals are allocated to various grid spaces, and feature data is obtained by calculating the radar signals in each grid space. The total number of grid spaces is N. L ×N W ×N H The size of the feature data of the radar signal in each grid space is represented as C×T, where C is the number of channels for the data features. Therefore, the size of the feature data of the input radar signal used for life form detection is represented as N. L ×N W ×N H ×C×T. Therefore, based on this input data, a trained machine learning model can be used to detect the corresponding life form information.

[0075] In the training process of the above machine learning model, the first step is to create a training dataset, that is, to collect a certain amount of radar data, process the radar data using the method described above, obtain input data for life detection, and calibrate the input data to obtain a training dataset. Then, the parameters of the machine learning model are optimized using the training dataset so that the machine learning model can fit the training dataset, thereby obtaining a trained machine learning model.

[0076] Once the trained machine learning model is obtained, it can be applied to new radar data to detect information about living organisms.

[0077] In some embodiments, the life form information can be the presence of life forms in the radar signal coverage area. The machine learning model outputs the probability of the presence of life forms. If the output is greater than a threshold (e.g., 0.5), then life forms exist in the radar signal coverage area; otherwise, life forms do not exist in the radar signal coverage area.

[0078] In other embodiments, the life form information may also be whether a life form exists in a certain grid space. The machine learning model outputs N. L ×N W The matrix, p i,jThis represents the probability that a life form exists on a grid with horizontal and width directions of (i,j). If p i,j If the probability exceeds a threshold, a living being exists on grid (i,j); otherwise, no living being exists on grid (i,j). Using this living being information as the output of a machine learning model, the localization of living beings can be achieved. Furthermore, if the probability of a living being existing on multiple grids meets the above requirements, then multiple living beings exist within the radar coverage area.

[0079] In some other embodiments, the life form information can also be whether a life form has fallen in a certain grid space. The machine learning model outputs N. L ×N W The matrix, p i,j Let p represent the probability of falling on a grid with horizontal and width directions of (i,j). i,j A life form on grid (i,j) with a probability greater than a threshold has fallen; conversely, a life form on grid (i,j) has not fallen or there is no life form on grid (i,j). Using this life form information as the output of a machine learning model, a fall detection function for life forms can be achieved. Furthermore, if the probability of falling on multiple grids meets the above requirements, then there are multiple target fall events within the radar coverage area.

[0080] In this embodiment, a machine learning model can also be constructed or trained based on radar feature data from each grid space over a period of time, so that the output of the machine learning model is the required detection information, thereby obtaining a trained machine learning model. In the above embodiment, radar feature data from each grid space over a period of time is input into the machine learning model, and the size of the input radar feature data is N. L ×N W ×N H ×C×T, where N L ×N W ×N H The number of grid spaces is represented by T, the time period is represented by C, and the number of channels for radar feature data is represented by C. The value of C is different for different radar feature data. For example, when only one of the mean of amplitude change and the mean of phase change of the radar signal is selected as radar feature data, C = 1; when the mean of amplitude change and phase change of the radar signal is selected as radar feature data, C = 2.

[0081] In the above embodiments, the detection of living beings is not limited by the number of people in the radar coverage area, and can detect multiple people. For example, by collecting data from multi-person scenes, labeling the information of all living beings, and training a machine learning model, the detection of multiple people can be achieved.

[0082] In the above embodiments, the required detection information may include whether a living being exists in the radar signal coverage area, the specific location of the living being, and whether the living being has fallen. A machine learning model is constructed based on one of the required detection information to output the required detection information. When the required detection information is whether a living being exists in the radar signal coverage area...

[0083] The machine learning model outputs the probability p of the presence of life in the radar signal coverage area based on radar feature data of each grid space over a period of time. life , and according to p life The numerical value determines whether there are living organisms in the radar signal coverage area. If p life If the value is greater than a preset threshold, then a living organism exists; otherwise, no living organism exists.

[0084] When the required detection information is the location of a life form within the radar coverage area.

[0085] The machine learning model outputs the probability of the presence of life on one or more grid cells based on radar feature data of each grid cell over a period of time. This probability can be expressed by the following formula: P life ={p i,j |1≤i≤N L ,1≤j≤N W}

[0086] Where, p i,j p represents a grid with coordinates (i,j) in both horizontal and width directions. i,j p on the grid life N represents the probability that a living being exists on this grid. L N represents the number of grid cells along the length direction. W This represents the number of grid cells in the width direction. Each grid cell in the horizontal plane of the radar coverage area is analyzed, and p is calculated for each cell. i,j If p life If the value is greater than a preset threshold, then there is a living organism in that grid; otherwise, there is no living organism in that grid. By analyzing whether there is a living organism in each grid on the horizontal plane, the function of locating living organisms can be realized.

[0087] In the above embodiments, p life The threshold is a real number greater than 0 and less than 1, which can be set to 0.5 or other values.

[0088] When the required detection information is whether a living being has fallen within the radar coverage area.

[0089] A machine learning model outputs the probability of an organism falling on one or more grids based on radar data from various grid spaces over a period of time. This probability can be expressed by the following formula: Pfall ={p′ i,j |1≤i≤N L ,1≤j≤N W}

[0090] Where, p′ i,j Let p′ represent a grid with coordinates (i,j) in both horizontal and width directions. i,j p′ on the grid fall N represents the probability that a living being will fall on this grid. L N represents the number of grid cells along the length direction. W This represents the number of grid cells in the width direction. Each grid cell in the horizontal plane of the radar coverage area is analyzed, and p′ for each cell is calculated. fall If p′ fall If the value exceeds a preset threshold, the creature in that grid has fallen; otherwise, the creature in that grid has not fallen. By analyzing whether the creature in each grid on the horizontal plane has fallen and combining the results of all grid spaces, the fall detection of the creature can be achieved.

[0091] Therefore, based on the characteristic data of radar signals in each grid space, different information about living organisms can be detected, thus improving the diversity of living organism information detection.

[0092] In the above embodiments, the machine learning model can be constructed using neural networks or other methods, and this application is not limited thereto. Below, taking the detection of the presence of living organisms in a radar signal coverage area as an example, the construction of a neural network-based machine learning model is explained.

[0093] First, a three-layer fully connected neural network is constructed to convert the radar feature data to be input into a form of size N. L ×N W ×N H A vector of size ×C×T is input into the first layer of the network to obtain a result of size N. feature The feature vectors are then processed by the second layer of the network to obtain a result of size N. feature The feature vector is then processed in the third layer of the network to obtain a one-dimensional output, which represents the probability that there is a life form in the space covered by the radar signal.

[0094] The above are just examples. Other detection methods, such as using machine learning models to detect the location of living beings or information about falling living beings in the radar signal coverage area, can refer to existing technologies and will not be elaborated upon in this application.

[0095] The above embodiments are described using the presence of a living being, the location of the living being, and whether the living being has fallen as detection results. This application is not limited to this, and the machine learning model can output other detection information according to the specific needs of actual operation.

[0096] The above embodiments are merely illustrative examples of the methods described in this application, but this application is not limited thereto, and appropriate modifications can be made based on the above embodiments. For example, the above embodiments can be used alone, or one or more of the above embodiments can be combined.

[0097] According to the embodiments of this application, the radar coverage area is gridded, and feature data is calculated and output for each grid, which ensures good detection accuracy while detecting multiple information about living organisms.

[0098] Second aspect of the embodiments

[0099] This application provides a living organism information detection device. Since the principle by which this device solves the problem is similar to the method of the first aspect embodiment, its specific implementation can refer to the implementation of the method of the first aspect embodiment, and the similarities will not be repeated.

[0100] Figure 4 is a schematic diagram of a life form information detection device according to an embodiment of this application. As shown in Figure 4, the life form information detection device 400 according to an embodiment of this application includes:

[0101] Division unit 401 divides the spatial area covered by radar signals into multiple grid spaces;

[0102] The determining unit 402 determines the radar signal at each spatial location in each of the grid spaces;

[0103] The calculation unit 403 calculates the feature data of the radar signal in the grid space within a first time period for each grid space, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period.

[0104] The detection unit 404 uses the feature data corresponding to all the grid spaces as input to a trained machine learning model to detect living organism information.

[0105] In some embodiments, the partitioning unit defines the spatial area covered by the radar signal as a cuboid space and divides the cuboid space into multiple grid spaces.

[0106] In some embodiments, for all spatial locations within the spatial area covered by the radar signal, the distance from said spatial location to the center of all said grid spaces is calculated;

[0107] The grid space closest to the spatial location is taken as the grid space to which the radar signal reflected from the spatial location belongs, thus obtaining the radar signal contained in each grid space.

[0108] In some embodiments, the first time period is longer than one respiratory cycle.

[0109] In some embodiments, the change in the radar signal is the amplitude fluctuation and / or phase change of the radar signal.

[0110] In some embodiments, when the change in the radar signal is an amplitude fluctuation, the mean value of the change in the radar signal is calculated using the following formula:

[0111] When the change in the radar signal is a phase change, the mean value of the change in the radar signal is calculated using the following formula:

[0112] in, This represents the amplitude of the radar signal at spatial location (x, y, z) at time t. This represents the mean amplitude of the radar signal at spatial location (x, y, z). This represents the phase of the radar signal at spatial location (x, y, z) at time t. N represents the mean phase of the radar signal at spatial location (x, y, z). G This indicates the number of radar signals belonging to a given grid space.

[0113] In some embodiments, as shown in FIG4, the life information detection device 400 further includes a training unit 450, which constructs the machine learning model based on the radar feature data of each grid space over a period of time, so that the output of the machine learning model is the required detection information, thereby obtaining the trained machine learning model.

[0114] In some embodiments, the output of the machine learning model is the probability that a living organism exists in the spatial area covered by the radar signal, and the trained machine learning model is used to detect whether a living organism exists in the spatial area covered by the radar.

[0115] In some embodiments, the output of the machine learning model is the probability of the presence of a living organism in one or more of the grid spaces, and the trained machine learning model is used to detect whether a living organism exists in the one or more grid spaces.

[0116] In some embodiments, the output of the machine learning model is the probability of an organism falling in one or more grid spaces, and the trained machine learning model is used to detect whether the organism has fallen in the one or more grid spaces.

[0117] It is worth noting that the above description only covers the components or modules relevant to this application, but this application is not limited thereto. The life information detection device 400 may also include other components or modules, and for details regarding these components or modules, please refer to relevant technologies.

[0118] For simplicity, Figure 4 only illustrates the connection relationships or signal flow between the various components or modules, but those skilled in the art should understand that various related technologies such as bus connections can be used. The aforementioned components or modules can be implemented using hardware facilities such as processors and memory; this application does not limit this.

[0119] The above embodiments are merely illustrative examples of embodiments of this application, but this application is not limited thereto, and appropriate modifications can be made based on the above embodiments. For example, the above embodiments can be used alone, or one or more of the above embodiments can be combined.

[0120] According to the embodiments of this application, the radar coverage area is gridded, and feature data is calculated and output for each grid, which ensures good detection accuracy while detecting multiple information about living organisms.

[0121] Third aspect of the embodiments

[0122] This application provides a computer device including a life information detection device 400 as described in the second aspect of the embodiment, the contents of which are incorporated herein by reference. This computer device may be, for example, a computer, server, workstation, laptop computer, smartphone, etc.; however, this application is not limited to these.

[0123] Figure 5 is a schematic diagram of a computer device according to an embodiment of this application. As shown in Figure 5, the computer device 500 may include a processor (e.g., a central processing unit, CPU) 510 and a memory 520; the memory 520 is coupled to the central processing unit 510. The memory 520 can store various data; in addition, it also stores an information processing program 521, and executes the program 521 under the control of the processor 510.

[0124] In some embodiments, the functionality of the life information detection device 400 is integrated into a processor 510. The processor 510 is configured to implement the life information detection method as described in the first aspect of the embodiment.

[0125] In some embodiments, the life information detection device 400 is configured separately from the processor 510. For example, the life information detection device 400 can be configured as a chip connected to the processor 510, and the functions of the life information detection device 400 can be realized through the control of the processor 510.

[0126] In addition, as shown in Figure 5, the computer device 500 may also include: input / output (I / O) devices 530 and a display 540, etc.; the functions of the above components are similar to those in the prior art, and will not be described in detail here. It is worth noting that the computer device 500 does not necessarily include all the components shown in Figure 5; furthermore, the computer device 500 may also include components not shown in Figure 5, which can be referred to in related technologies.

[0127] This application also provides a computer-readable program, wherein when the program is executed in a life information detection device, the program causes the life information detection device to perform the method described in the first embodiment.

[0128] This application provides a storage medium storing a computer-readable program, wherein the computer-readable program causes a life information detection device to perform the method described in the first aspect of the embodiment.

[0129] The apparatus and methods described above in this application can be implemented in hardware or in combination with software. This application relates to a computer-readable program that, when executed by a logic component, enables the logic component to implement the apparatus or components described above, or to implement the various methods or steps described above. This application also relates to storage media for storing the above programs, such as hard disks, magnetic disks, optical disks, DVDs, flash memory, etc.

[0130] The methods / apparatus described in conjunction with the embodiments of this application can be directly embodied in hardware, software modules executed by a processor, or a combination of both. For example, one or more and / or combinations of one or more functional block diagrams shown in the figures can correspond to various software modules in a computer program flow, or to various hardware modules. These software modules can correspond to the various steps shown in the figures, respectively. These hardware modules can be implemented, for example, using a field-programmable gate array (FPGA) to embed these software modules.

[0131] The software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. A storage medium can be coupled to the processor, enabling the processor to read information from and write information to the storage medium; or the storage medium can be an integral part of the processor. The processor and storage medium can reside in an ASIC. The software module can be stored in the memory of a mobile terminal or in a memory card that can be inserted into the mobile terminal. For example, if the device (such as a mobile terminal) uses a high-capacity MEGA-SIM card or a high-capacity flash memory device, the software module can be stored in the MEGA-SIM card or the high-capacity flash memory device.

[0132] One or more and / or one or more combinations of functional blocks described in the accompanying drawings can be implemented as a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, or any suitable combination thereof for performing the functions described herein. One or more and / or one or more combinations of functional blocks described in the accompanying drawings can also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in communication with a DSP, or any other such configuration.

[0133] The present application has been described above with reference to specific embodiments. However, those skilled in the art should understand that these descriptions are exemplary and not intended to limit the scope of protection of the present application. Those skilled in the art can make various modifications and variations to the present application based on its spirit and principles, and these modifications and variations are also within the scope of the present application.

[0134] Regarding the above-described embodiments disclosed in this example, the following notes are also disclosed:

[0135] 1. A method for detecting life form information, comprising:

[0136] The spatial area covered by radar signals is divided into multiple grid spaces;

[0137] Determine the radar signal at each spatial location within each of the grid spaces;

[0138] For each grid space, the feature data of the radar signal in the grid space within the first time period is calculated, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period;

[0139] The feature data corresponding to all the grid spaces are input into the trained machine learning model to detect living organism information.

[0140] 2. According to the method described in Appendix 1, dividing the spatial area covered by the radar signal into multiple grid spaces includes:

[0141] The partitioning unit defines the spatial area covered by the radar signal as a cuboid space and divides the cuboid space into multiple grid spaces.

[0142] 3. According to the method described in Appendix 2, determining the radar signal for each spatial location contained in each of the grid spaces includes:

[0143] For all spatial locations within the spatial area covered by the radar signal, calculate the distance from each spatial location to the center of all the grid spaces;

[0144] The grid space closest to the spatial location is taken as the grid space to which the radar signal reflected from the spatial location belongs, thus obtaining the radar signal contained in each grid space.

[0145] 4. According to the method described in Appendix 1, wherein,

[0146] The first time period is longer than one respiratory cycle.

[0147] 5. According to the method described in Appendix 1, wherein,

[0148] The change in the radar signal is the amplitude fluctuation and / or phase change of the radar signal.

[0149] 6. According to the method described in Appendix 5, wherein,

[0150] When the change in the radar signal is an amplitude fluctuation, the mean value of the change in the radar signal is calculated using the following formula:

[0151] When the change in the radar signal is a phase change, the mean value of the change in the radar signal is calculated using the following formula:

[0152] in, This represents the amplitude of the radar signal at spatial location (x, y, z) at time t. This represents the mean amplitude of the radar signal at spatial location (x, y, z). This represents the phase of the radar signal at spatial location (x, y, z) at time t. N represents the mean phase of the radar signal at spatial location (x, y, z).G This indicates the number of radar signals belonging to a given grid space.

[0153] 7. The method according to Appendix 1, wherein the method further comprises:

[0154] The machine learning model is constructed based on the radar feature data of each grid space over a period of time, so that the output of the machine learning model is the required detection information, and the trained machine learning model is obtained.

[0155] 8. According to the method described in Appendix 7, wherein,

[0156] The output of the machine learning model is the probability that a living organism exists in the spatial area covered by the radar signal. The trained machine learning model is used to detect whether a living organism exists in the spatial area covered by the radar.

[0157] 9. According to the method described in Appendix 7, wherein,

[0158] The output of the machine learning model is the probability of the existence of a living organism in one or more of the grid spaces. The trained machine learning model is used to detect whether a living organism exists in the one or more grid spaces.

[0159] 10. The method according to Appendix 7, wherein,

[0160] The output of the machine learning model is the probability of an organism falling down in one or more grid spaces. The trained machine learning model is used to detect whether the organism has fallen down in the one or more grid spaces.

Claims

1. A life form information detection device, wherein, The device includes: The division unit divides the spatial area covered by the radar signal into multiple grid spaces; A determining unit, which determines the radar signal at each spatial location in each of the grid spaces; The calculation unit calculates the feature data of the radar signal in the grid space within a first time period for each grid space, wherein the feature data is the average value of the change in radar signal at all locations in the grid space at each time point within the first time period. The detection unit uses the feature data corresponding to all the grid spaces as input to a trained machine learning model to detect living organism information.

2. The apparatus according to claim 1, wherein, The partitioning unit divides the spatial area covered by the radar signal into multiple grid spaces, including: The partitioning unit defines the spatial area covered by the radar signal as a cuboid space and divides the cuboid space into multiple grid spaces.

3. The apparatus according to claim 2, wherein, The determining unit determines the radar signals for each spatial location contained in each of the grid spaces, including: For all spatial locations within the spatial area covered by the radar signal, calculate the distance from each spatial location to the center of all the grid spaces; The grid space closest to the spatial location is taken as the grid space to which the radar signal reflected from the spatial location belongs, thus obtaining the radar signal contained in each grid space.

4. The apparatus according to claim 1, wherein, The first time period is longer than one respiratory cycle.

5. The apparatus according to claim 1, wherein, The change in the radar signal is the amplitude fluctuation and / or phase change of the radar signal.

6. The apparatus according to claim 5, wherein, When the change in the radar signal is an amplitude fluctuation, the mean value of the change in the radar signal is calculated using the following formula: When the change in the radar signal is a phase change, the mean value of the change in the radar signal is calculated using the following formula: in, This represents the amplitude of the radar signal at spatial location (x, y, z) at time t. This represents the mean amplitude of the radar signal at spatial location (x, y, z). This represents the phase of the radar signal at spatial location (x, y, z) at time t. N represents the mean phase of the radar signal at spatial location (x, y, z). G This indicates the number of radar signals belonging to a given grid space.

7. The apparatus according to claim 1, wherein, The device further includes: The training unit constructs the machine learning model based on radar feature data of each grid space over a period of time, so that the output of the machine learning model is the required detection information, thus obtaining the trained machine learning model.

8. The apparatus according to claim 7, wherein, The output of the machine learning model is the probability that a living organism exists in the spatial area covered by the radar signal. The trained machine learning model is used to detect whether a living organism exists in the spatial area covered by the radar.

9. The apparatus according to claim 7, wherein, The output of the machine learning model is the probability of the existence of a living organism in one or more of the grid spaces. The trained machine learning model is used to detect whether a living organism exists in the one or more grid spaces.

10. The apparatus according to claim 7, wherein, The output of the machine learning model is the probability of an organism falling down in one or more grid spaces. The trained machine learning model is used to detect whether the organism has fallen down in the one or more grid spaces.

Images