Vehicle alarm method and device, vehicle, storage medium and program product

Abstract

The invention relates to the technical field of vehicle safety, and discloses a vehicle alarm method and device, a vehicle, a storage medium and a program product, the vehicle comprises a vehicle-mounted camera and a radar which are arranged on a vehicle body, and the method comprises the following steps: determining whether an object is an abnormal object according to a monitoring result of the radar; when the object is an abnormal object, obtaining a current image collected by a vehicle-mounted camera; personnel detection is carried out on the current image, a detection result is obtained, and the detection result comprises current position information of at least one person; and determining whether to trigger alarm according to the detection result. On one hand, the running time of detection algorithms such as a deep learning algorithm can be shortened, so that power consumption is reduced, on the other hand, dual detection is performed through the monitoring result and the current image, whether a person performs suspicious activity around the vehicle body or not is determined, the possibility of false detection can be reduced, and the alarm accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle safety technology, and in particular to a vehicle alarm method, device, vehicle, storage medium and program product. Background Art

[0002] Sentry Mode is an intelligent vehicle safety feature. When activated, the vehicle's surroundings are monitored using existing sensors or onboard cameras. If suspicious activity is detected, the vehicle is activated to sound an alarm, providing a warning. Currently, detection of unusually close proximity is typically based on deep learning algorithms. However, to ensure the speed and accuracy of deep learning algorithms, high-performance hardware (such as graphics processing units) is typically required. These hardware devices themselves consume a lot of energy, and running deep learning algorithms for extended periods of time further increases power consumption. Summary of the Invention

[0003] In view of this, the present invention provides a vehicle alarm method, device, vehicle, storage medium and program product to improve the problem of high power consumption caused by long-term operation of deep learning algorithms.

[0004] In a first aspect, the present invention provides a vehicle alarm method, wherein the vehicle includes a vehicle-mounted camera and a radar arranged on the vehicle body, and the method includes: determining whether an object is an abnormal object based on the monitoring results of the radar; when the object is an abnormal object, obtaining a current image captured by the vehicle-mounted camera; performing personnel detection on the current image to obtain a detection result, wherein the detection result includes the current location information of at least one person; and determining whether to trigger an alarm based on the detection result.

[0005] The vehicle alarm method provided in this embodiment first determines whether an object is an abnormal object based on radar monitoring results. If the object is an abnormal object, it then obtains the current image captured by the vehicle camera, then performs personnel detection on the current image, and determines whether to trigger an alarm based on the obtained detection results. This embodiment first uses the radar monitoring results to determine whether the object is an abnormal object, and if the object is an abnormal object, it then performs personnel detection on the current image to determine the detection result. On the one hand, it can reduce the running time of detection algorithms such as deep learning algorithms, thereby reducing power consumption. On the other hand, by performing dual detection of monitoring results and the current image, it can reduce the possibility of false detection and improve the accuracy of the alarm.

[0006] In an optional embodiment, determining whether an object is an abnormal object based on the monitoring results of the radar includes: determining the degree of fluctuation of a first distance within the detection time period based on the monitoring results within the detection time period, wherein the first distance is the distance between the object and the radar; and determining whether the object is an abnormal object based on the degree of fluctuation of the first distance within the detection time period.

[0007] In this embodiment, after obtaining the monitoring results of the radar, the fluctuation degree of the first distance within the detection time period is determined based on the monitoring results within the detection time period, and then based on the fluctuation degree of the first distance within the detection time period, it is determined whether the object is an abnormal object. Based on the fluctuation degree of the first distance, it can be more convenient and efficient to determine whether the monitored object is an abnormal object.

[0008] In an optional embodiment, whether the object is an abnormal object is determined based on the degree of fluctuation of the first distance within the detection time period, including: when the degree of fluctuation of the first distance is greater than or equal to a preset fluctuation value, and the maximum first distance is less than or equal to the preset distance, the object is determined to be an abnormal object.

[0009] In an optional embodiment, before performing personnel detection on the current image to obtain the detection result, the method also includes: performing grayscale processing on the current image to obtain a grayscale image; performing mean processing on the grayscale image to determine the brightness value; processing the current image to obtain the detection result, including: when the brightness value is greater than or equal to the brightness threshold, processing the current image to obtain the detection result.

[0010] In this embodiment, the brightness of the current image is filtered before performing person detection on the current image, which can reduce the running power consumption of the posture estimation algorithm.

[0011] In an optional implementation, obtaining the current image of the vehicle-mounted camera includes: obtaining the current image of the vehicle-mounted camera by frame extraction.

[0012] In an optional embodiment, determining whether to trigger an alarm based on the detection results includes: updating the status information of the person based on the detection results, the status information including time information reflecting that the person is within the alarm range of the vehicle; and determining whether to trigger an alarm based on the updated status information of the person.

[0013] In an optional implementation, determining whether to trigger an alarm based on the updated status information of the personnel includes: triggering an alarm when the time information corresponding to any one of the personnel is greater than a preset alarm duration.

[0014] In this embodiment, an alarm is triggered when the duration and / or continuous duration of a person within the alarm range at any viewing angle is greater than the preset alarm duration, so that an abnormal situation can be quickly responded to and problems can be discovered in time.

[0015] In an optional embodiment, there are multiple on-board cameras, and the multiple on-board cameras include a front camera, a rear camera, a left camera and a right camera; determining whether to trigger an alarm based on the updated status information of the person also includes: determining a first time duration based on the updated status information of the person, wherein the first time duration is the cumulative value of the maximum value in the time information of at least one person collected by the front camera, the maximum value in the time information of at least one person collected by the rear camera, the maximum value in the time information of at least one person collected by the left camera, and the maximum value in the time information of at least one person collected by the right camera; triggering an alarm when the first time duration is greater than or equal to the preset alarm duration; or, the status information also includes indication information of whether the person is missing; triggering an alarm when the status information of the person collected by the front camera or the rear camera indicates lost, and the current images corresponding to the left camera and the right camera do not detect the person; or triggering an alarm when the status information of the person collected by the left camera and the right camera both indicate lost, and the current images corresponding to the front camera or the rear camera do not detect the person.

[0016] In this embodiment, the comprehensive determination of whether to trigger an alarm based on multiple perspectives provides a more comprehensive understanding of human activity across the entire alarm range, more accurately reflecting the overall state of human activity from different perspectives and reducing the likelihood of missed alerts. Monitoring blind spots between different perspectives effectively compensates for the shortcomings of single- and multi-perspective alarms in detecting blind spots, reducing the possibility of individuals exploiting blind spots to hide or engage in suspicious activity, and further enhancing vehicle safety.

[0017] In an optional embodiment, the alarm range includes a first alarm zone set around the vehicle body and a second alarm zone set around the first alarm zone, and the time information includes the duration and / or continuous duration of the person in the first alarm zone, the duration and / or continuous duration of the person in the second alarm zone, the timestamp of the person's first entry into the first alarm zone, and the timestamp of the person's first entry into the second alarm zone.

[0018] In this embodiment, the alarm range is divided into two areas, and computing resources can be reasonably allocated according to the importance and activity frequency of different areas.

[0019] In a second aspect, the present invention provides a vehicle alarm device, comprising: a determination module for determining whether an object is an abnormal object based on the monitoring results of a radar; an acquisition module for acquiring a current image captured by a vehicle-mounted camera when the object is an abnormal object; a detection module for performing personnel detection on the current image to obtain a detection result, the detection result including the current location information of at least one person; and an alarm module for determining whether to trigger an alarm based on the detection result.

[0020] In an optional embodiment, the determination module includes: a first determination unit, used to determine the degree of fluctuation of a first distance within the detection time period based on the monitoring results within the detection time period, wherein the first distance is the distance between the object and the radar; and a second determination unit, used to determine whether the object is an abnormal object based on the degree of fluctuation of the first distance within the detection time period.

[0021] In an optional embodiment, the second determination unit includes: a first determination subunit, configured to determine the object as an abnormal object when the fluctuation degree of the first distance is greater than or equal to a preset fluctuation value and the maximum first distance is less than or equal to a preset distance.

[0022] In an optional embodiment, the device also includes: a grayscale processing module, which is used to perform grayscale processing on the current image to obtain a grayscale image; a brightness statistics module, which is used to perform mean processing on the grayscale image to determine the brightness value; the detection module includes: a processing unit, which is used to process the current image when the brightness value is greater than or equal to the brightness threshold to obtain a detection result.

[0023] In an optional implementation, the acquisition module includes: an acquisition unit, configured to acquire a current image of the vehicle-mounted camera by extracting frames.

[0024] In an optional embodiment, the alarm module includes: an updating unit for updating the status information of the person based on the detection results, the status information including time information reflecting that the person is within the alarm range of the vehicle; a fourth determination unit for determining whether to trigger an alarm based on the updated status information of the person.

[0025] In an optional implementation, the fourth determining unit includes: a first triggering unit, configured to trigger an alarm when the time information corresponding to any one person is greater than a preset alarm duration.

[0026] In an optional embodiment, there are multiple on-board cameras, and the multiple on-board cameras include a front camera, a rear camera, a left camera and a right camera; the alarm module also includes: a fifth determination unit, which is used to determine a first time duration based on the updated status information of the person, wherein the first time duration is the cumulative value of the maximum value of the time information of at least one person collected by the front camera, the maximum value of the time information of at least one person collected by the rear camera, the maximum value of the time information of at least one person collected by the left camera, and the maximum value of the time information of at least one person collected by the right camera; a second trigger unit, which is used to trigger an alarm when the first time duration is greater than or equal to the preset alarm time duration; or, the status information also includes indication information of whether the person is missing; a third trigger unit, which is used to trigger an alarm when the status information of the person collected by the front camera or the rear camera indicates that the person is lost, and the current images corresponding to the left camera and the right camera do not detect the person; or a fourth trigger unit, which is used to trigger an alarm when the status information of the person collected by the left camera and the right camera both indicate that the person is lost, and the current images corresponding to the front camera or the rear camera do not detect the person.

[0027] In an optional embodiment, the alarm range includes a first alarm zone set around the vehicle body and a second alarm zone set around the first alarm zone, and the time information includes the duration and / or continuous duration of the person in the first alarm zone, the duration and / or continuous duration of the person in the second alarm zone, the timestamp of the person's first entry into the first alarm zone, and the timestamp of the person's first entry into the second alarm zone.

[0028] In a third aspect, the present invention provides a vehicle comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the vehicle alarm method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0029] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a vehicle to execute the vehicle alarm method of the first aspect or any corresponding embodiment thereof.

[0030] In a fifth aspect, the present invention provides a computer program product comprising computer instructions for causing a vehicle to execute the vehicle alarm method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in related technologies, the following briefly introduces the drawings required for use in the specific embodiments or related technical descriptions. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 is a schematic diagram of a radar installation position in a vehicle according to an embodiment of the present invention;

[0033] Figure 2 is a flow chart of a vehicle alarm method according to an embodiment of the present invention;

[0034] Figure 3 is a schematic diagram of a process flow of key points of a human body according to an embodiment of the present invention;

[0035] Figure 4 is a flow chart of another vehicle alarm method according to an embodiment of the present invention;

[0036] Figure 5 is a graph showing a change in a first distance within a first detection time period according to an embodiment of the present invention;

[0037] Figure 6 is a graph showing changes in the first distance within a second detection time period according to an embodiment of the present invention;

[0038] Figure 7 is a graph showing changes in the first distance within a third detection time period according to an embodiment of the present invention;

[0039] Figure 8 is a graph showing changes in the first distance within a fourth detection time period according to an embodiment of the present invention;

[0040] Figure 9 is a schematic diagram of an alarm range according to an embodiment of the present invention;

[0041] Figure 10 is a flow chart of another vehicle alarm method according to an embodiment of the present invention;

[0042] Figure 11 This is a schematic diagram of a specific flow chart of a vehicle alarm method according to an embodiment of the present invention;

[0043] Figure 12 is a structural block diagram of a vehicle alarm device according to an embodiment of the present invention;

[0044] Figure 13 4 is a schematic diagram of the hardware structure of a vehicle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. According to the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of the present invention.

[0046] The present invention provides a vehicle alarm method, which first uses the monitoring results of the radar to determine whether the object is an abnormal object. If the object is an abnormal object, the current image is then subjected to personnel detection to determine whether there are any people engaging in suspicious activities around the vehicle body. On the one hand, the running time of detection algorithms such as deep learning algorithms can be reduced, thereby reducing power consumption. On the other hand, dual detection of the monitoring results and the current image is performed to determine whether there are any people engaging in suspicious activities around the vehicle body, which can reduce the possibility of false detection and improve the accuracy of the alarm.

[0047] The vehicle alarm method provided by the present invention is applied to a vehicle, which includes a vehicle-mounted camera and a radar arranged on the vehicle body. The vehicle-mounted camera can be a surround-view camera, etc., and the radar can be an ultrasonic radar or a millimeter-wave radar, etc.

[0048] The present invention does not limit the number of on-board cameras and radars. There can be one or more on-board cameras, and one or more radars. For example, a vehicle can be equipped with four on-board cameras, which can be located at the front, rear, left, and right of the vehicle body, respectively. For ease of distinction, the present invention refers to the on-board camera located at the front of the vehicle body as the front camera, the on-board camera located at the rear of the vehicle body as the rear camera, the on-board camera located on the left of the vehicle body as the left camera, and the on-board camera located on the right of the vehicle body as the right camera.

[0049] like Figure 1 As shown, the vehicle 100 may include 12 radars 110, of which 4 radars may be arranged in the front of the vehicle body, 4 radars may be arranged in the rear of the vehicle body, 2 radars may be arranged in the left of the vehicle body, and 2 radars may be arranged in the right of the vehicle body. The radars located in the front and rear of the vehicle body may be arranged at intervals in the horizontal direction, and the radars located in the left and right of the vehicle body may be arranged at intervals in the vertical direction. Figure 1 This is just one example of how the radar can be set up, and is not limited to this. Figure 1 The dotted area represents the monitoring range of the radar 110.

[0050] The vehicle alarm method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0051] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a vehicle controller such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0052] In this embodiment, a vehicle alarm method is provided, which can be used in a vehicle controller. Figure 2 FIG. 1 is a flow chart of a vehicle alarm method according to an embodiment of the present invention. Figure 2 As shown, the method includes the following steps:

[0053] Step S201: Determine whether the object is an abnormal object based on the monitoring result of the radar.

[0054] Specifically, when the vehicle is in sentry mode, the radar on the vehicle can continuously scan the vehicle's surrounding environment to determine relevant information about objects appearing around the vehicle body (such as the distance, speed or movement trajectory of the object, etc.). The relevant information about the objects around the vehicle body detected by the radar is the radar monitoring result.

[0055] Objects can be people, electric vehicles, animals (such as cats or dogs), or obstacles (such as roadblocks or railings). Abnormal objects can refer to objects that do not conform to the normal surrounding environment characteristics or behavior patterns of the vehicle, such as moving objects that stay within the vehicle's alarm range for a long time.

[0056] The alarm range can be a preset area centered on the vehicle. The preset area can be circular or rectangular, etc. The alarm range can be configured by the designer according to needs. For example, the alarm range can be a circle center area centered on the centroid of the vehicle body with a radius of 2m.

[0057] For example, when the radar is an ultrasonic radar, the radar can determine the monitoring result by the time between transmitting and receiving the ultrasonic wave. Taking the monitoring result as distance as an example, the change in the distance between the object and the vehicle body can be continuously monitored using the following formula (1):

[0058]

[0059] In formula (1), s represents the distance between the vehicle and the object, t represents the time difference between the emission and reception of ultrasonic waves, and v represents the propagation speed of ultrasonic waves, which propagate through air at approximately 340 meters per second (m / s). By measuring the time difference between the emission and reception of ultrasonic echo signals, the distance between the vehicle-mounted radar and the object can be determined.

[0060] For example, when the radar is a millimeter wave radar (such as 30 gigahertz (GHz) to 300 GHz), the radar can determine the monitoring results (distance and / or speed, etc.) by emitting millimeter wave signals and receiving echoes reflected by objects, and analyzing information such as the time delay and frequency change of the echoes, and then determine whether the object is an abnormal object.

[0061] In some embodiments, whether an object is an abnormal object may be determined based on distance and speed. For example, if the distance reflected by the monitoring result is less than a preset distance and the speed is not 0, the object is considered to be an abnormal object.

[0062] It should be understood that when there are multiple radars, the process of the vehicle controller determining whether an object is an abnormal object based on the radar's monitoring results remains unchanged, but the monitoring range of each radar is different due to its different installation location.

[0063] Step S202: When the object is an abnormal object, a current image captured by the vehicle-mounted camera is obtained.

[0064] Specifically, when it is determined that the object is an abnormal object, the vehicle-mounted camera is started to obtain the image captured by the vehicle-mounted camera at the current moment (i.e., the current image); when it is determined that the object is not an abnormal object, the vehicle controller continues to obtain the monitoring results of the radar to determine whether the object is an abnormal object.

[0065] The onboard camera and radar have the same monitoring area. If there are multiple onboard cameras and radars, the onboard camera with the same monitoring area as the radar can be determined based on the radar's position in the acquired monitoring results. In this case, if the object is an abnormal object, the current image captured by the onboard camera corresponding to the radar is obtained.

[0066] Step S203: Perform person detection on the current image to obtain a detection result.

[0067] The detection result includes the current location information of at least one person. It should be understood that there may be no one around the vehicle at the current moment, so the detection result also includes the situation where there is no one.

[0068] Exemplarily, a target detection algorithm and a posture estimation algorithm may be used to perform person detection on the current image to obtain a detection result.

[0069] Specifically, the target detection algorithm is used to identify whether there is a person in the current image. If there is a person, a detection frame can be used to identify the person's position; the posture estimation algorithm is used to estimate the key points of the person in the detection frame and determine the coordinates of the person (i.e., current position information) based on the key points. For example, the key points of a person can be as follows: Figure 3As shown, the estimated key points of a person can be 17, namely nose 0, right eye 1, left eye 2, right ear 3, left ear 4, right shoulder 5, left shoulder 6, right elbow 7, left elbow 8, right wrist 9, left wrist 10, right hip 11, left hip 12, right knee 13, left knee 14, right ankle 15 and left ankle 16.

[0070] The target detection algorithm may be a conventional target detection algorithm in the art, and the posture estimation algorithm may also be a conventional posture estimation algorithm in the art, which will not be described in detail here.

[0071] Step S204: Determine whether to trigger an alarm based on the detection result.

[0072] Specifically, after determining the detection result, the time information of the person within the alarm range can be determined, and then the time information can be used to determine whether to trigger the alarm. For example, if the time information exceeds a preset time length, the alarm is triggered. The preset time length can be configured by the designer according to actual needs, for example, the preset time length can be 2 minutes (min) or 3 minutes, etc.

[0073] For example, triggering an alarm can specifically involve the vehicle controller controlling a sound device (such as a horn) on the vehicle to emit a continuous or intermittent honking sound to attract attention within a certain range and alert suspicious individuals. Triggering an alarm can also involve the vehicle controller controlling a lighting device such as the vehicle's headlights or hazard lights to flash, attracting attention from nearby personnel and alerting suspicious individuals. Triggering an alarm can also involve the vehicle controller pushing an alarm notification message to the owner's smart terminal (such as a mobile phone or smartwatch), which can include the alarm time and a description of the abnormal situation (such as "It is detected that a person has stayed around the vehicle for too long, which may be an abnormality"). Triggering an alarm can also be a combination of at least two of the above methods, and is not limited to this.

[0074] The vehicle alarm method provided in this embodiment first determines whether an object is an abnormal object based on radar monitoring results. If the object is an abnormal object, it then obtains the current image captured by the vehicle camera, then performs personnel detection on the current image, and determines whether to trigger an alarm based on the obtained detection results. This embodiment first uses the radar monitoring results to determine whether the object is an abnormal object, and if the object is an abnormal object, it then performs personnel detection on the current image to determine the detection result. On the one hand, it can reduce the running time of detection algorithms such as deep learning algorithms, thereby reducing power consumption. On the other hand, by performing dual detection of monitoring results and the current image, it can reduce the possibility of false detection and improve the accuracy of the alarm.

[0075] In this embodiment, another vehicle alarm method is provided, which can be used for the above-mentioned vehicle controller. Figure 4 FIG. 1 is a flow chart of another vehicle alarm method according to an embodiment of the present invention. Figure 4As shown, the method includes the following steps:

[0076] Step S401: Determine whether the object is an abnormal object based on the radar monitoring result.

[0077] Specifically, the above step S401 includes:

[0078] Step S4011: Determine the degree of fluctuation of the first distance within the detection time period according to the monitoring result within the detection time period.

[0079] The first distance is the distance between the object and the radar. The detection period can be a preset value, set by the designer based on requirements. For example, the detection period can be 5 seconds or 10 seconds. The detection period can also be non-fixed. For example, the detection period can be the length of time the radar monitors the vehicle's surroundings. For example, if the radar is activated at 8:00, and the current time is 8:03, the detection period can be 30 seconds; if the current time is 8:01, the detection period can be 1 minute.

[0080] Specifically, taking the monitoring result of the first distance as an example, after determining multiple first distances within the detection time period, the variance can be calculated to reflect the degree of fluctuation of the first distance within the detection time period. The variance is proportional to the degree of fluctuation of the first distance. A larger variance indicates a greater degree of fluctuation of the first distance, and a smaller variance indicates a smaller degree of fluctuation of the first distance.

[0081] In some embodiments, after determining multiple first distances within the detection time period, the maximum value D of the first distance within the detection time period may also be used. max and minimum value D min , calculate the range R1(D max -D min ), the range reflects the degree of fluctuation of the first distance during the detection period. The range is directly proportional to the degree of fluctuation of the first distance. A larger range indicates a larger range of distance fluctuation and a greater degree of fluctuation, while a smaller range indicates a smaller range of distance fluctuation and a smaller degree of fluctuation.

[0082] Step S4012: Determine whether the object is an abnormal object based on the fluctuation degree of the first distance within the detection time period.

[0083] For example, when the fluctuation degree of the first distance is greater than or equal to a preset fluctuation value, and the maximum first distance is less than or equal to the preset distance, the object is determined to be an abnormal object. When the fluctuation degree of the first distance is less than the preset fluctuation value and / or the maximum first distance is greater than the preset distance, the object is determined to be a normal object.

[0084] The preset fluctuation value can be determined by the designer based on experience, and the preset distance is determined based on the vehicle's alarm range. For example, if the vehicle's alarm range is a circular area with a radius of 2 meters, the preset distance can be 1.8 meters or 2 meters. The preset distance can be used to indicate the boundary of the vehicle's alarm range, i.e., the alert line.

[0085] Specifically, the fluctuation degree of the first distance can reflect whether the object is in a moving state. When the fluctuation degree is large (greater than the preset fluctuation value), it indicates that the object is a moving object, and the maximum first distance within the detection time period can reflect whether the object is within the alarm range of the vehicle.

[0086] For example, if the variation curve of the first distance in the detection time period is as follows Figure 5 As shown, by calculating the variance, it can be determined that the fluctuation of the distance curve is very small (smoother), and the distance is outside the alarm range (the minimum first distance is greater than the preset distance). At this time, it is considered that the object is an object at a fixed position outside the alarm range (such as a roadblock or a parked electric vehicle, etc.), and the object is determined to be a normal object.

[0087] If the variation curve of the first distance in the detection time period is as follows Figure 6 As shown, it can be seen that the fluctuation of the distance curve is very small, but the distance is within the alarm range (the maximum first distance is less than or equal to the preset distance). At this time, the object is considered to be an object at a fixed position within the alarm range (such as a roadblock, a parked electric vehicle or a wall, etc.), and the object is determined to be a normal object.

[0088] If the variation curve of the first distance in the detection time period is as follows Figure 7 As shown, by calculating the variance, it can be determined that the fluctuation of the distance curve is large, but the distance is outside the alarm range. At this time, the object is considered to be a moving object outside the alarm range. Since it is not within the alarm range, the object is also determined to be a normal object.

[0089] If the variation curve of the first distance in the detection time period is as follows Figure 8 As shown, by calculating the variance, it can be determined that the fluctuation of the distance curve is large and the distance is within the alarm range. At this time, it is considered that the object is a moving object within the alarm range and is relatively suspicious, so the object is determined to be an abnormal object.

[0090] Step S402: When the object is an abnormal object, a current image captured by the vehicle-mounted camera is obtained.

[0091] For details, please see Figure 2 Step S202 of the illustrated embodiment will not be described in detail here.

[0092] Step S403: Perform person detection on the current image to obtain a detection result.

[0093] For details, please see Figure 2 Step S203 of the illustrated embodiment will not be described in detail here.

[0094] Step S404: Determine whether to trigger an alarm based on the detection result.

[0095] Specifically, the above step S404 may include:

[0096] Step S4041: Update the personnel status information according to the detection result.

[0097] The status information includes time information reflecting when a person is within the alarm range of the vehicle.

[0098] Exemplarily, the status information may include the duration and / or continuous duration that a person is within the alarm range of the vehicle. The duration may refer to the accumulated value of the duration that a person is within the alarm range of the vehicle, and the continuous duration may refer to the duration that a person is within the alarm range of the vehicle at one time.

[0099] For example, if a person enters the vehicle's alarm range at 8:01, the person is detected in the current image until 8:02, when the person disappears. Then, the person is detected again in the current image at 8:05, and the person is detected again in the current image until 8:08. At this time, the duration of the person's stay in the alarm range is 4 minutes = (2 minutes - 1 minute) + (8 minutes - 5 minutes). The continuous duration of the person's stay in the alarm range can be 1 minute (the duration between 8:02 and 8:01) and 3 minutes (the duration between 8:08 and 8:05).

[0100] Specifically, when the detection result includes the current location information of the person, it can be determined whether the person is within the alarm range of the vehicle based on the current location information of the person, and the status information of the person can be updated based on the information of whether the person is within the alarm range of the vehicle and the acquisition time of the current image.

[0101] In some examples, the status information may also include information indicating whether a person is missing. If the detection result is that no one is present, the updated status information may also include information indicating that the person is missing and the time when the person is missing.

[0102] Step S4042: Determine whether to trigger an alarm based on the updated status information of the personnel.

[0103] Specifically, after determining the person's updated status information, if the updated status information meets the alarm conditions, an alarm is triggered. If the updated status information does not meet the alarm conditions, the current image is reacquired and steps S403 and S404 are executed. If the updated status information still does not meet the alarm conditions after the target duration, the person's status information is cleared, radar monitoring is resumed, and steps S401 to S404 are executed. The target duration can be determined by the designer based on requirements, for example, the target duration can be 30 minutes or 40 minutes.

[0104] In some embodiments, the alarm condition can be a single-view alarm, that is, when the time information corresponding to any person (such as the duration and / or the continuous duration) exceeds the preset alarm duration, the alarm is triggered. The preset alarm duration can be determined by the designer based on actual needs. For example, the preset alarm duration can be 2 minutes or 3 minutes.

[0105] In this embodiment, an alarm is triggered when the duration and / or continuous duration of a person within the alarm range at any viewing angle is greater than the preset alarm duration, so that an abnormal situation can be quickly responded to and problems can be discovered in time.

[0106] In other embodiments, there are multiple onboard cameras, including a front camera, a rear camera, a left camera, and a right camera. The alarm condition may be a multi-view alarm. In this case, the above step S4042 may include step a1 and step a2:

[0107] Step a1: Determine a first duration based on the updated status information of the personnel.

[0108] Among them, the first time length is the cumulative value of the maximum value in the time information of at least one person collected by the front camera, the maximum value in the time information of at least one person collected by the rear camera, the maximum value in the time information of at least one person collected by the left camera, and the maximum value in the time information of at least one person collected by the right camera.

[0109] For example, if the time information of two people is obtained based on the current image captured by the front camera, the duration of one person is 2 minutes, and the duration of the other person is 1 minute; the time information of one person is obtained based on the current image captured by the rear camera, and the duration of this person is 1 minute; the time information of three people is obtained based on the current image captured by the left camera, the duration of the first person is 0.5 minutes, the duration of the second person is 1 minute, and the duration of the third person is 0.5 minutes; the time information of one person is obtained based on the current image captured by the right camera, and the duration of this person is 0.5 minutes, then the first duration is 4.5 minutes = max{2 minutes, 1 minute} + 1 minute + max{0.5 minutes, 1 minute, 0.5 minutes} + 0.5 minutes.

[0110] Step a2: triggering an alarm when the first duration is greater than or equal to a preset alarm duration.

[0111] Specifically, the maximum value of the time information (such as duration or continuous duration) corresponding to at least one person in each viewing angle (on-board camera) is accumulated. When the accumulated duration exceeds the preset alarm duration, an alarm is triggered. Wherein, at least one viewing angle can detect a person.

[0112] In this embodiment, whether to trigger an alarm is determined comprehensively based on multiple perspectives, which can more comprehensively grasp the activities of people within the entire alarm range, more accurately reflect the overall activity status of people at different perspectives, identify abnormal situations where the target person stays for a short time at each perspective but wanders around the vehicle body, and reduce the possibility of missed reports.

[0113] In some other embodiments, there are multiple on-board cameras, and the status information further includes indication information of whether a person is missing. The indication information may be a string or a symbol. If a string exists, it indicates that a person is missing. The alarm condition may be a blind spot alarm. In this case, the above step S4042 may further include step b1 or step b2:

[0114] Step b1: When the status information of a person collected by the front camera or the rear camera indicates that the person is lost, and no person is detected in the current images corresponding to the left camera and the right camera, an alarm is triggered.

[0115] Step b2: When the status information of the person collected by the left camera and the right camera both indicate that the person is lost, and the current image corresponding to the front camera or the rear camera does not detect the person, an alarm is triggered.

[0116] In this embodiment, monitoring the blind spots between different perspectives can effectively make up for the shortcomings of single-perspective alarms and multi-perspective alarms in monitoring blind spots, reduce the possibility of people using blind spots to hide or conduct suspicious activities, and further improve the safety of the vehicle.

[0117] It should be understood that the left camera is generally located on the left rearview mirror, and the right camera is generally located on the right rearview mirror. When the vehicle is parked (in a parking space or has been parked for longer than a preset time), the left and right rearview mirrors are retracted, and blind spots may exist at corresponding locations in the left and right front of the vehicle. Accordingly, when the left and right cameras are located in other locations, blind spots may also exist at corresponding locations in the left and right rear of the vehicle.

[0118] In some examples, such as Figure 9 As shown, the alarm range includes a first alarm area 901 set around the vehicle body and a second alarm area 902 set around the first alarm area 901, and the time information includes the duration and / or continuous duration of the person in the first alarm area 901, the duration and / or continuous duration of the person in the second alarm area 902, the timestamp of the person's first entry into the first alarm area 901, and the timestamp of the person's first entry into the second alarm area 902.

[0119] Specifically, the range of the first alarm zone 901 and the second alarm zone 902 can be configured by the designer according to needs. For example, the first alarm zone can be within 1 meter from the vehicle body, and the first alarm zone is a high-risk area; the second alarm zone can be within 3 meters from the vehicle body, and the second alarm zone can be a low-risk area.

[0120] In this embodiment, the alarm range is divided into two areas, and computing resources can be reasonably allocated according to the importance and activity frequency of different areas.

[0121] In this embodiment, after obtaining the monitoring results of the radar, the fluctuation degree of the first distance within the detection time period is determined based on the monitoring results within the detection time period, and then based on the fluctuation degree of the first distance within the detection time period, it is determined whether the object is an abnormal object. Based on the fluctuation degree of the first distance, it can be more convenient and efficient to determine whether the monitored object is an abnormal object.

[0122] In this embodiment, another vehicle alarm method is provided, which can be used in a vehicle controller. Figure 10 FIG. 1 is a flow chart of another vehicle alarm method according to an embodiment of the present invention. Figure 10 As shown, the method includes the following steps:

[0123] Step S1001: Determine whether the object is an abnormal object based on the monitoring result of the radar.

[0124] For details, please see Figure 4 Step S401 of the illustrated embodiment will not be described in detail here.

[0125] Step S1002: When the object is an abnormal object, a current image of the vehicle-mounted camera is obtained by extracting frames.

[0126] Specifically, a video stream can be acquired from an onboard camera. An image stream can then be obtained from the real-time video stream. A frame can then be extracted from the image stream as the current image. The current image and the previous image do not need to be consecutive. For example, one frame can be extracted every three frames as the current image. Because the difference between any two frames is small, the frame rate can be adjusted appropriately.

[0127] Step S1003: grayscale processing is performed on the current image to obtain a grayscale image.

[0128] Specifically, performing grayscale processing on the current image is a process of converting a color image into a grayscale image. The grayscale processing method may be a weighted average method, a maximum method, a minimum method, or an average method.

[0129] The weighted average method assigns different weights to the red (R), green (G), and blue (B) color channels based on the human eye's varying sensitivity to different colors. The grayscale value of each pixel is then calculated based on the pixel values ​​and weights of the three channels. The maximum method uses the maximum value of the three color channels for each pixel in a color image as the grayscale value of that pixel. The minimum method uses the minimum value of the three color channels for each pixel in a color image as the grayscale value of that pixel. The average method uses the average of the three color channels for each pixel in a color image as the grayscale value of that pixel.

[0130] In this embodiment, grayscale processing is performed on the image to avoid an increase in the overall average brightness of the image due to a certain area being too bright.

[0131] Step S1004: perform mean processing on the grayscale image to determine the brightness value.

[0132] Step S1005: When the brightness value is greater than or equal to the brightness threshold, perform person detection on the current image to obtain a detection result.

[0133] The detection result includes current location information of at least one person.

[0134] Specifically, when the brightness value is less than a brightness threshold, the current image is reacquired. The brightness threshold can be set by the designer. If the brightness value is less than the brightness threshold, it means that the current image is too dark and it is difficult to detect whether there is a person in the current image.

[0135] Specifically, the above step S1005 includes:

[0136] Step c1: Process the current image to obtain coordinate information of two boundary points of the bounding box and coordinate information of key points of the human body.

[0137] Specifically, the current image may be input into a posture estimation algorithm model, and the current image may be processed by the posture estimation algorithm model to obtain coordinate information of two boundary points of the bounding box and coordinate information of key points of the human body.

[0138] For example, the bounding box may be a rectangle. In this case, the two boundary points of the bounding box may be the boundary point at the upper left corner and the boundary point at the lower right corner. The coordinate information of the two boundary points of the bounding box may be Box{X min ,Y min ,X max ,Y max}, the key points of the human body can be Figure 3 As shown, the coordinate information of the key points of the human body can be Keypoints{[X0,Y0],[X1,Y1],…,[X 16 ,Y 16 ]}.

[0139] Step c2: determining the detection result based on the coordinate information of the two boundary points of the bounding box and the coordinate information of the key points of the human body.

[0140] Specifically, after processing the current image, if the coordinate information of the two boundary points of the bounding box and the coordinate information of the key points of the human body are obtained, it can be determined that there is a person in the current image and the current position information of the person. At this time, the detection result can be the current position information of at least one person; if the coordinate information of the two boundary points of the bounding box and the coordinate information of the key points of the human body are not obtained, the detection result can be that there is no person.

[0141] For example, when a person exists, the center of the two ankles (right ankle 15 and left ankle 16) can be used as the current position information K{K x ,K y}.

[0142] In some embodiments, before performing personnel detection on the current image and obtaining the detection result, the vehicle alarm method further includes: training a deep learning model based on a database to obtain a posture estimation algorithm model.

[0143] Specifically, when training deep learning models, data augmentation (including image stitching, random scaling, color space transformation, and random flipping) can be used to increase the diversity of training samples in the database and improve the robustness of the pose estimation algorithm model.

[0144] When training the deep learning model, the output of the deep learning model takes into account both the bounding box and key points of the human body, and the loss function is modified structurally. Specifically, the prediction of the bounding box is modified from predicting the center point and width and height of the bounding box to predicting the coordinate information of the two boundary points (upper left corner and lower right corner) of the bounding box, thereby improving the model accuracy.

[0145] Taking into account the running speed of the vehicle controller, the number of 17 key points output by the deep learning model is modified, and the midpoint of the two ankles is used as the current position information K{K x ,K y}, that is, only one key point is output. At the same time, in the decoding process of the bounding box and key point information, the decoding process is integrated into the convolution, which greatly improves the running speed.

[0146] Exemplarily, the deep learning model is an end-to-end algorithm model. Compared with the traditional detection algorithm model, the end-to-end algorithm model directly outputs the bounding box and human posture key points, which not only improves the running speed but also outputs the detection box and human posture key points.

[0147] Step S1006: Determine whether to trigger an alarm based on the detection result.

[0148] For details, please see Figure 4 Step S404 of the illustrated embodiment will not be described in detail here.

[0149] In this embodiment, the current image is acquired by extracting frames and the brightness of the current image is filtered before performing personnel detection on the current image, so that the operating power consumption of the posture estimation algorithm can be reduced.

[0150] The following combination Figure 11 , the vehicle alarm method provided by the present invention is described in detail with specific examples.

[0151] like Figure 11 As shown, when the vehicle is in sentry mode, the radar is turned on to monitor the environment around the vehicle. The vehicle controller can determine whether the object is an abnormal object based on the radar's monitoring results. If the object is not an abnormal object, the radar continues to monitor the environment around the vehicle.

[0152] If the object is an abnormal object, a video stream is obtained from the vehicle camera, and an image stream is obtained from the video stream. Then, the current image is obtained by extracting frames. Then, in order to further reduce power consumption, the brightness of the current image is filtered. When the brightness is low (the brightness value is less than the brightness threshold), the current image is obtained again from the image stream.

[0153] When the brightness is greater than or equal to the brightness threshold, the current image is input into the end-to-end algorithm model (posture estimation algorithm model), the detection result is determined according to the output of the end-to-end algorithm model, and whether there is an abnormal situation of the person is determined based on the detection result, and an alarm is triggered when there is an abnormal situation of the person.

[0154] When there is no abnormality among the personnel, determine whether there has been no alarm for a long time (the duration without alarm is greater than the first duration). If there has been no alarm for a long time, re-control the radar to monitor the environment around the vehicle body; otherwise, re-acquire the current image.

[0155] In this embodiment, a vehicle alarm device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments. Details already described will not be repeated here. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0156] This embodiment provides a vehicle alarm device, such as Figure 12 Shown, including:

[0157] The determination module 1201 is used to determine whether the object is an abnormal object based on the monitoring result of the radar;

[0158] An acquisition module 1202 is used to acquire a current image captured by the vehicle-mounted camera when the object is an abnormal object;

[0159] A detection module 1203 is configured to perform person detection on the current image and obtain a detection result, wherein the detection result includes current location information of at least one person;

[0160] The alarm module 1204 is used to determine whether to trigger an alarm based on the detection result.

[0161] In some optional implementations, the determining module 1201 includes:

[0162] a first determining unit, configured to determine a degree of fluctuation of a first distance within the detection time period based on a monitoring result within the detection time period, wherein the first distance is a distance between the object and the radar;

[0163] The second determining unit is configured to determine whether the object is an abnormal object according to a fluctuation degree of the first distance within a detection time period.

[0164] In some optional implementations, the second determining unit includes:

[0165] The first determining subunit is configured to determine the object as an abnormal object when the fluctuation degree of the first distance is greater than or equal to a preset fluctuation value and the maximum first distance is less than or equal to a preset distance.

[0166] In some optional embodiments, the device further comprises:

[0167] Grayscale processing module, used for performing grayscale processing on the current image to obtain a grayscale image;

[0168] Brightness statistics module, used to perform mean processing on grayscale images and determine brightness values;

[0169] The detection module 1203 includes:

[0170] The processing unit is used to process the current image to obtain a detection result when the brightness value is greater than or equal to the brightness threshold.

[0171] In some optional implementations, the acquisition module 1202 includes:

[0172] The acquisition unit is used to acquire the current image of the vehicle-mounted camera by extracting frames.

[0173] In some optional implementations, the alarm module 1204 includes:

[0174] An updating unit, configured to update the status information of the person according to the detection result, the status information including time information reflecting that the person is within the alarm range of the vehicle;

[0175] The fourth determining unit is used to determine whether to trigger an alarm based on the updated status information of the personnel.

[0176] In some optional implementations, the fourth determining unit includes:

[0177] The first trigger unit is used to trigger an alarm when the time information corresponding to any person is greater than a preset alarm duration.

[0178] In some optional implementations, there are multiple onboard cameras, including a front camera, a rear camera, a left camera, and a right camera; the alarm module 1204 further includes:

[0179] a fifth determining unit, configured to determine a first duration based on the updated status information of the person, wherein the first duration is a cumulative value of a maximum value among time information of at least one person collected by the front camera, a maximum value among time information of at least one person collected by the rear camera, a maximum value among time information of at least one person collected by the left camera, and a maximum value among time information of at least one person collected by the right camera;

[0180] A second triggering unit is configured to trigger an alarm when the first duration is greater than or equal to a preset alarm duration; or the status information further includes information indicating whether a person is missing;

[0181] A third triggering unit is configured to trigger an alarm when the status information of a person captured by the front camera or the rear camera indicates that the person is lost and no person is detected in the current images corresponding to the left camera and the right camera; or

[0182] The fourth trigger unit is used to trigger an alarm when the status information of the person collected by the left camera and the right camera both indicate that the person is lost, and the current image corresponding to the front camera or the rear camera does not detect the person.

[0183] In some optional embodiments, the alarm range includes a first alarm zone set around the vehicle body and a second alarm zone set around the first alarm zone, and the time information includes the duration and / or continuous duration of the person in the first alarm zone, the duration and / or continuous duration of the person in the second alarm zone, the timestamp of the person's first entry into the first alarm zone, and the timestamp of the person's first entry into the second alarm zone.

[0184] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0185] The vehicle alarm device in this embodiment is presented in the form of a functional unit, where the unit refers to an application specific integrated circuit (ASIC), a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0186] The embodiment of the present invention further provides a vehicle, such as Figure 13 As shown, the vehicle includes: one or more processors 1310, a memory 1320, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses and can be installed on a common motherboard or in other ways as needed. The processor can process instructions executed in the vehicle, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Figure 13 A processor 1310 is taken as an example.

[0187] Processor 1310 may be a central processing unit (CPU), a network processor (NPU), or a combination thereof. Processor 1310 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CPLD), a field programmable gate array (FPGA), a general purpose array logic (GAL), or any combination thereof.

[0188] The memory 1320 stores instructions that can be executed by at least one processor 1310, so as to enable the at least one processor 1310 to execute the method shown in the above embodiment.

[0189] The memory 1320 may include a program storage area and a data storage area. The program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on vehicle usage, etc. Furthermore, the memory 1320 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some optional embodiments, the memory 1320 may optionally include a memory remotely located relative to the processor 1310, and these remote memories may be connected to the vehicle via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0190] The memory 1320 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid state drive; the memory 1320 may also include a combination of the above types of memory.

[0191] The vehicle also includes a communication interface 1330 for the vehicle to communicate with other devices or a communication network.

[0192] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0193] A portion of the present invention may be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the vehicle. Those skilled in the art will appreciate that the computer program instructions may exist in a computer-readable medium in the form of, but are not limited to, source files, executable files, installation package files, and the like. Accordingly, the manner in which the computer program instructions are executed by the vehicle includes, but is not limited to: the vehicle directly executing the instruction, or the vehicle compiling the instruction and then executing the corresponding compiled program, or the vehicle reading and executing the instruction, or the vehicle reading and installing the instruction and then executing the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the vehicle.

[0194] In the description of this specification, the description with reference to the terms "this embodiment", "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0195] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0196] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations shall all fall within the scope defined by the present invention.

Claims

1. A vehicle alarm method, characterized in that: The vehicle includes an on-board camera and a radar disposed on the vehicle body, and the method includes: determining whether the object is an abnormal object based on the monitoring result of the radar; When the object is the abnormal object, obtaining a current image captured by the vehicle-mounted camera; Performing person detection on the current image to obtain a detection result, wherein the detection result includes current location information of at least one person; According to the detection result, it is determined whether to trigger an alarm.

2. The method according to claim 1, characterized in that Determining whether the object is an abnormal object according to the monitoring result of the radar includes: determining, based on monitoring results within a detection time period, a degree of fluctuation of a first distance within the detection time period, wherein the first distance is a distance between the object and the radar; Whether the object is an abnormal object is determined according to the fluctuation degree of the first distance within the detection time period.

3. The method according to claim 2, characterized in that The determining whether the object is an abnormal object according to the fluctuation degree of the first distance within the detection time period includes: When the fluctuation degree of the first distance is greater than or equal to a preset fluctuation value, and the maximum first distance is less than or equal to a preset distance, the object is determined to be an abnormal object.

4. The method according to claim 1, wherein Before performing person detection on the current image to obtain a detection result, the method further includes: Performing grayscale processing on the current image to obtain a grayscale image; Performing mean processing on the grayscale image to determine a brightness value; The performing person detection on the current image to obtain a detection result includes: When the brightness value is greater than or equal to the brightness threshold, the current image is processed to obtain a detection result.

5. The method according to any one of claims 1 to 4, characterized in that The acquiring of the current image of the vehicle-mounted camera includes: The current image of the vehicle-mounted camera is obtained by frame extraction.

6. The method according to any one of claims 1 to 4, characterized in that Determining whether to trigger an alarm based on the detection result includes: updating the status information of the person according to the detection result, wherein the status information includes time information reflecting that the person is within the alarm range of the vehicle; Determine whether to trigger an alarm based on the updated status information of the personnel.

7. The method according to claim 6, characterized in that The determining whether to trigger an alarm according to the updated status information of the personnel includes: When the time information corresponding to any person is greater than the preset alarm duration, an alarm is triggered.

8. The method according to claim 7, characterized in that There are multiple on-board cameras, including a front camera, a rear camera, a left camera, and a right camera; and determining whether to trigger an alarm based on the updated status information of the person further includes: Determining a first duration according to the updated status information of the person, wherein the first duration is a cumulative value of a maximum value among time information of at least one person collected by the front camera, a maximum value among time information of at least one person collected by the rear camera, a maximum value among time information of at least one person collected by the left camera, and a maximum value among time information of at least one person collected by the right camera; When the first duration is greater than or equal to the preset alarm duration, an alarm is triggered; or, the status information further includes information indicating whether the person is missing; When the status information of a person collected by the front camera or the rear camera indicates loss and no person is detected in the current images corresponding to the left camera and the right camera, an alarm is triggered; or, When the status information of the person collected by the left camera and the right camera both indicate that the person is lost, and the current image corresponding to the front camera or the rear camera does not detect the person, an alarm is triggered.

9. The method according to claim 6, characterized in that The alarm range includes a first alarm zone set around the vehicle body and a second alarm zone set around the first alarm zone, and the time information includes the duration and / or continuous duration of the person in the first alarm zone, the duration and / or continuous duration of the person in the second alarm zone, the timestamp of the person's first entry into the first alarm zone, and the timestamp of the person's first entry into the second alarm zone.

10. A vehicle alarm device, characterized in that: The device comprises: A determination module, used to determine whether the object is an abnormal object based on the monitoring results of the radar; an acquisition module, configured to acquire a current image captured by the vehicle-mounted camera when the object is the abnormal object; a detection module, configured to perform person detection on the current image to obtain a detection result, wherein the detection result includes current location information of at least one person; The alarm module is used to determine whether to trigger an alarm based on the detection result.

11. A vehicle, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the vehicle alarm method according to any one of claims 1 to 9 by executing the computer instructions.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a vehicle to execute the vehicle alarm method according to any one of claims 1 to 9.

13. A computer program product, characterized in that The method comprises computer instructions for causing a vehicle to execute the vehicle alarm method according to any one of claims 1 to 9.

Images