Vehicle early warning method, electronic equipment, vehicle, storage medium and computer product

Abstract

The invention discloses a vehicle early warning method, electronic equipment, a vehicle, a storage medium and a computer product, and relates to the technical field of vehicles, and the vehicle early warning method specifically comprises the steps that a wireless signal transmitting module is controlled to transmit a plurality of wireless detection signals to the periphery of the vehicle; generating a target living body image based on the plurality of wireless detection signals, and extracting image features contained in the target living body image; and determining a target early warning strategy corresponding to the risk personnel under the condition of identifying that the risk personnel exist in the target living body image according to the image features, and executing an early warning operation according to the target early warning strategy. According to the invention, the technical effect of improving the recognition accuracy of the risk personnel is achieved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a vehicle early warning method, electronic equipment, vehicle, storage medium and computer program product. Background Art

[0002] With the continuous development of the automobile industry, new energy vehicles have become the preferred means of transportation for more and more users. In order to improve the safety of new energy vehicles, technicians usually configure multiple fisheye cameras on the vehicle to better identify risky people around the vehicle.

[0003] In the related art, after the vehicle enters the locked state, the vehicle usually controls the fisheye camera to capture the vehicle's surroundings to obtain image data, and identifies whether there are any risky persons around the vehicle based on the image data.

[0004] However, since fisheye cameras are easily affected by ambient light during the process of collecting image data, the quality of the captured image data is low. This can easily lead to a low accuracy rate in the vehicle's recognition of risky persons, greatly reducing the safety of the vehicle. Summary of the invention

[0005] The main purpose of this application is to provide a vehicle warning method, electronic equipment, vehicle, storage medium and computer program product, aiming to solve the technical problem in related technologies that the accuracy of identifying risk persons is low, thereby reducing vehicle safety.

[0006] To achieve the above purpose, the present application proposes a vehicle early warning method, which is applied to an electronic device, the electronic device includes a wireless signal transmission module, and the vehicle early warning method includes:

[0007] Controlling the wireless signal transmitting module to transmit a plurality of wireless detection signals around the vehicle;

[0008] generating a target living body image based on the plurality of wireless detection signals, and extracting image features contained in the target living body image;

[0009] In the case where a risk person is identified in the target living body image according to the image features, a target early warning strategy corresponding to the risk person is determined, and an early warning operation is performed according to the target early warning strategy.

[0010] In one embodiment, the step of determining the target early warning strategy corresponding to the risk personnel includes:

[0011] Determine the position information of the risk person according to the image feature, and determine the distance parameter according to the position information of the risk person, wherein the distance parameter is the distance parameter between the risk person and the vehicle;

[0012] When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, the preset first warning strategy is determined as the target warning strategy.

[0013] In one embodiment, after the step of determining the distance parameter according to the personnel location information, the method further includes:

[0014] When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, determining the actual position information of the risk person according to the image feature;

[0015] When it is determined that the actual posture information matches the preset posture information, determining the preset second warning strategy as the target warning strategy;

[0016] When it is determined that the actual posture information does not match the preset posture information, the preset third warning strategy is determined as the target warning strategy.

[0017] In one embodiment, the step of generating a target living body image based on the plurality of wireless detection signals comprises:

[0018] Detecting initial channel state information respectively matched by the plurality of wireless detection signals, and processing each of the initial channel state information to obtain each complete channel state information;

[0019] Each of the complete channel state information is screened to determine target channel state information, and a target living body image is generated based on each of the target channel state information, wherein the target channel state information is channel state information corresponding to a wireless detection signal that detects a living body detection target.

[0020] In one embodiment, the step of detecting initial channel state information that matches each of the plurality of wireless detection signals includes:

[0021] Detecting wireless receiving signals that are matched to each of the plurality of wireless detection signals;

[0022] Determining a signal change ratio between each of the plurality of wireless detection signals and a matching wireless reception signal;

[0023] Based on each of the signal change ratios, initial channel state information that matches each of the plurality of wireless detection signals is determined.

[0024] In one embodiment, the step of processing each of the initial channel state information to obtain each complete channel state information includes:

[0025] Determining an initial amplitude parameter and an initial phase parameter respectively included in each of the initial channel state information;

[0026] Performing phase expansion on each of the initial amplitude parameters and each of the initial phase parameters, and filtering and eliminating each of the phase expanded initial amplitude parameters and each of the initial phase parameters, so as to obtain each target amplitude parameter and each target phase parameter;

[0027] Each of the target amplitude parameters and the corresponding matching target phase parameters are fused to obtain each complete channel state information.

[0028] In one embodiment, the step of screening each of the complete channel state information to determine the target channel state information includes:

[0029] Acquire preset standard channel state information, wherein the standard channel state information is channel state information corresponding to a wireless detection signal that detects a non-living detection target;

[0030] Each of the complete channel state information is compared with the standard channel state information respectively, so as to determine the complete channel state information inconsistent with the standard channel state information as the target channel state information.

[0031] In one embodiment, the step of generating a target living body image based on each target channel state information includes:

[0032] Extracting one-dimensional vector features contained in each of the target channel state information;

[0033] The one-dimensional vector features are fused to obtain two-dimensional vector features, and a target living body image is generated based on the two-dimensional vector features.

[0034] In addition, to achieve the above-mentioned purpose, the present application also proposes an electronic device, which includes: a wireless signal transmission module, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the vehicle warning method described above.

[0035] In addition, to achieve the above objectives, the present application also proposes a vehicle, which includes the electronic device as described above.

[0036] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the vehicle warning method described above are implemented.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the vehicle warning method described above are implemented.

[0038] The vehicle warning method provided in the embodiment of the present application is applied to an electronic device, which includes a wireless signal transmission module, and controls the wireless signal transmission module to transmit multiple wireless detection signals to the surroundings of the vehicle; generates a target living image based on the multiple wireless detection signals, and extracts image features contained in the target living image; when a risk person is identified in the target living image according to the image features, determines a target warning strategy corresponding to the risk person, and performs a warning operation according to the target warning strategy.

[0039] In this embodiment, during operation, the electronic device first controls its configured wireless signal transmission module to transmit multiple wireless detection signals to the surrounding of the vehicle. After that, the electronic device generates a target living image based on the multiple wireless detection signals, and extracts the image features contained in the target living image. Finally, the electronic device identifies the image features, and when it is determined that the living detection target contained in the target living image is a risky person according to the image features, it determines a target early warning strategy that matches the risky person, and then performs the early warning operation according to the target early warning strategy to issue a warning to the driver of the vehicle and / or the risky person.

[0040] In this way, the present application solves the technical problem in the related technology that the recognition accuracy of risky persons is low, thereby reducing the safety of the vehicle. That is, the present application utilizes the characteristic that the wireless detection signal will generate different channel state information when it comes into contact with a living body, and can accurately identify the risky persons around the vehicle, thereby avoiding the situation where the image data collected by the fisheye camera is of low quality due to the influence of ambient light, making it difficult to accurately identify the risky persons. The technical effect of improving the recognition accuracy of risky persons is achieved, and the safety of the vehicle is greatly increased. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 This is a schematic diagram of the structure of an electronic device involved in an embodiment of the vehicle warning method of the present application;

[0044] Figure 2 A flow chart of the first embodiment of the vehicle warning method of the present application;

[0045] Figure 3 A schematic diagram of a target living body image involved in an embodiment of the vehicle warning method of the present application;

[0046] Figure 4 A brief flowchart of the vehicle early warning method of this application;

[0047] Figure 5 This is a schematic diagram of the module structure of the vehicle warning device according to an embodiment of the present application;

[0048] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the vehicle warning method in the embodiment of the present application.

[0049] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0050] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0051] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0052] In this embodiment, for ease of description, please refer to Figure 1 , Figure 1 The structure diagram of the electronic device involved in the vehicle warning method of the present application is as follows: Figure 1As shown, an electronic device with a wireless signal transmission module configured inside, or a mobile terminal, data storage control terminal, PC and other terminals connected to an electronic control unit supporting the electronic device are explained as the execution subject. Among them, the wireless signal transmission module includes two parts: TCAM (Telematics & Connectivity Antenna Module, smart antenna module) and DHU (Digital Cockpit Head Unit, digital cockpit unit), among which TCAM includes a first Wifi Service, Core Service, MCU (Microcontroller Unit, micro control unit), and similarly, the DHU includes a second Wifi Service, QNX Service (QNX virtual machine), Android Service (Android subsystem), Algorithm Service (data processing unit), SOC (System on Chip, system-level chip) and MCU.

[0053] Based on the above electronic equipment, the overall concept of the vehicle warning method of the present application is proposed here.

[0054] With the continuous development of the automobile industry, new energy vehicles have become the preferred means of transportation for more and more users in their daily travel. In order to improve the safety of new energy vehicles, technicians usually configure multiple fisheye cameras on the vehicle to better identify risky people around the vehicle. In related technologies, after entering the locked state, the vehicle usually controls the fisheye camera to shoot the surrounding environment of the vehicle to obtain image data, and identifies whether there are risky people around the vehicle based on the image data. However, since the fisheye camera is easily affected by ambient light in the process of collecting image data, the quality of the captured image data is low. This can easily lead to a low accuracy rate in the vehicle's recognition of risky people, greatly reducing the safety of the vehicle.

[0055] In response to the above phenomenon, the present application provides a vehicle warning method, which is applied to an electronic device, and the electronic device includes a wireless signal transmission module. The vehicle warning method includes: controlling the wireless signal transmission module to transmit multiple wireless detection signals to the surroundings of the vehicle; generating a target living image based on the multiple wireless detection signals, and extracting image features contained in the target living image; when a risk person is identified in the target living image according to the image features, determining a target warning strategy corresponding to the risk person, and performing a warning operation according to the target warning strategy.

[0056] In this way, the present application solves the technical problem in the related technology that the recognition accuracy of risky persons is low, thereby reducing the safety of the vehicle. That is, the present application utilizes the characteristic that the wireless detection signal will generate different channel state information when it comes into contact with a living body, and can accurately identify the risky persons around the vehicle, thereby avoiding the situation where the image data collected by the fisheye camera is of low quality due to the influence of ambient light, making it difficult to accurately identify the risky persons. The technical effect of improving the recognition accuracy of risky persons is achieved, and the safety of the vehicle is greatly increased.

[0057] Based on the overall concept of the vehicle warning method of the present application, the embodiment of the present application provides a vehicle warning method, referring to Figure 2 , Figure 2 This is a flow chart of the first embodiment of the vehicle warning method of the present application.

[0058] In this embodiment, the vehicle early warning method is applied to an electronic device, the electronic device includes a wireless signal transmission module, and the vehicle early warning method includes steps S10 to S30:

[0059] Step S10: Control the wireless signal transmitting module to transmit a plurality of wireless detection signals around the vehicle;

[0060] Step S20: generating a target living body image based on the plurality of wireless detection signals, and extracting image features contained in the target living body image;

[0061] It should be noted that the wireless detection signal may specifically be a Wifi signal. In addition, the target living body image is image data containing the living body detection target generated after the living body detection target existing around the vehicle is detected by the Wifi signal. In addition, the image feature is an image feature contained in the target living body image that can reflect information such as the position and posture of the living body detection target.

[0062] In this embodiment, when the electronic device detects that the vehicle enters a locked state, it first calls the wireless signal transmission module configured by itself to continuously transmit multiple wireless detection signals to the surrounding environment of the vehicle. After that, the electronic device controls the wireless signal transmission module to receive the wireless reflection signals generated by each of the multiple wireless detection signals. The wireless signal transmission module then obtains multiple initial channel state information based on each wireless detection signal and each wireless reflection signal, and inputs the multiple initial channel state information into the convolutional neural network model configured by itself. The convolutional neural network model generates a target living image containing living detection targets existing around the vehicle according to each received initial channel state information. The convolutional neural network model performs phase separation processing on the target living image to obtain multiple sub-living images, and calls the encoder to encode the multiple sub-living images to identify the convolutional feature weights corresponding to each of the multiple sub-living images. The convolutional neural network model then fuses the convolutional feature weights to obtain posture fusion data, performs inference merging operation on the posture fusion data, and performs sampling processing on the posture fusion data after inference merging to extract target features.

[0063] Exemplarily, for example, when the electronic device detects that the vehicle enters the locked state, it first controls the wireless transmission module configured by itself to control the TCAM in the wireless transmission module to start, so that the first Wifi Service in the TCAM is controlled to transmit multiple first Wifi signals to the surrounding environment of the vehicle, and the first Wifi Service then receives the first reflection signals generated by the multiple first Wifi signals, thereby determining the initial CSI data corresponding to the multiple first Wifi signals according to the first reflection signals, and the TCAM then inputs the obtained initial CSI data into the DHU in the wireless transmission module, so that the initial CSI data enters the data processing unit Algorithm Service configured in the DHU; at the same time, the electronic device controls the DHU in the wireless transmission module to start, so that the second Wifi Service in the DHU transmits multiple second Wifi signals to the surrounding environment of the vehicle, and the second Wifi Service then receives the second reflection signals generated by the multiple second Wifi signals, thereby determining the initial CSI (Channel State Information) data corresponding to the multiple second Wifi signals according to the second reflection signals, and the DHU then inputs the obtained initial CSI data into the data processing unit Algorithm Service configured in the DHU. Service; Algorithm Service after obtaining each initial CSI data, and each CSI data is input to the RCNN (Regions with Convolutional Neural Networks, regional convolutional neural network) configured in the electronic device, and the RCNN generates target image data containing the live detection target based on each CSI data. After constructing the target live image, the RCNN also needs to phase the target live image to obtain multiple sub-live images. The RCNN calls the encoder to encode the multiple sub-live images respectively to identify the convolution feature weights corresponding to each of the multiple sub-live images. The RCNN then fuses the obtained convolution feature weights to obtain preliminary posture fusion data. The RCNN performs inference operations through 8-bit 3-channel RGB reshape, thereby merging the preliminary posture fusion data into 24-bit data. The RCNN then performs upsampling encoding operations and downsampling decoding operations on the 24-bit data in sequence, thereby outputting a 3-channel image, and finally determines the 3-channel image as the image feature corresponding to the target live image.

[0064] It should be noted that, in this embodiment and another embodiment, when the electronic device detects that the vehicle enters a locked state, it can also first control the initialization of the QNX virtual machine (QNX Service) configured in the DHU, so as to start the second Wifi Service and the Android subsystem (Android Service) in the DHU through the QNX virtual machine. At the same time, the electronic device also controls the Socket communication between the TCAM and the DHU to ensure that the TCAM can send the acquired initial CSI data to the Algorithm Service in the DHU.

[0065] In addition, in this embodiment and another embodiment, after controlling the TCAM to start, the electronic device also needs to first control the Core Service configured in the TCAM and the QNX virtual machine configured in the DHU to detect the communication status, so as to initiate a communication re-request in time when the heartbeat communication between the TCAM and the DHU is detected to be interrupted.

[0066] In this way, the vehicle calls the wireless transmission module to transmit multiple wireless detection signals to the cockpit, so as to identify the living detection targets around the vehicle through the multiple wireless detection signals, and generate a target living image containing the living detection target, and then extract the image features that can reflect the position, posture and other information of the living detection target through the target living image.

[0067] In a feasible implementation manner, the step of “generating a target living body image based on the multiple wireless detection signals” in the above step S20 may specifically include steps S201 to S202:

[0068] Step S201: Detecting initial channel state information matched by each of the plurality of wireless detection signals, and processing each of the initial channel state information to obtain each complete channel state information;

[0069] Step S202: Screening each of the complete channel state information to determine target channel state information, and generating a target living body image based on each of the target channel state information, wherein the target channel state information is the channel state information corresponding to the wireless detection signal that detects the living body detection target.

[0070] It should be noted that the initial channel state information is CSI data that has not been screened and cleaned. In addition, the complete channel state information is CSI data that has been cleaned and filtered and does not contain outliers. In addition, the target state channel information is the CSI data corresponding to the Wifi signal of the living detection target around the vehicle. It can be understood that the CSI data is a series of complex decimal sequences that can represent the signal wave ratio between the wireless detection signal and the corresponding wireless reception signal.

[0071] In this embodiment, after controlling the wireless signal transmission module to transmit multiple wireless detection signals to the surrounding environment, the electronic device further controls the wireless signal transmission module to receive wireless reflection signals generated by each of the multiple wireless detection signals, and controls the wireless signal transmission module to obtain initial channel state information that has not been cleaned and matches each of the multiple wireless detection signals according to the received wireless detection signals and the wireless reflection signals. The electronic device then controls the wireless signal transmission module to input each initial channel state information into a data processing unit configured in the electronic device, so that the data processing unit calls the convolutional neural network model to perform cleaning and denoising processing on each initial channel state information to obtain each complete channel state information that does not contain abnormal values. After that, the convolutional neural network model screens each complete channel state information to determine the target channel state information contained in each complete channel state information and corresponding to the wireless detection signal that detects the living detection target, and then generates a target living body image containing the target living body existing around the vehicle based on each target channel state information.

[0072] Exemplarily, for example, after the electronic device controls the first Wifi Service configured in the TCAM to continuously transmit multiple first Wifi signals to the surrounding environment of the vehicle, the electronic device controls the first Wifi Service to continuously receive the first reflected signals generated by the multiple first Wifi signals, and determines the initial CSI data that matches the multiple first Wifi signals based on the first reflected signals and the first Wifi signals. At the same time, after the electronic device controls the second Wifi Service configured in the DHU to continuously transmit multiple second Wifi signals to the surrounding environment, the electronic device controls the second Wifi Service to continuously receive the second reflected signals generated by the multiple second Wifi signals, and determines the initial CSI data that matches the multiple second Wifi signals based on the second reflected signals and the second Wifi signals. The electronic device further controls the TCAM and the DHU to input the initial CSI data obtained respectively into the data processing unit Algorithm Service configured in the DHU. The Algorithm Service performs a data cleaning operation on the initial CSI data to eliminate the abnormal data contained in the initial CSI data, thereby obtaining complete CSI data that does not contain the abnormal data. Then, the Algorithm Service performs a data cleaning operation on the initial CSI data to eliminate the abnormal data contained in the initial CSI data, thereby obtaining complete CSI data that does not contain the abnormal data. The Service filters each first Wifi signal and each second Wifi signal based on each complete CSI data, thereby determining each target first Wifi signal and each target second Wifi signal that scans the liveness detection target among each Wifi signal. The Algorithm Service determines the complete CSI data corresponding to each target first Wifi signal and each target second Wifi signal as the target CSI data for determining whether the liveness detection target is a human body, and calls the RCNN configured by itself to generate a target liveness image containing the liveness detection targets around the vehicle based on each target CSI data.

[0073] In this way, the vehicle calls the wireless transmission module to transmit multiple wireless detection signals to the cockpit, thereby identifying the living detection targets around the vehicle through the multiple wireless detection signals, and generating a target living image containing the living detection target, and then extracting the image features that can reflect the location, posture and other information of the living detection target through the target living image.

[0074] In a feasible implementation manner, the step of "detecting initial channel state information that matches each of the plurality of wireless detection signals" in the above step S201 may specifically include steps S2011 to S2013:

[0075] Step S2011: Detecting wireless receiving signals that match each of the plurality of wireless detection signals;

[0076] Step S2012: determining a signal change ratio between each of the plurality of wireless detection signals and a matching wireless reception signal;

[0077] Step S2013: determining initial channel state information that matches each of the plurality of wireless detection signals based on the signal change ratios.

[0078] In this embodiment, after controlling the wireless signal transmission module to continuously transmit multiple wireless detection signals to the surrounding environment, the electronic device further controls the wireless signal transmission module to receive the wireless reflection signals generated by each of the multiple wireless detection signals. Afterwards, the wireless signal transmission module compares the multiple wireless detection signals with each of the matching wireless reception signals, thereby determining the signal change ratio generated between each of the multiple wireless detection signals and the matching wireless reception signals. Finally, the wireless signal transmission module determines the signal change ratio corresponding to each of the multiple wireless detection signals as the initial channel state information matched by each of the multiple wireless detection signals, and inputs each initial channel state information into the above-mentioned data processing unit.

[0079] Exemplarily, for example, after controlling the first Wifi Service configured by the TCAM to transmit multiple first Wifi signals to the surrounding environment of the vehicle, the electronic device further controls the TCAM to receive first Wifi reflected signals formed by each of the first Wifi signals after being reflected by an obstacle. Then, the TCAM compares each of the first Wifi signals with the first Wifi reflected signals that match them, thereby determining a first signal wave ratio formed between each of the first Wifi signals and the matched first Wifi reflected signals. Finally, the TCAM determines the first signal wave ratio corresponding to each of the first Wifi signals as initial CSI data corresponding to each of the first Wifi signals.

[0080] Similarly, after controlling the second Wifi Service configured by the DHU to transmit multiple second Wifi signals into the surrounding environment of the vehicle, the electronic device further controls the DHU to receive second Wifi reflected signals formed by each second Wifi signal after being reflected by an obstacle. After that, the DHU compares each second Wifi signal with its matching second Wifi reflected signal, thereby determining a second signal wave ratio formed between each second Wifi signal and its matching second Wifi reflected signal. Finally, the TCAM determines the second signal wave ratio corresponding to each second Wifi signal as the second initial CSI data corresponding to each second Wifi signal. The electronic device controls the TCAM and the DHU to input the collected initial CSI data into the data processing unit Algorithm Service configured in the DHU.

[0081] In this way, the electronic device can obtain the initial channel state information generated by each of the multiple wireless detection signals by calling the wireless transmission module to transmit the multiple wireless detection signals to the surrounding environment of the vehicle.

[0082] In a feasible implementation manner, the step of “processing each of the initial channel state information to obtain each complete channel state information” in the above step S201 may specifically include steps S2014 to S2016:

[0083] Step S2014: determining an initial amplitude parameter and an initial phase parameter respectively included in each of the initial channel state information;

[0084] Step S2015: performing phase unwrapping on each of the initial amplitude parameters and each of the initial phase parameters, and filtering and eliminating each of the phase-unwrapped initial amplitude parameters and each of the initial phase parameters to obtain each target amplitude parameter and each target phase parameter;

[0085] Step S2016: Fusing the target amplitude parameters and the respective matching target phase parameters to obtain complete channel state information.

[0086] In this embodiment, after acquiring each initial channel state information, the data processing unit first reads the initial amplitude parameters and initial phase parameters contained in each initial channel state information, and then the data processing unit performs a phase unwrapping operation on each initial amplitude parameter and each initial phase parameter. The data processing unit then performs a filtering and elimination operation on each initial amplitude parameter and each initial phase parameter after the phase unwrapping to eliminate the abnormal values ​​contained in each initial amplitude parameter and each initial phase parameter, and obtains each target amplitude parameter and each target phase parameter. Finally, the data processing unit fuses each target amplitude parameter with each matching target phase parameter, so as to obtain each complete channel state information that does not contain abnormal values.

[0087] Exemplarily, for example, after the data processing unit Algorithm Service obtains each initial CSI data input by TCAM and DHU, Algorithm Service first extracts the initial amplitude parameters and initial phase parameters contained in each initial CSI data, and then Algorithm Service performs phase unfolding processing on each initial amplitude parameter and each initial phase parameter, so that the unfolded amplitude curve and phase curve are restored to continuous curves. At the same time, Algorithm Service performs median filtering and uniform filtering on each unfolded initial amplitude parameter and each initial phase parameter to eliminate outliers of the unfolded initial amplitude parameter and initial phase parameter in the time and frequency domains, thereby obtaining target amplitude parameters and target phase parameters with target outliers eliminated. Finally, Algorithm Service fuses each target amplitude parameter with its matching target phase parameter to obtain complete CSI data that does not contain abnormal values.

[0088] It should be noted that the initial CSI data has the problem of random drift and reversal of amplitude and phase. Therefore, if the initial CSI data is not denoised, the abnormal values ​​contained in the initial CSI data will have a certain impact on the final detection result. In addition, the amplitude and phase in the initial CSI data can be calculated, and in the process of calculating the amplitude and phase, the phase and amplitude that exceed the preset function range will be folded, resulting in discontinuity of phase and amplitude. In this case, it is necessary to perform a phase expansion operation to restore each phase and amplitude to a continuous state. It can be understood that the specific calculation process of the phase expansion operation is a prior art, so it will not be repeated here. In addition, the specific process of median filtering and uniform filtering of the expanded phase value and amplitude parameters by Algorithm Service is a prior art, so it will not be repeated here.

[0089] In this way, the electronic device can remove the abnormal values ​​contained in each initial channel state information, thereby obtaining each complete channel state information that does not contain the abnormal value, thereby ensuring the accuracy of the detection result of the person left in the cabin.

[0090] In a feasible implementation manner, the step of “screening each of the complete channel state information to determine the target channel state information” in the above step S202 may specifically include steps S2021 to S2022:

[0091] Step S2021: obtaining preset standard channel state information, wherein the standard channel state information is channel state information corresponding to a wireless detection signal detecting a non-living detection target;

[0092] Step S2022: Compare each of the complete channel state information with the standard channel state information respectively, so as to determine the complete channel state information inconsistent with the standard channel state information as the target channel state information.

[0093] It should be noted that the standard channel state information is the CSI data corresponding to the Wifi signal that detects a non-living detection target. It can be understood that when multiple Wifi signals detect non-living detection targets, the signal change ratios between the multiple Wifi signals and their corresponding Wifi reflection signals should be consistent, that is, when the CSI data corresponding to the multiple Wifi signals changes, it can be determined that the Wifi signal detects a moving living detection target.

[0094] In this embodiment, after obtaining each complete channel state information, the data processing unit first reads the storage module configured in the electronic device to obtain the preset standard channel state information corresponding to the wireless detection signal indicating that a non-living detection target is detected. After that, the data processing unit compares each obtained complete channel state information with the standard channel state information respectively to obtain multiple first comparison results. The data processing unit reads the multiple first comparison results to determine the complete channel state information that is inconsistent with the standard channel state information as the target channel state information corresponding to the wireless detection signal that detects a living detection target based on the multiple first comparison results.

[0095] Exemplarily, for example, after obtaining each complete CSI data, the Algorithm Service first reads a storage module configured in the electronic device to obtain preset standard CSI data corresponding to when the Wifi signal detects a non-living detection target. Then, the Algorithm Service compares each complete CSI data with the standard CSI data respectively to obtain multiple first comparison results. Finally, the Algorithm Service screens each complete CSI data according to the multiple first comparison results, thereby determining the complete CSI data inconsistent with the standard CSI data as the target CSI data that can generate a living image that matches the living detection target.

[0096] In this way, the electronic device can determine the target channel state information corresponding to the wireless detection signal that detects the living body detection target from the multiple complete channel state information transmitted, and then generate a target living body image containing the living body detection target based on the target channel state information.

[0097] In a feasible implementation manner, the step of “generating a target living body image based on each target channel state information” in the above step S202 may specifically include steps S2024 to S2025:

[0098] Step S2024: extracting the one-dimensional vector features contained in each of the target channel state information;

[0099] Step S2025: fusing the one-dimensional vector features to obtain two-dimensional vector features, and generating a target living body image based on the two-dimensional vector features.

[0100] It should be noted that the one-dimensional vector feature is the latent spatial feature contained in the target channel state information. The one-dimensional feature vector specifically includes an amplitude tensor and a phase tensor, wherein the size of each tensor is 150×3×3 (5 consecutive samples, 30 frequencies, 3 transmitters and 3 receivers). In this way, RCNN (Regions with Convolutional Neural Networks) can extract the spatial features contained in the one-dimensional vector feature through the last two dimensions of each tensor.

[0101] In this embodiment, after determining multiple target channel state information, the data processing unit first inputs the multiple target channel state information into the convolutional neural network model configured by itself, so that the convolutional neural network model calls multiple encoders to extract the one-dimensional feature vectors contained in each of the multiple target channel state information. After that, the data processing unit fuses each one-dimensional feature vector to obtain a two-dimensional fused feature, and converts the two-dimensional fused feature into a feature map in the spatial domain, and then upsamples the converted two-dimensional feature map to obtain a target living image containing a living detection target.

[0102] For example, see Figure 3 , Figure 3This is a schematic diagram of a target living body image involved in an embodiment of the vehicle warning method of this application, Algorithm After determining multiple target CSI data, Service first inputs the multiple target CSI data into the RCNN configured by itself, so that RCNN calls two different encoders to process each target CSI data to extract the amplitude tensor and phase tensor of size 150×3×3 corresponding to each of the multiple target CSI data. RCNN then determines the amplitude tensor and phase tensor corresponding to each target CSI data as 1D features. After that, RCNN extracts the spatial information contained in each 1D feature based on the last two dimensions (3 transmitters and 3 receivers) of each amplitude tensor and each phase tensor, and fuses each 1D feature according to each spatial information to obtain a fused 1D feature. RCNN then reshapes the fused 1D feature into an initial 2D feature map of 24×24, and extracts the spatial information contained in the initial 2D feature map through the two convolution blocks configured by itself to obtain a 2D feature vector with a spatial dimension of 6×6. Finally, RCNN performs an upsampling operation on the 2D feature vector with a spatial dimension of 6×6 to adjust the 2D feature vector to the following: Figure 3 The 3×720×1280 size shown contains a target living image in the image domain of the living body detection target.

[0103] In this way, the electronic device can construct a target living body image including a living body detection target based on the filtered target channel state information.

[0104] Step S30: when a risk person is identified in the target living body image according to the image features, determining a target early warning strategy corresponding to the risk person, and performing an early warning operation according to the target early warning strategy;

[0105] In this embodiment, after extracting the image features, the convolutional neural network inputs the image features into an image processing module configured for the electronic device. The image processing module first determines whether the liveness detection target is a risk person based on the image features. When the image processing module detects that the liveness detection target is a risk person, the image processing module determines a target warning strategy that matches the risk person based on the image features, and uploads the target warning strategy to the electronic device. The electronic device performs a warning operation according to the target warning strategy to issue a warning to the liveness detection target through the vehicle and / or issue a warning to the driver through the vehicle.

[0106] Exemplarily, for example, after RCNN extracts image features, the electronic device inputs the features extracted by RCNN to the SOC configured in the DHU, and the SOC calls the preset YOLO image processing model to process the image features to determine whether the liveness detection target in the target live image is a human being based on the image features, and when the liveness detection target is detected to be a human being, the human being is determined as a risky person, and the SOC further determines a target warning strategy that matches the risky person based on the image features, and uploads the target warning strategy to the MCU, and the MCU performs the warning operation according to the target warning strategy to issue a warning to the liveness detection target through the vehicle / issue a warning to the driver through the vehicle.

[0107] In this way, the electronic device can identify whether there are risky persons around the vehicle based on image features, and when risky persons are detected around the vehicle, further identify the possibility that the risky persons may cause harm to the vehicle, thereby determining an appropriate early warning strategy to warn the risky persons and / or the driver through the vehicle.

[0108] In a feasible implementation manner, the step of "determining the target warning strategy corresponding to the risk personnel" in the above step S30 may specifically include steps S301 to S302:

[0109] Step S301: determining the position information of the risk person according to the image features, and determining a distance parameter according to the position information, wherein the distance parameter is a distance parameter between the risk person and the vehicle;

[0110] Step S302: when it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, determining the preset first warning strategy as a target warning strategy.

[0111] In this embodiment, when the image processing module detects that the liveness detection target contained in the target liveness image is a risk person, the image processing module further determines the position of the risk person in the image based on the image features, and based on the actual position information of the risk person in the real environment, at the same time, the image processing module reads the above-mentioned storage module to obtain the vehicle position information of the vehicle, and determines the distance parameter generated between the vehicle and the risk person based on the actual position information and the vehicle position information. Finally, the image processing module reads the storage module to obtain a preset distance parameter threshold, and compares the distance parameter with the distance parameter threshold, so that when it is detected that the distance parameter is less than or equal to the distance parameter threshold, the first warning strategy that needs to issue warnings to the driver of the vehicle and the risk person at the same time is determined as the target warning strategy that needs to be executed.

[0112] Exemplarily, for example, when the SOC identifies that the liveness detection target contained in the target liveness image is a risk person based on the image feature, the SOC further extracts the position feature information contained in the image feature, and the SOC determines the actual position information of the risk person based on the position feature information. After that, the SOC reads the storage module configured in the electronic device to obtain the vehicle position information of the vehicle, and determines the risk distance parameter generated between the risk person and the vehicle based on the actual position information and the vehicle position information. Finally, the SOC reads the storage module to obtain a preset distance parameter threshold, and compares the risk distance parameter with the distance parameter threshold. When the SOC detects that the risk distance parameter is less than or equal to the distance parameter threshold, it determines that the risk person is at risk of contacting the vehicle. The SOC then determines that the first warning strategy that issues a warning to the risk person and the driver at the same time through the vehicle is the target warning strategy among the preset multiple warning strategies, and the SOC sends the preset warning information to the MCU (Microcontroller) according to the first warning strategy. The MCU converts the warning information into a CAN signal, and sends the CAN signal to the buzzer, LED headlights and other hardware devices connected to the DHU through the BUS in the vehicle, so as to control the buzzer, LED headlights and other hardware devices to enter the operating state. At the same time, the CAN signal is sent to the TGM through the BUS, so that the TGM sends the warning information to the mobile device used by the driver by wireless transmission, so as to remind the driver through the mobile device that the vehicle is at risk.

[0113] In addition, in this embodiment and another embodiment, after determining the personal location information of the risk person, the SOC can, in addition to obtaining the vehicle location information by reading the storage module, also determine the vehicle location information corresponding to the vehicle through the GPS module configured on the vehicle, and determine the distance parameters between the risk person and the vehicle based on the vehicle location information and the personal location information.

[0114] In addition, in this embodiment and another embodiment, in addition to determining the distance parameters between the risk person and the vehicle based on the vehicle position information and the person position information, the SOC can also further call the camera device configured on the vehicle to photograph the risk person to capture target image data after determining the person position information of the risk person based on the image features, and then determine the distance parameters between the risk person and the vehicle based on the second image feature of the target image data and the person position information.

[0115] In addition, in this embodiment and another embodiment, after extracting the image features, the SOC can also calculate the distance parameters between the risk person and the vehicle based on the characteristic parameters such as the signal reception time and the signal transmission time contained in the image features. It can be understood that since the target living body image is image data generated based on the CSI data, the characteristics of the living body image data include the corresponding Wifi signal transmission time and Wifi signal reception time. In this way, the SOC can calculate the distance parameters between the risk person and the vehicle based on the Wifi signal transmission time, the Wifi signal reception time and the preset Wifi signal propagation speed.

[0116] In this embodiment, when the electronic device detects that the vehicle enters the locked state, it first calls the wireless signal transmission module configured by itself to continuously transmit multiple wireless detection signals to the surrounding environment of the vehicle. After that, the electronic device controls the wireless signal transmission module to receive the wireless reflection signals generated by each of the multiple wireless detection signals. The wireless signal transmission module then obtains multiple initial channel state information based on each wireless detection signal and each wireless reflection signal, and inputs the multiple initial channel state information into the convolutional neural network model configured by itself. The convolutional neural network model generates a target living image containing the living detection targets existing around the vehicle based on the received initial channel state information. The convolutional neural network model performs phase separation processing on the target living image to obtain multiple sub-living images, and calls the encoder to encode the multiple sub-living images to identify The convolution feature weights corresponding to the multiple sub-living images are respectively obtained, and the convolution neural network model then fuses the convolution feature weights to obtain posture fusion data, performs inference merging operation on the posture fusion data, and performs sampling processing on the posture fusion data after inference merging to extract target features. Finally, the convolution neural network inputs the image features into the image processing module configured by the electronic device, and the image processing module first determines whether the living body detection target is a risk person based on the image features. When the image processing module detects that the living body detection target is a risk person, it determines the target warning strategy that matches the risk person according to the image features, and uploads the target warning strategy to the electronic device. The electronic device performs the warning operation according to the target warning strategy to issue a warning to the living body detection target through the vehicle and / or issue a warning to the driver through the vehicle.

[0117] In this way, the present application solves the technical problem in the related technology that the recognition accuracy of risky persons is low, thereby reducing the safety of the vehicle. That is, the present application utilizes the characteristic that the wireless detection signal will generate different channel state information when it comes into contact with a living body, and can accurately identify the risky persons around the vehicle, thereby avoiding the situation where the image data collected by the fisheye camera is of low quality due to the influence of ambient light, making it difficult to accurately identify the risky persons. The technical effect of improving the recognition accuracy of risky persons is achieved, and the safety of the vehicle is greatly increased.

[0118] Based on the first embodiment of the present application, a second embodiment of the present application is proposed. In the second embodiment of the present application, the same or similar contents as the above embodiments can be referred to the above introduction, and will not be repeated later. On this basis, after the above step S302, the vehicle warning method of the present application can also include steps A10 to A30:

[0119] Step A10: when it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, determining the actual position information of the risk person according to the image feature;

[0120] Step A20: When it is determined that the actual posture information matches the preset posture information, determining the preset second warning strategy as the target warning strategy;

[0121] Step A30: When it is determined that the actual posture information does not match the preset posture information, determine the preset third warning strategy as the target warning strategy.

[0122] In this embodiment, after detecting the distance parameter generated between the risk person and the vehicle, the image processing module can also compare the distance parameter with the above-mentioned distance parameter threshold, so that when it is detected that the distance parameter is less than or equal to the distance parameter threshold, the posture feature contained in the image feature is further extracted, and the actual posture information of the risk person is determined according to the posture feature. After that, the image processing module reads the above-mentioned storage module to obtain a plurality of preset posture information representing dangerous contact operations, and compares the actual posture information with the plurality of preset posture information respectively to obtain a plurality of second comparison results. The image processing module reads the plurality of second comparison results, and then, based on the plurality of second comparison results, when it is determined that the actual posture information matches at least one preset posture information, the preset second warning strategy that needs to issue a warning to the driver and the risk person at the same time is determined as the target warning strategy. Similarly, based on the plurality of second comparison results, when it is determined that the actual posture information does not match a plurality of preset standard postures, the image processing module determines that the third warning strategy that only issues a warning to the risk person is the target warning strategy.

[0123] Exemplarily, for example, after detecting the distance parameter generated between the risk person and the vehicle, the SOC can also compare the distance parameter with the above-mentioned distance parameter threshold, so as to determine that the risk person is likely to be in direct contact with the vehicle when detecting that the distance parameter is less than or equal to the distance parameter threshold. The SOC further extracts the posture feature information contained in the image feature, and determines the actual posture information adopted by the risk person when contacting the vehicle based on the posture feature information. After that, the SOC reads the above-mentioned storage module to obtain a plurality of preset posture information corresponding to risk actions such as scratching the vehicle and stealing oil, and compares the actual posture information with each preset posture information to obtain a plurality of second comparison results. The SOC reads the plurality of second comparison results, and when it is determined that among the plurality of second comparison results, there is at least one second comparison result that matches the actual posture information and the preset posture information, it is determined that the risk person has the intention to harm the vehicle. The SOC then uses the second warning strategy that simultaneously issues warnings to the driver and the risk person as the target warning strategy, and the SOC then issues the preset warning according to the second warning strategy. The information is sent to the MCU, which converts the warning information into CAN signals, and sends the CAN signals to the buzzer, LED headlights and other hardware devices connected to the DHU through the BUS bus in the vehicle to control the buzzer, LED headlights and other hardware devices to enter the operating state. At the same time, the CAN signal is sent to the TGM through the BUS bus, so that the warning information is sent to the mobile device used by the driver by wireless transmission through the TGM, so as to prompt the driver through the mobile device that the vehicle is at risk. Similarly, when the SOC detects that multiple second comparison results are that the actual posture information and the preset posture information do not match, it determines that the risk person has no intention of harming the vehicle. At this time, the SOC will use the third warning strategy of only issuing warnings to the risk person as the target warning strategy. The SOC then sends the preset warning information to the MCU according to the third warning strategy, and the MCU converts the warning information into CAN signals, and sends the CAN signals to the buzzer, LED headlights and other hardware devices connected to the DHU through the BUS bus in the vehicle to control the buzzer, LED headlights and other hardware devices to enter the operating state.

[0124] It should be noted that the specific content of the second early warning strategy may be consistent with the above-mentioned first early warning strategy.

[0125] In this way, when the electronic device detects the possibility of a risk person coming into contact with the vehicle, it can further identify the posture adopted by the risk person when contacting the vehicle, and thus judge whether the risk person will cause damage to the vehicle based on the posture, and then adopt different early warning strategies based on the judgment result, thereby avoiding the situation where the driver is directly warned only when the presence of a risk person is detected around the vehicle. This achieves the technical effect of ensuring that the electronic device can identify whether the risk person has the intention to harm the vehicle, and issue a warning to the driver when it is identified whether the risk person has the intention to harm the vehicle.

[0126] For example, in order to help understand the implementation process of the vehicle warning method obtained by combining this embodiment with the above embodiments, please refer to Figure 4 , Figure 4 This is a brief flow chart of the vehicle warning method of this application, specifically:

[0127] In this embodiment, when the electronic device detects that the vehicle enters the locked state, it first starts the TCAM and DHU configured by itself, and the Core Service in the TCAM detects whether the TCAM and the DHU can communicate normally. When the CoreService detects that the TCAM and the DHU communicate normally, the electronic device controls the first Wifi Service in the TCAM to transmit multiple first Wifi signals to the surrounding environment of the vehicle, and controls the first Wifi Service to receive the first Wifi reflection signals generated by each first Wifi signal. The first Wifi Service generates multiple first initial CSI data according to each first Wifi signal and each first Wifi reflection signal, and the TCAM then sends the generated first initial CSI data to the data processing unit in the DHU.

[0128] At the same time, the electronic device controls the second Wifi Service in the DHU to transmit multiple second Wifi signals to the surrounding environment of the vehicle, and controls the second Wifi Service to receive the second Wifi reflected signals generated by each of the second Wifi signals, so that the second Wifi Service generates multiple second initial CSI data according to each of the second Wifi signals and each of the second Wifi reflected signals, and the DHU further sends the generated second CSI data to the data processing unit;

[0129] Afterwards, the data processing unit performs noise reduction processing on each of the first initial CSI data and each of the second initial CSI data to obtain a plurality of complete CSI data. At this time, the data processing unit reads a storage module in the electronic device to obtain preset CSI data corresponding to the Wifi signal when a non-living detection target is detected, and compares each of the complete CSI data with each of the preset CSI data to obtain a plurality of comparison results. The data processing unit thereby determines the target CSI data corresponding to the Wifi signal that detects the living detection target in each of the complete CSI data based on the plurality of comparison results.

[0130] Afterwards, the data processing unit inputs each target CSI data into a preset convolutional neural network model, and the convolutional neural network model extracts a one-dimensional feature vector contained in each target CSI data, and fuses each one-dimensional feature vector to generate a target detection image containing a live detection target. The convolutional neural network model then extracts image features of the target detection image, and sends the image features to the SOC in the electronic device, and the SOC matches the image features with preset human features. When the SOC determines that the image features match the human features, the SOC determines the live detection target as a risk person in the vehicle's surrounding environment, and determines the distance parameter generated between the risk person and the vehicle based on the image features. When the SOC detects that the distance parameter is less than or equal to the preset distance parameter threshold, it determines that the first warning strategy that needs to issue a warning to the driver of the vehicle and the risk person at the same time is the target warning strategy;

[0131] Finally, the electronic equipment performs warning operations according to the target warning strategy to control the vehicle's hardware devices such as buzzers and headlights to enter the operating state, thereby issuing warnings to risk personnel. At the same time, the electronic equipment sends warning information to the mobile terminal held by the driver according to the target warning strategy, so as to display prompt information to the driver through the mobile terminal, thereby prompting the driver that there is a risk in the vehicle.

[0132] In addition, when the SOC detects that the distance parameter is less than or equal to the preset distance parameter threshold, the SOC can also determine the real-time posture information of the risk person based on the image features, and judge whether the risk person has the intention to harm the vehicle according to the real-time posture information. After that, when the SOC determines that the risk person has the intention to harm the vehicle, it determines the preset second warning strategy (consistent with the above-mentioned first warning strategy) as the target warning strategy, and executes the warning operation to control the vehicle to issue prompts to the risk person and the driver at the same time. Similarly, when the SOC determines that the risk person has no intention to harm the vehicle, it determines the third warning strategy of only issuing warnings to the risk person as the target warning strategy, and executes the warning operation according to the third warning strategy to control the vehicle to only issue prompts to the risk person;

[0133] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the vehicle warning method of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0134] This application also provides a vehicle warning device, please refer to Figure 5 The vehicle warning device is applied to an electronic device, the electronic device includes a wireless signal transmission module, and the device includes:

[0135] The signal detection module 10 is used to control the wireless signal transmission module to transmit a plurality of wireless detection signals around the vehicle;

[0136] A feature extraction module 20, configured to generate a target living body image based on the plurality of wireless detection signals, and extract image features contained in the target living body image;

[0137] The decision screening module 30 is used to determine the target early warning strategy corresponding to the risk person when a risk person is identified in the target living body image according to the image features, and perform an early warning operation according to the target early warning strategy.

[0138] In a feasible implementation manner, the decision screening module 30 is further used to:

[0139] Determine the position information of the risk person according to the image feature, and determine the distance parameter according to the position information of the risk person, wherein the distance parameter is the distance parameter between the risk person and the vehicle;

[0140] When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, the preset first warning strategy is determined as the target warning strategy.

[0141] In a feasible implementation manner, the decision screening module 30 is further used to:

[0142] When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, determining the actual position information of the risk person according to the image feature;

[0143] When it is determined that the actual posture information matches the preset posture information, determining the preset second warning strategy as the target warning strategy;

[0144] When it is determined that the actual posture information does not match the preset posture information, the preset third warning strategy is determined as the target warning strategy.

[0145] In a feasible implementation manner, the feature extraction module 20 is further used for:

[0146] Detecting initial channel state information respectively matched by the plurality of wireless detection signals, and processing each of the initial channel state information to obtain each complete channel state information;

[0147] Each of the complete channel state information is screened to determine target channel state information, and a target living body image is generated based on each of the target channel state information, wherein the target channel state information is channel state information corresponding to a wireless detection signal that detects a living body detection target.

[0148] In a feasible implementation manner, the feature extraction module 20 is further used for:

[0149] Detecting wireless receiving signals that are matched to each of the plurality of wireless detection signals;

[0150] Determining a signal change ratio between each of the plurality of wireless detection signals and a matching wireless reception signal;

[0151] Based on each of the signal change ratios, initial channel state information that matches each of the plurality of wireless detection signals is determined.

[0152] In a feasible implementation manner, the feature extraction module 20 is further used for:

[0153] Determining an initial amplitude parameter and an initial phase parameter respectively included in each of the initial channel state information;

[0154] Performing phase expansion on each of the initial amplitude parameters and each of the initial phase parameters, and filtering and eliminating each of the phase expanded initial amplitude parameters and each of the initial phase parameters, so as to obtain each target amplitude parameter and each target phase parameter;

[0155] Each of the target amplitude parameters and the corresponding matching target phase parameters are fused to obtain each complete channel state information.

[0156] In a feasible implementation manner, the feature extraction module 20 is further used for:

[0157] Acquire preset standard channel state information, wherein the standard channel state information is channel state information corresponding to a wireless detection signal that detects a non-living detection target;

[0158] Each of the complete channel state information is compared with the standard channel state information respectively, so as to determine the complete channel state information inconsistent with the standard channel state information as the target channel state information.

[0159] In a feasible implementation manner, the feature extraction module 20 is further used for:

[0160] Extracting one-dimensional vector features contained in each of the target channel state information;

[0161] The one-dimensional vector features are fused to obtain two-dimensional vector features, and a target living body image is generated based on the two-dimensional vector features.

[0162] The vehicle warning device provided by the present application adopts the vehicle warning method in the above embodiment, which can solve the problem of low accuracy in identifying risk personnel in the related technology. Compared with the prior art, the beneficial effects of the vehicle warning device provided by the present application are the same as the beneficial effects of the vehicle warning method provided by the above embodiment, and the other technical features in the vehicle warning device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0163] The present application provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the vehicle warning method in the above-mentioned embodiment one.

[0164] Reference below Figure 6 , which shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present application. The electronic device in the embodiments of the present application may include, but is not limited to, an electronic device with a wireless signal transmission module configured therein, or a mobile terminal, a data storage control terminal, a PC, or other terminals connected to an electronic control unit supporting the electronic device. Figure 6 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0165] like Figure 6As shown, the electronic device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the electronic device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows an electronic device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0166] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0167] The electronic device provided by the present application adopts the vehicle early warning method in the above embodiment, which can solve the problem of low accuracy in identifying risk personnel in the related technology. Compared with the prior art, the beneficial effects of the electronic device provided by the present application are the same as the beneficial effects of the vehicle early warning method provided by the above embodiment, and the other technical features in the electronic device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0168] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0169] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0170] The present application provides a vehicle having the electronic device as described above, and the electronic device is used to execute the vehicle early warning method in the above embodiment.

[0171] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the vehicle warning method in the above-mentioned embodiment.

[0172] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0173] The computer-readable storage medium may be included in the electronic device, or may exist independently without being installed in the electronic device.

[0174] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by an electronic device, the electronic device: controls the wireless signal transmission module to transmit multiple wireless detection signals around the vehicle; generates a target living image based on the multiple wireless detection signals, and extracts image features contained in the target living image; when a risk person is identified in the target living image according to the image features, determines a target early warning strategy corresponding to the risk person, and performs an early warning operation according to the target early warning strategy.

[0175] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0176] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0177] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0178] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned vehicle early warning method, and can solve the problem of low accuracy in identifying risk personnel in related technologies. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the vehicle early warning method provided in the above-mentioned embodiment, and will not be repeated here.

[0179] The present application also provides a computer program product, including a computer program, which implements the steps of the vehicle warning method as described above when executed by a processor.

[0180] The computer program product provided by this application can solve the problem of low accuracy in identifying risk personnel in related technologies. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the vehicle warning method provided by the above embodiment, which will not be repeated here.

[0181] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A vehicle early warning method, characterized in that: The vehicle early warning method is applied to an electronic device, the electronic device includes a wireless signal transmission module, and the vehicle early warning method includes: Controlling the wireless signal transmitting module to transmit a plurality of wireless detection signals around the vehicle; generating a target living body image based on the plurality of wireless detection signals, and extracting image features contained in the target living body image; In the case where a risk person is identified in the target living body image according to the image features, a target early warning strategy corresponding to the risk person is determined, and an early warning operation is performed according to the target early warning strategy.

2. The vehicle warning method according to claim 1, characterized in that: The step of determining the target early warning strategy corresponding to the risk personnel includes: Determine the position information of the risk person according to the image feature, and determine the distance parameter according to the position information of the risk person, wherein the distance parameter is the distance parameter between the risk person and the vehicle; When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, the preset first warning strategy is determined as the target warning strategy.

3. The vehicle early warning method according to claim 2, characterized in that: After the step of determining the distance parameter according to the personnel position information, the method further includes: When it is detected that the distance parameter is less than or equal to a preset distance parameter threshold, determining the actual position information of the risk person according to the image feature; When it is determined that the actual posture information matches the preset posture information, determining the preset second warning strategy as the target warning strategy; When it is determined that the actual posture information does not match the preset posture information, the preset third warning strategy is determined as the target warning strategy.

4. The vehicle warning method according to claim 1, characterized in that: The step of generating a target living body image based on a plurality of the wireless detection signals comprises: Detecting initial channel state information respectively matched by the plurality of wireless detection signals, and processing each of the initial channel state information to obtain each complete channel state information; Each of the complete channel state information is screened to determine target channel state information, and a target living body image is generated based on each of the target channel state information, wherein the target channel state information is channel state information corresponding to a wireless detection signal that detects a living body detection target.

5. The vehicle warning method according to claim 4, characterized in that: The step of detecting initial channel state information that matches each of the plurality of wireless detection signals comprises: Detecting wireless receiving signals that are matched to each of the plurality of wireless detection signals; Determining a signal change ratio between each of the plurality of wireless detection signals and a matching wireless reception signal; Based on each of the signal change ratios, initial channel state information that matches each of the plurality of wireless detection signals is determined.

6. The vehicle warning method according to claim 4, characterized in that: The step of processing each of the initial channel state information to obtain each complete channel state information comprises: Determining an initial amplitude parameter and an initial phase parameter respectively included in each of the initial channel state information; Performing phase expansion on each of the initial amplitude parameters and each of the initial phase parameters, and filtering and eliminating each of the phase expanded initial amplitude parameters and each of the initial phase parameters, so as to obtain each target amplitude parameter and each target phase parameter; Each of the target amplitude parameters and the corresponding matching target phase parameters are fused to obtain each complete channel state information.

7. The vehicle warning method according to claim 4, characterized in that: The step of screening each of the complete channel state information to determine the target channel state information comprises: Acquire preset standard channel state information, wherein the standard channel state information is channel state information corresponding to a wireless detection signal that detects a non-living detection target; Each of the complete channel state information is compared with the standard channel state information respectively, so as to determine the complete channel state information inconsistent with the standard channel state information as the target channel state information.

8. The vehicle early warning method according to claim 4, characterized in that: The step of generating a target living body image based on each target channel state information comprises: Extracting one-dimensional vector features contained in each of the target channel state information; The one-dimensional vector features are fused to obtain two-dimensional vector features, and a target living body image is generated based on the two-dimensional vector features.

9. An electronic device, characterized in that: The device comprises: a wireless signal transmission module, a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the vehicle warning method as claimed in any one of claims 1 to 8.

10. A vehicle, characterized in that: The vehicle comprises the electronic device as claimed in claim 9.

11. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the vehicle warning method according to any one of claims 1 to 8 are implemented.

12. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the vehicle warning method according to any one of claims 1 to 8 are implemented.

Images