In-vehicle personnel category identification method and device, electronic equipment and storage medium

Abstract

The invention discloses an in-vehicle personnel category identification method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring multiple groups of in-vehicle radar echo features; respectively inputting the multiple groups of in-vehicle radar echo features into a trained personnel category recognition model to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features; and performing statistics based on the first personnel category recognition result corresponding to each group of in-vehicle radar echo features to obtain a personnel category number, and determining a second personnel category recognition result based on the personnel category number and a category judgment threshold. According to the technical scheme, through the personnel category recognition model and the personnel category recognition method for threshold judgment, coarse and fine cascade recognition of personnel categories is realized, and the accuracy of a personnel category recognition result is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of smart cabin technology, and in particular to a method, device, electronic equipment and storage medium for identifying the category of people in a vehicle. Background Art

[0002] With the intelligent development of the automobile industry, the construction of smart cabins is an inevitable trend. In addition to completing various intelligent interactions, automobile safety is also an important part of the smart cabin.

[0003] Among them, Child Presence Detection (CPD) is an indicator that has received much attention in recent years for automobile safety. It is used to detect and remind drivers whether there are children in the vehicle to prevent accidental locking of the vehicle or children being forgotten in the vehicle.

[0004] In the process of implementing the present invention, it is found that there are at least the following technical problems in the prior art: the existing in-vehicle person category recognition solution has the problem of low accuracy of person category recognition results. Summary of the invention

[0005] The present invention provides a method, a device, an electronic device and a storage medium for identifying the category of a person in a vehicle, so as to improve the accuracy of the result of identifying the category of a person in the vehicle.

[0006] According to one aspect of the present invention, a method for identifying the category of a person in a vehicle is provided, comprising:

[0007] Obtain multiple sets of in-vehicle radar echo features;

[0008] Inputting the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively, to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features;

[0009] The number of personnel categories is obtained based on the statistics of the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and the second personnel category recognition result is determined based on the number of personnel categories and a category judgment threshold.

[0010] According to another aspect of the present invention, there is provided a device for identifying the type of a person in a vehicle, comprising:

[0011] An in-vehicle radar echo feature acquisition module, used to acquire multiple sets of in-vehicle radar echo features;

[0012] A first person category recognition result prediction module, used to input the multiple groups of in-vehicle radar echo features into the trained person category recognition model respectively, to obtain a first person category recognition result corresponding to each group of in-vehicle radar echo features;

[0013] The second personnel category recognition result determination module is used to obtain the number of personnel categories based on the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and determine the second personnel category recognition result based on the number of personnel categories and the category judgment threshold.

[0014] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0015] at least one processor;

[0016] and a memory communicatively coupled to the at least one processor;

[0017] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying the category of people in a vehicle as described in any embodiment of the present invention.

[0018] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying the category of a person in a vehicle as described in any embodiment of the present invention when executed.

[0019] The technical solution of the embodiment of the present invention obtains multiple sets of in-vehicle radar echo features, and then inputs the multiple sets of in-vehicle radar echo features into the trained personnel category recognition model to obtain the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then obtains the number of personnel categories based on the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then determines the second personnel category recognition result based on the number of personnel categories and the category judgment threshold. The above technical solution realizes the coarse and fine cascade recognition of personnel categories through the personnel category recognition model and the personnel category recognition method of threshold judgment, and effectively improves the accuracy of the personnel category recognition result.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1is a flow chart of a method for identifying the category of a person in a vehicle provided according to the first embodiment of the present invention;

[0023] Figure 2 is a flow chart of a method for identifying the category of people in a vehicle provided according to Embodiment 2 of the present invention;

[0024] Figure 3 is a flow chart of a method for identifying the category of a person in a vehicle provided according to Embodiment 3 of the present invention;

[0025] Figure 4 is a flow chart of a method for identifying the category of a person in a vehicle provided according to a fourth embodiment of the present invention;

[0026] Figure 5 is a flow chart of a method for identifying the category of a person in a vehicle provided according to an embodiment of the present invention;

[0027] Figure 6 is a schematic diagram of the structure of a device for identifying the type of people in a vehicle provided according to a fifth embodiment of the present invention;

[0028] Figure 7 It is a schematic diagram of the structure of an electronic device for implementing the method for identifying the category of a person in a vehicle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0031] Embodiment 1

[0032] Figure 1 This is a flow chart of a method for identifying the type of people in a vehicle provided in the first embodiment of the present invention. This embodiment can be applied to the case where a millimeter-wave radar is used to detect life in a locked vehicle cabin. The method can be performed by a device for identifying the type of people in a vehicle. The device can be implemented in the form of hardware and / or software. The device can be configured in electronic devices such as vehicle-mounted terminals. Figure 1 As shown, the method includes:

[0033] S110: Acquire multiple sets of in-vehicle radar echo features.

[0034] In the embodiment of the present invention, the in-vehicle radar echo feature refers to the feature extracted from the radar echo data collected in the vehicle, and each set of in-vehicle radar echo features may include but is not limited to the energy feature, volume feature, and duration feature of the in-vehicle target. Each set of in-vehicle radar echo features may correspond to one frame of in-vehicle radar echo data.

[0035] Exemplarily, a millimeter-wave radar can be used to obtain multiple frames of in-vehicle radar echo data, and then generate point cloud data corresponding to each frame of in-vehicle radar echo data, and then perform feature extraction on the point cloud data corresponding to each frame of in-vehicle radar echo data to obtain multiple groups of in-vehicle radar echo features.

[0036] It should be noted that compared with in-vehicle personnel recognition methods based on infrared sensors or visual sensors, the embodiments of the present invention use millimeter-wave radar as a sensor, which is not affected by ambient light, dust, smoke and temperature, and effectively improves the reliability and accuracy of personnel category recognition results.

[0037] S120: Input the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features.

[0038] In the embodiment of the present invention, the person category recognition model refers to a pre-trained neural network model, which can be used to roughly predict the category of the person in the car. The person category recognition result can be an adult category, a child category, or an empty car category, etc., which is not specifically limited here.

[0039] Exemplarily, a trained personnel category recognition model can be used to predict the category of people in the vehicle based on multiple groups of in-vehicle radar echo features, thereby obtaining a first personnel category recognition result corresponding to each group of in-vehicle radar echo features. The first personnel category recognition results corresponding to each group can be the same or different.

[0040] S130. Obtain the number of personnel categories based on the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and determine the second personnel category recognition result based on the number of personnel categories and a category judgment threshold.

[0041] In the embodiment of the present invention, the number of personnel categories refers to the result of counting the number of personnel category recognition results corresponding to multiple sets of features, and the number of personnel categories may include the number of adult categories, the number of children categories, and the number of empty vehicle categories. The category judgment threshold is used to finely judge the personnel category, and the number of the category judgment threshold may be multiple.

[0042] Exemplarily, the number of personnel categories counted may include: the number of adult categories may be 20, the number of child categories may be 0, and the number of empty vehicle categories may be 0; if the number of adult categories is greater than the category judgment threshold, the second personnel category identification result is the adult category.

[0043] The technical solution of the embodiment of the present invention obtains multiple sets of in-vehicle radar echo features, and then inputs the multiple sets of in-vehicle radar echo features into the trained personnel category recognition model to obtain the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then obtains the number of personnel categories based on the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then determines the second personnel category recognition result based on the number of personnel categories and the category judgment threshold. The above technical solution effectively improves the accuracy of the personnel category recognition result through the multi-level personnel category recognition method of the personnel category recognition model and the threshold judgment.

[0044] Embodiment 2

[0045] Figure 2 The flowchart of a method for identifying the category of a person in a vehicle provided in the second embodiment of the present invention is provided. The method of this embodiment can be combined with each optional solution in the method for identifying the category of a person in a vehicle provided in the above embodiment. The method for identifying the category of a person in a vehicle provided in this embodiment is further optimized. Optionally, obtaining multiple sets of in-vehicle radar echo features includes: obtaining multiple frames of in-vehicle radar echo data; obtaining time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data based on coherent time accumulation parameters; performing distance-Doppler estimation on the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data to obtain distance-Doppler spectrum data corresponding to each frame of in-vehicle radar echo data; performing target detection on the distance-Doppler spectrum data corresponding to each frame of in-vehicle radar echo data to obtain the distance-Doppler information of the target radar echo; performing angle estimation on the distance-Doppler information of the target radar echo to obtain point cloud data corresponding to each frame of in-vehicle radar echo data; performing feature extraction on the point cloud data corresponding to each frame of in-vehicle radar echo data to obtain multiple sets of in-vehicle radar echo features.

[0046] like Figure 2 As shown, the method includes:

[0047] S210: Acquire multiple frames of in-vehicle radar echo data.

[0048] Among them, the in-vehicle radar echo data refers to the radar echo data inside the vehicle cabin collected by millimeter-wave radar.

[0049] Exemplarily, the in-vehicle radar echo data can be expressed as x(n,m,k), where n=1,2,…,N, n represents the nth sampling point in the fast time dimension, and N represents the number of distance units in a chirp sequence; m=1,2,…,M, m represents the mth linear frequency modulated continuous wave signal in the slow time dimension, and M represents the number of chirps in a frame; k is the antenna dimension, representing the received signal of the kth antenna channel.

[0050] S220. Based on the coherent time accumulation parameter, obtain the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data, perform range-Doppler estimation on the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data, and obtain the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data.

[0051] Wherein, time accumulation refers to storing multiple frames of radar echo range spectrum data according to coherent time accumulation parameters. The coherent time accumulation parameters may be dynamically adjusted parameters, and may be used to slide and store multiple frames of radar echo range spectrum data.

[0052] Specifically, the distance of each frame of in-vehicle radar echo data is estimated to obtain the radar echo distance data corresponding to each frame of in-vehicle radar echo data, and then the clutter suppression is performed on the radar echo distance data corresponding to each frame of in-vehicle radar echo data to obtain the radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data; and then based on the coherent time accumulation parameter, the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data is obtained, and the distance-Doppler estimation is performed on the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data to obtain the distance-Doppler spectrum data corresponding to each frame of in-vehicle radar echo data.

[0053] Exemplarily, the distance estimation step includes:

[0054] The fast time dimension data frame of each frame of the in-vehicle radar echo data is windowed and filtered, and the fast Fourier transform (FFT) of the fast time dimension is performed to obtain the radar echo distance data y(l, m, k) corresponding to each frame of the in-vehicle radar echo data. The calculation process can be expressed as:

[0055]

[0056] Wherein, w(n) represents the window function, N represents the number of points of the fast time dimension FFT, and l represents the range unit index number.

[0057] Furthermore, clutter suppression methods such as DC removal and sliding average are used to suppress clutter on the radar echo distance data y(l,m,k) corresponding to each frame of in-vehicle radar echo data, and the radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data is obtained. Understandably, Most of the static clutter is removed, and it can be used to extract Doppler information of micro-motion vital signs such as breathing in the cabin.

[0058] Furthermore, in order to ensure sufficient resolution, time accumulation of multiple frames of in-vehicle radar echo data is required. Based on the coherent time accumulation parameters, multiple frames of radar echo range spectrum data are stored in a sliding manner. For example, the range-Doppler estimation of the nFth frame data is currently being performed, and the time-accumulated radar echo range spectrum data is:

[0059]

[0060] Among them, α represents the coherent time accumulation parameter. Further, based on the time-accumulated radar echo range spectrum data For range-Doppler estimation, since the above time accumulation method is a non-uniform structure along the slow time dimension, sparse recovery methods such as non-uniform Fast Fourier Transform (NUFFT) or orthogonal matching pursuit (OMP) can be used for range-Doppler estimation to obtain range-Doppler spectrum data. f d Represents the Doppler shift frequency.

[0061] S230: Perform target detection on the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data to obtain range-Doppler information of the target radar echo.

[0062] Among them, human targets in the vehicle radar echo data present the characteristics of multiple scattering points. Human targets will occupy multiple range units and Doppler units. Constant False-Alarm Rate (CFAR) and other target detection methods can be used to detect the range-Doppler spectrum data. Perform target detection and obtain the distance-Doppler information of the target radar echo Where i represents the i-th range-Doppler target point detected.

[0063] S240 , performing angle estimation on the range-Doppler information of the target radar echo to obtain point cloud data corresponding to each frame of in-vehicle radar echo data.

[0064] For example, based on the detected range-Doppler target point along the antenna dimension, the range-Doppler information of the target radar echo is estimated to obtain the point cloud data in, They respectively represent the azimuth and elevation angles of the i-th range-Doppler target point. The angle estimation may adopt an angle estimation method such as FFT or DBF (Directional Beamforming), which is not specifically limited here.

[0065] S250, performing feature extraction on the point cloud data corresponding to each frame of the in-vehicle radar echo data to obtain multiple groups of in-vehicle radar echo features.

[0066] Specifically, the point cloud data corresponding to each frame of in-vehicle radar echo data is clustered to exclude some noise points in the point cloud data, and then feature extraction is performed on the clustered point cloud data to extract target features to provide feature information for subsequent multi-personnel category recognition.

[0067] In some optional embodiments, feature extraction is performed on point cloud data corresponding to each frame of in-vehicle radar echo data to obtain multiple groups of in-vehicle radar echo features, including: target energy feature extraction is performed on point cloud data corresponding to each frame of in-vehicle radar echo data to obtain target energy features corresponding to each frame of in-vehicle radar echo data; target volume feature extraction is performed on point cloud data corresponding to each frame of in-vehicle radar echo data to obtain target volume features corresponding to each frame of in-vehicle radar echo data; target duration feature extraction is performed on point cloud data corresponding to each frame of in-vehicle radar echo data to obtain target duration features corresponding to each frame of in-vehicle radar echo data; multiple groups of in-vehicle radar echo features are generated based on target energy features corresponding to each frame of in-vehicle radar echo data, target volume features corresponding to each frame of in-vehicle radar echo data, and target duration features corresponding to each frame of in-vehicle radar echo data.

[0068] The target energy feature may be the distance-Doppler spectrum energy or angle spectrum energy of the target point, etc. The target volume feature may be the radius contained in the boundary of the target point cloud, and the target duration feature may be the number of times the target appears continuously in this area.

[0069] S260: Input the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features.

[0070] S270. Obtain the number of personnel categories based on statistics of the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and determine the second personnel category recognition result based on the number of personnel categories and a category judgment threshold.

[0071] Optionally, based on the coherent time accumulation parameters, the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data is obtained, including: obtaining historical personnel category recognition results; determining the coherent time accumulation parameters corresponding to the historical personnel category recognition results; based on the coherent time accumulation parameters corresponding to the historical personnel category recognition results, obtaining the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data.

[0072] Among them, the historical personnel category recognition result refers to the personnel category recognition result predicted based on the historical in-vehicle radar echo characteristics. For example, the historical in-vehicle radar echo characteristics may be the in-vehicle radar echo characteristics before the current group of in-vehicle radar echo characteristics.

[0073] It should be noted that by dynamically adjusting the coherent time accumulation parameters based on historical personnel category recognition results, rapid recognition of adult categories and signal enhancement of child categories are achieved, effectively improving the accuracy of personnel category recognition results.

[0074] Specifically, determining the coherent time accumulation parameter corresponding to the historical personnel category recognition result includes: when the historical personnel category recognition result is the adult category, obtaining the first coherent time accumulation parameter; when the historical personnel category recognition result is the child category, obtaining the second coherent time accumulation parameter; when the historical personnel category recognition result is the empty vehicle category, obtaining the third coherent time accumulation parameter; wherein the first coherent time accumulation parameter is less than the second coherent time accumulation parameter, and the second coherent time accumulation parameter is less than the third coherent time accumulation parameter.

[0075] For example, setting α A For the first coherent time accumulation parameter corresponding to the adult category, set α B For the second coherent time accumulation parameter corresponding to the child category, set α E is the third coherent time accumulation parameter corresponding to the empty car category, α A <α B <α E It can be understood that the coherent time accumulation parameter is the smallest under the adult category, so that the effect of rapid adult category recognition can be achieved. That is, if the historical person category recognition result is the adult category, the coherent time accumulation parameter is dynamically adjusted to α A ; In the child category, the echo signal is weak, so the coherent time accumulation parameter is dynamically adjusted to α B, to achieve the effect of dynamic enhancement of weak signals, quickly and stably identify the child category; in the empty car category, the coherent time accumulation parameter is dynamically adjusted to α E , so as to accurately judge the unmanned status of the cabin.

[0076] The technical solution of the embodiment of the present invention determines the coherent time accumulation parameters corresponding to the historical personnel category recognition results, and then obtains the time-accumulated radar echo distance spectrum data corresponding to each frame of the in-vehicle radar echo data based on the coherent time accumulation parameters corresponding to the historical personnel category recognition results. The above technical solution dynamically adjusts the coherent time accumulation parameters based on the historical personnel category recognition results, realizes the rapid recognition of the adult category and the signal enhancement of the child category, and effectively improves the accuracy of the personnel category recognition results.

[0077] Embodiment 3

[0078] Figure 3 A flowchart of a method for identifying the category of people in a vehicle provided in the third embodiment of the present invention, the method of this embodiment can be combined with the various optional schemes in the method for identifying the category of people in a vehicle provided in the above embodiments. The method for identifying the category of people in a vehicle provided in this embodiment is further optimized. Optionally, the determining of the second person category identification result based on the number of person categories and the category judgment threshold comprises: obtaining historical person category identification results; determining the category judgment threshold corresponding to the historical person category identification results, and determining the second person category identification result based on the number of person categories and the category judgment threshold corresponding to the historical person category identification results.

[0079] like Figure 3 As shown, the method includes:

[0080] S310: Acquire multiple sets of in-vehicle radar echo features.

[0081] S320: Input the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features.

[0082] S330. Obtain the number of personnel categories based on statistics of the first personnel category recognition results corresponding to each group of in-vehicle radar echo features.

[0083] S340, obtaining historical personnel category recognition results; determining a category judgment threshold corresponding to the historical personnel category recognition results, and determining a second personnel category recognition result based on the number of personnel categories and the category judgment threshold corresponding to the historical personnel category recognition results.

[0084] The embodiment of the present invention adjusts the category judgment threshold by using historical personnel category recognition results, thereby achieving dynamic adjustment of the category judgment threshold, thereby making the second personnel category recognition result obtained by judgment more accurate, and effectively improving the accuracy of the personnel category recognition result.

[0085] Optionally, the first person category recognition result corresponding to each group of in-vehicle radar echo features is an adult category, a child category or an empty vehicle category; the number of person categories includes the number of adult categories, the number of child categories and the number of empty vehicle categories; the category judgment threshold includes the adult category judgment threshold, the child category quantity judgment threshold and the empty vehicle category judgment threshold; accordingly, determining the category judgment threshold corresponding to the historical person category recognition result, and determining the second person category recognition result based on the number of person categories and the category judgment threshold corresponding to the historical person category recognition result, including: when the historical person category recognition result is an adult category, obtaining the first adult category judgment threshold, the first child category quantity judgment threshold and the first empty vehicle category judgment threshold; based on the number of adult categories, the number of child categories, the number of empty vehicles, the first adult category judgment threshold, the first child category quantity judgment threshold value and the first empty car category judgment threshold to determine the second person category recognition result; in the case where the historical person category recognition result is the child category, obtain the second adult category judgment threshold, the second child category quantity judgment threshold and the second empty car category judgment threshold; determine the second person category recognition result based on the adult category quantity, the child category quantity, the empty car category quantity, the second adult category judgment threshold, the second child category quantity judgment threshold and the second empty car category judgment threshold; in the case where the historical person category recognition result is the empty car category, obtain the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty car category judgment threshold; determine the second person category recognition result based on the adult category quantity, the child category quantity, the empty car category quantity, the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty car category judgment threshold.

[0086] Exemplarily, when the historical person category recognition result is an adult category, the first adult category judgment threshold, the first child category quantity judgment threshold, and the first empty vehicle category judgment threshold are obtained; if the number of adult categories is greater than the first adult category judgment threshold, the second person category recognition result is determined as an adult category. When the historical person category recognition result is a child category, the second adult category judgment threshold, the second child category quantity judgment threshold, and the second empty vehicle category judgment threshold are obtained; if the number of empty vehicles is greater than the second empty vehicle category judgment threshold, the second person category recognition result is determined as an empty vehicle category.

[0087] It should be noted that the principle of adjusting the category judgment threshold can be: the continuation of the same personnel category state can appropriately relax the threshold requirement, and the conversion of different personnel categories should appropriately increase the threshold requirement. For example, when the historical personnel category recognition result is the adult category, the current adult category judgment threshold is TH_AD_L, that is, if Num_AD>TH_AD_L, the second personnel category recognition result is determined as the adult category; when the historical personnel category recognition result is the child or empty vehicle category, the adult category judgment threshold converted to the adult category is TH_AD_H, TH_AD_H>TH_AD_L, that is, if Num_AD>TH_AD_H, the second personnel category recognition result is determined as the adult category, where Num_AD represents the number of adult categories.

[0088] The technical solution of the embodiment of the present invention adjusts the category judgment threshold by using historical personnel category recognition results, thereby realizing dynamic adjustment of the category judgment threshold, thereby making the second personnel category recognition result obtained by judgment more accurate, and effectively improving the accuracy of the personnel category recognition result.

[0089] Embodiment 4

[0090] Figure 4 A flowchart of a method for identifying the category of people in a vehicle provided in the fourth embodiment of the present invention, the method of this embodiment can be combined with the various optional schemes in the method for identifying the category of people in a vehicle provided in the above embodiments. The method for identifying the category of people in a vehicle provided in this embodiment is further optimized. Optionally, after the multiple groups of in-vehicle radar echo features are respectively input into the trained personnel category recognition model to obtain the first personnel category recognition result corresponding to each group of in-vehicle radar echo features, it also includes: obtaining historical personnel category recognition results; determining the non-coherent time accumulation parameters corresponding to the historical personnel category recognition results; obtaining accumulated point cloud data based on the non-coherent time accumulation parameters; judging whether there is a target in the vehicle based on the accumulated point cloud data; and correcting the first personnel category recognition result based on the in-vehicle target judgment result.

[0091] like Figure 4 As shown, the method includes:

[0092] S410: Acquire multiple sets of in-vehicle radar echo features.

[0093] S420: Input the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features.

[0094] S430, obtaining historical personnel category recognition results; determining the non-coherent time accumulation parameters corresponding to the historical personnel category recognition results; obtaining accumulated point cloud data based on the non-coherent time accumulation parameters; judging whether there is a target in the vehicle based on the accumulated point cloud data; and correcting the first personnel category recognition result based on the in-vehicle target judgment result.

[0095] The target refers to a person or other target with life characteristics. The non-coherent time accumulation parameter is used to enhance the radar echo characteristics inside the vehicle, that is, the multi-frame point cloud data can be accumulated according to the non-coherent time accumulation parameter, and then the accumulated point cloud data can be clustered to determine whether there are life characteristics of a person in the vehicle, and the first person category recognition result can be corrected. For example, if the first person category recognition result is an empty vehicle category and the target judgment result inside the vehicle is a child category, the first person category recognition result is corrected to a child category.

[0096] Specifically, when the historical personnel category recognition result is the adult category, obtain the first non-coherent time accumulation parameter; obtain the first accumulated point cloud data based on the first non-coherent time accumulation parameter; judge whether there is a target in the car based on the first accumulated point cloud data; when the historical personnel category recognition result is the child category, obtain the second non-coherent time accumulation parameter; obtain the second accumulated point cloud data based on the second non-coherent time accumulation parameter; judge whether there is a target in the car based on the second accumulated point cloud data; when the historical personnel category recognition result is the empty car category, obtain the third non-coherent time accumulation parameter; obtain the third accumulated point cloud data based on the third non-coherent time accumulation parameter; judge whether there is a target in the car based on the third accumulated point cloud data; correct the first personnel category recognition result based on the target judgment result in the car.

[0097] It is understandable that when children and infants, especially infants, accidentally slip into severely obstructed locations in the cabin such as footwells, the signal-to-noise ratio of the radar echo signal is very low. Therefore, when the first person category recognition result is an empty car category and a child category, it is necessary to further determine whether there are life characteristics of people in the car to improve the accuracy of the detection results.

[0098] For example, the parameter β is set A , β B , β E Represent the non-coherent time accumulation parameters for adult category, child category, and empty car category, respectively, and satisfy β A <β B <β E The signal-to-noise ratio of targets in the adult category is relatively high, and the non-coherent time accumulation parameter β = β can be dynamically adjusted. A =1; in the children category, the non-coherent time accumulation parameter β can be dynamically adjusted = β B; In the empty car category, the non-coherent time accumulation parameter β can be adjusted E , to accurately judge the unmanned status of the cabin.

[0099] S440. Obtain the number of personnel categories based on the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and determine the second personnel category recognition result based on the number of personnel categories and a category judgment threshold.

[0100] The technical solution of the embodiment of the present invention dynamically adjusts the non-coherent time accumulation parameters through historical personnel category recognition results, thereby achieving signal enhancement of the child category and the empty vehicle category, and effectively improving the accuracy of the personnel category recognition results.

[0101] Figure 5 It is a flow chart of a method for identifying the category of people in a car according to an embodiment of the present invention. Specifically, the in-car radar echo data is subjected to distance estimation, clutter suppression, distance-Doppler estimation of coherent time accumulation parameters, target detection, angle estimation and feature extraction to obtain the in-car radar echo features; then the in-car radar echo features are input into the personnel category recognition model based on the long short-term memory network (LSTM, Long Short-Term Memory) to obtain the first personnel category recognition result, and then based on the accumulated point cloud data of the non-coherent time accumulation parameters, it is judged whether there is life in the car, the correction of the first personnel category recognition result is realized, and the corrected first personnel category recognition result is output. A state window of a preset time range is set, the number of personnel categories in the state window is counted, and then a dynamic multi-threshold judgment is performed according to the number of personnel categories and the category judgment threshold to obtain the second personnel category recognition result; when the second personnel category recognition result is a child category, an alarm is issued. Among them, the coherent time accumulation parameter, the non-coherent time accumulation parameter and the category judgment threshold can be dynamically adjusted according to the historical personnel category recognition result (i.e., the second personnel category recognition result corresponding to the previous in-car radar echo data).

[0102] The above technical solution effectively improves the accuracy of personnel category recognition results through a multi-level personnel category recognition method based on a personnel category recognition model and threshold judgment and a parameter adjustment strategy.

[0103] Embodiment 5

[0104] Figure 6 This is a schematic diagram of the structure of a device for identifying the type of people in a vehicle provided by Embodiment 5 of the present invention. Figure 6 As shown, the device comprises:

[0105] An in-vehicle radar echo feature acquisition module 510 is used to acquire multiple sets of in-vehicle radar echo features;

[0106] A first person category recognition result prediction module 520 is used to input the multiple groups of in-vehicle radar echo features into the trained person category recognition model respectively to obtain a first person category recognition result corresponding to each group of in-vehicle radar echo features;

[0107] The second person category identification result determination module 530 is used to obtain the number of person categories based on the first person category identification result corresponding to each group of in-vehicle radar echo features, and determine the second person category identification result based on the number of person categories and the category judgment threshold.

[0108] The technical solution of the embodiment of the present invention obtains multiple sets of in-vehicle radar echo features, and then inputs the multiple sets of in-vehicle radar echo features into the trained personnel category recognition model to obtain the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then obtains the number of personnel categories based on the first personnel category recognition result corresponding to each set of in-vehicle radar echo features, and then determines the second personnel category recognition result based on the number of personnel categories and the category judgment threshold. The above technical solution effectively improves the accuracy of the personnel category recognition result through the multi-level personnel category recognition method of the personnel category recognition model and the threshold judgment.

[0109] In some optional implementations, the in-vehicle radar echo feature acquisition module 510 includes:

[0110] A multi-frame in-vehicle radar echo data unit, used to obtain multi-frame in-vehicle radar echo data;

[0111] A time-accumulated radar echo range spectrum data acquisition unit, used to acquire the time-accumulated radar echo range spectrum data corresponding to each frame of in-vehicle radar echo data based on a coherent time accumulation parameter;

[0112] a range-Doppler estimation unit, configured to perform range-Doppler estimation on the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data, to obtain the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data;

[0113] A target detection unit, used to perform target detection on the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data, to obtain the range-Doppler information of the target radar echo;

[0114] An angle estimation unit, used to perform angle estimation on the range-Doppler information of the target radar echo to obtain point cloud data corresponding to each frame of in-vehicle radar echo data;

[0115] The feature extraction unit is used to extract features from the point cloud data corresponding to each frame of the in-vehicle radar echo data to obtain multiple groups of in-vehicle radar echo features.

[0116] In some optional implementations, the time-accumulated radar echo range spectrum data acquisition unit includes:

[0117] A historical personnel category recognition result acquisition subunit is used to acquire historical personnel category recognition results;

[0118] A coherent time accumulation parameter dynamic adjustment subunit, used to determine the coherent time accumulation parameter corresponding to the historical personnel category recognition result;

[0119] The time-accumulated radar echo distance spectrum acquisition subunit is used to acquire the time-accumulated radar echo distance spectrum data corresponding to each frame of the in-vehicle radar echo data based on the coherent time accumulation parameters corresponding to the historical personnel category recognition results.

[0120] In some optional implementations, the coherent time accumulation parameter dynamic adjustment subunit may be specifically used to:

[0121] When the historical person category recognition result is an adult category, obtaining a first coherent time accumulation parameter;

[0122] When the historical person category recognition result is a child category, obtaining a second coherent time accumulation parameter;

[0123] When the historical personnel category recognition result is an empty vehicle category, obtaining a third coherent time accumulation parameter;

[0124] The first coherent time accumulation parameter is smaller than the second coherent time accumulation parameter, and the second coherent time accumulation parameter is smaller than the third coherent time accumulation parameter.

[0125] In some optional implementations, the second person category identification result determination module 530 includes:

[0126] A historical personnel category recognition result acquisition unit, used to acquire historical personnel category recognition results;

[0127] The dynamic category judgment threshold personnel category identification unit is used to determine the category judgment threshold corresponding to the historical personnel category identification result, and determine the second personnel category identification result based on the number of personnel categories and the category judgment threshold corresponding to the historical personnel category identification result.

[0128] In some optional implementations, the first person category recognition result corresponding to each group of in-vehicle radar echo features is an adult category, a child category or an empty vehicle category; the number of person categories includes the number of adult categories, the number of child categories and the number of empty vehicle categories; the category judgment threshold includes an adult category judgment threshold, a child category number judgment threshold and an empty vehicle category judgment threshold;

[0129] Accordingly, the dynamic category judgment threshold personnel category recognition unit can be specifically used for:

[0130] In the case where the historical person category recognition result is the adult category, obtaining a first adult category judgment threshold, a first child category quantity judgment threshold, and a first empty vehicle category judgment threshold; determining a second person category recognition result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the first adult category judgment threshold, the first child category quantity judgment threshold, and the first empty vehicle category judgment threshold;

[0131] In the case where the historical person category recognition result is the child category, obtaining a second adult category judgment threshold, a second child category quantity judgment threshold, and a second empty vehicle category judgment threshold; determining a second person category recognition result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the second adult category judgment threshold, the second child category quantity judgment threshold, and the second empty vehicle category judgment threshold;

[0132] When the historical personnel category identification result is an empty vehicle category, obtain the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty vehicle category judgment threshold; determine the second personnel category identification result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty vehicle category judgment threshold.

[0133] In some optional implementations, the device for identifying the type of person in a vehicle further includes:

[0134] The first personnel category recognition result correction unit is used to obtain historical personnel category recognition results; determine the non-coherent time accumulation parameters corresponding to the historical personnel category recognition results; obtain accumulated point cloud data based on the non-coherent time accumulation parameters; judge whether there is a target in the vehicle based on the accumulated point cloud data; and correct the first personnel category recognition result based on the target judgment result in the vehicle.

[0135] The device for identifying the type of a person in a vehicle provided by an embodiment of the present invention can execute the method for identifying the type of a person in a vehicle provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0136] Embodiment 6

[0137] Figure 7A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0138] like Figure 7 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The I / O interface 15 is also connected to the bus 14.

[0139] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0140] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a method for identifying the category of a person in a vehicle, which includes:

[0141] Obtain multiple sets of in-vehicle radar echo features;

[0142] Inputting the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively, to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features;

[0143] The number of personnel categories is obtained based on the statistics of the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and the second personnel category recognition result is determined based on the number of personnel categories and a category judgment threshold.

[0144] In some embodiments, the method for identifying the category of a person in the vehicle can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for identifying the category of a person in the vehicle described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the method for identifying the category of a person in the vehicle by any other appropriate means (for example, by means of firmware).

[0145] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0146] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0147] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0148] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0149] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0150] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0151] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0152] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying the type of people in a vehicle, characterized in that: include: Obtain multiple sets of in-vehicle radar echo features; Inputting the multiple groups of in-vehicle radar echo features into the trained personnel category recognition model respectively, to obtain a first personnel category recognition result corresponding to each group of in-vehicle radar echo features; The number of personnel categories is obtained based on the statistics of the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and the second personnel category recognition result is determined based on the number of personnel categories and a category judgment threshold.

2. The method according to claim 1, characterized in that The obtaining of multiple sets of in-vehicle radar echo features includes: Obtain multiple frames of in-vehicle radar echo data; Based on the coherent time accumulation parameters, the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data is obtained; Performing range-Doppler estimation on the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data to obtain the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data; Performing target detection on the range-Doppler spectrum data corresponding to each frame of the in-vehicle radar echo data to obtain the range-Doppler information of the target radar echo; Performing angle estimation on the range-Doppler information of the target radar echo to obtain point cloud data corresponding to each frame of in-vehicle radar echo data; Feature extraction is performed on the point cloud data corresponding to each frame of the in-vehicle radar echo data to obtain multiple groups of in-vehicle radar echo features.

3. The method according to claim 2, characterized in that The method of obtaining the time-accumulated radar echo range spectrum data corresponding to each frame of the in-vehicle radar echo data based on the coherent time accumulation parameter includes: Obtain historical personnel category recognition results; Determine the coherent time accumulation parameter corresponding to the historical personnel category recognition result; Based on the coherent time accumulation parameters corresponding to the historical personnel category recognition results, the time-accumulated radar echo distance spectrum data corresponding to each frame of in-vehicle radar echo data is obtained.

4. The method according to claim 3, characterized in that The determining of the coherent time accumulation parameter corresponding to the historical personnel category recognition result includes: When the historical person category recognition result is an adult category, obtaining a first coherent time accumulation parameter; When the historical person category recognition result is a child category, obtaining a second coherent time accumulation parameter; When the historical personnel category recognition result is an empty vehicle category, obtaining a third coherent time accumulation parameter; The first coherent time accumulation parameter is smaller than the second coherent time accumulation parameter, and the second coherent time accumulation parameter is smaller than the third coherent time accumulation parameter.

5. The method according to claim 1, characterized in that The determining of the second person category recognition result based on the number of person categories and the category judgment threshold comprises: Obtain historical personnel category recognition results; A category judgment threshold corresponding to the historical personnel category recognition result is determined, and a second personnel category recognition result is determined based on the number of personnel categories and the category judgment threshold corresponding to the historical personnel category recognition result.

6. The method according to claim 5, characterized in that The first person category recognition result corresponding to each group of vehicle interior radar echo features is an adult category, a child category or an empty vehicle category; the number of person categories includes the number of adult categories, the number of child categories and the number of empty vehicle categories; the category judgment threshold includes the adult category judgment threshold, the child category number judgment threshold and the empty vehicle category judgment threshold; Accordingly, determining a category judgment threshold corresponding to the historical personnel category recognition result, and determining a second personnel category recognition result based on the number of personnel categories and the category judgment threshold corresponding to the historical personnel category recognition result, includes: In the case where the historical person category recognition result is the adult category, obtaining a first adult category judgment threshold, a first child category quantity judgment threshold, and a first empty vehicle category judgment threshold; determining a second person category recognition result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the first adult category judgment threshold, the first child category quantity judgment threshold, and the first empty vehicle category judgment threshold; In the case where the historical person category recognition result is the child category, obtaining a second adult category judgment threshold, a second child category quantity judgment threshold, and a second empty vehicle category judgment threshold; determining a second person category recognition result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the second adult category judgment threshold, the second child category quantity judgment threshold, and the second empty vehicle category judgment threshold; When the historical personnel category identification result is an empty vehicle category, obtain the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty vehicle category judgment threshold; determine the second personnel category identification result based on the adult category quantity, the child category quantity, the empty vehicle category quantity, the third adult category judgment threshold, the third child category quantity judgment threshold and the third empty vehicle category judgment threshold.

7. The method according to any one of claims 1 to 6, characterized in that: After the plurality of groups of in-vehicle radar echo features are respectively input into the trained personnel category recognition model to obtain the first personnel category recognition result corresponding to each group of in-vehicle radar echo features, the method further includes: Obtain historical personnel category recognition results; Determine the non-coherent time accumulation parameter corresponding to the historical personnel category recognition result; Acquire accumulated point cloud data based on the non-coherent time accumulation parameter; and determine whether there is a target in the vehicle based on the accumulated point cloud data; The first person category recognition result is corrected based on the in-vehicle target judgment result.

8. A device for identifying the type of people in a vehicle, characterized in that: include: An in-vehicle radar echo feature acquisition module, used to acquire multiple sets of in-vehicle radar echo features; The first person category recognition result prediction module is used to input the multiple groups of in-vehicle radar echo features into the trained person category recognition model respectively, and obtain the corresponding person category of each group of in-vehicle radar echo features. First person category identification result; The second personnel category recognition result determination module is used to obtain the number of personnel categories based on the first personnel category recognition results corresponding to each group of in-vehicle radar echo features, and determine the second personnel category recognition result based on the number of personnel categories and the category judgment threshold.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying the category of people in a vehicle according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying the category of a person in a vehicle according to any one of claims 1 to 7 when executed.

Images