Toilet safety monitoring method and device, computer equipment and storage medium

Through the integrated security monitoring system of millimeter wave radar and voice recognition technology, the status of users in the bathroom is monitored in real time and alarmed automatically, the problem of insufficient privacy protection and real-time monitoring is solved, and the safety and equipment of the elderly and those with limited mobility is improved. Intelligent.

CN119028083AInactive Publication Date: 2024-11-26加一感知(深圳)科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411272576.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The prior art is difficult to monitor the physical condition of users in the bathroom in real time without infringing on privacy, and automatically issue an alarm without the need for active user operations to prevent fall accidents and provide timely rescue.

Method used

Using a security monitoring system that integrates millimeter-wave radar technology and voice recognition technology, the user's motion status and attitude information is obtained in real time through the radar equipment, determine whether the dangerous conditions are met, and actively inquire and data through the voice equipment when an abnormal state is detected. Collect, confirm whether the user is in a dangerous state, and generate alarm information.

Benefits of technology

It realizes comprehensive monitoring and automatic alarms for bathroom users, improves the effectiveness and intelligence of monitoring, ensures the safety of the elderly and those with limited mobility, reduces the possibility of false alarms and underreports, and provides timely rescue measures. .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119028083A_ABST
    Figure CN119028083A_ABST
Patent Text Reader

Abstract

The invention relates to a toilet safety monitoring method and device, computer equipment and a storage medium, and aims to solve the problems of privacy protection and real-time monitoring. Posture information of a user is obtained in real time through millimeter wave radar equipment arranged in a toilet, and whether the user is in a falling or long-time static state or not is judged. When abnormity is detected, active voice inquiry is carried out by using voice equipment, and whether a dangerous state exists or not is judged based on voice data. And if danger exists indeed, alarm information is generated. According to the method, real-time monitoring and automatic alarm under privacy protection are effectively realized, and the reliability and timeliness of toilet safety monitoring are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a toilet safety monitoring method, device, computer equipment and storage medium. Background Art

[0002] In the daily lives of the elderly and people with limited mobility, the bathroom is a high-risk area. Due to the slippery floor, narrow space, and the body balance problems these people may face during use, the incidence of falls is high. Therefore, the demand for bathroom safety monitoring equipment is increasing. At present, the existing technical solutions mainly rely on camera monitoring, pull-cord alarms, and SOS button alarms. However, these methods have certain limitations in practical applications and are difficult to fully meet the needs of bathroom safety monitoring.

[0003] Although camera monitoring can capture the user's dynamic information in real time, the privacy of the bathroom greatly limits the application of cameras in this environment. Even if special privacy protection measures are adopted, the user's privacy may still be violated, causing psychological rejection by the user. In addition, the camera monitoring system requires more complicated installation and maintenance, which is costly and difficult to popularize for ordinary families. Therefore, how to ensure the safety of users in the bathroom while protecting privacy has become an important research direction.

[0004] Although pull cord alarms and SOS button alarms can provide simple emergency assistance functions, these methods rely on the user to actively send out a signal for help when an accident occurs. However, in many cases, users who fall or suddenly faint may not be able to trigger these alarm devices in time, thus missing the best time for rescue. In addition, these alarm methods cannot monitor the user's real-time status, lack active alarm functions, and cannot provide timely warnings. Therefore, how to monitor the user's physical condition in real time and automatically issue an alarm without the user's active operation is a problem that needs to be solved urgently in current technology. Summary of the invention

[0005] The main purpose of the present invention is to provide a bathroom safety monitoring method, device, computer equipment and storage medium to solve the problem of how to monitor the physical condition of users in the bathroom in real time without infringing on privacy, and automatically issue an alarm without active operation, so as to effectively prevent falling accidents and provide timely rescue.

[0006] To achieve the above-mentioned purpose, the present invention provides a safety monitoring method for a toilet, comprising the following steps: based on a radar device arranged in the toilet, obtaining a current real-time radar signal, and based on the current real-time radar signal, determining the posture information of the person in the toilet, wherein the person situation information at least includes: the person is in a falling state and the person is in a staying state; when the person situation information is that the person is in a falling state, obtaining the current real-time radar signal again based on the radar device arranged in the toilet, and determining whether the person status in the toilet meets a first preset condition according to the current real-time radar signal, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; in the case where the person situation information is that the person is in a falling state When the information shows that a person is in a resident state, the current real-time radar signal is obtained again based on the radar equipment arranged in the bathroom, and according to the current real-time radar signal, it is determined whether the state of the person in the bathroom meets the second preset condition, wherein the second preset condition at least includes: no action state and exceeds the preset time; when the state of the person in the bathroom meets the first preset condition or the second preset condition, a voice inquiry and voice recording are performed to the person in the bathroom through the voice equipment arranged in the bathroom, and the voice data obtained by the voice recording is obtained, and based on the recorded voice data, it is determined whether the person in the bathroom is in a dangerous state, and when the person in the bathroom is in a dangerous state, an alarm message is generated.

[0007] The present invention also provides a safety monitoring system for a bathroom, the safety monitoring system also includes: a control processor, a voice device, a radar device and a lighting device, the control processor is communicatively connected to the voice device, the radar device and the lighting device, and is used to jointly execute the steps of the above-mentioned method; wherein the safety monitoring system also includes: a wireless communication module, which is communicatively connected to the control processor, and is used to upload the data processed by the control processor to the server; the server side is communicatively connected to the mobile terminal, and is used to store the data content uploaded by the wireless communication module, and upload the alarm information in the data content to the mobile terminal; the mobile terminal is used to check real-time alarm information and historical alarm information.

[0008] The present invention also provides a safety monitoring device for a toilet, the safety monitoring device comprising: an acquisition unit, for acquiring a current real-time radar signal based on a radar device arranged in the toilet, and determining posture information of a person in the toilet based on the current real-time radar signal, wherein the person situation information at least includes: the person is in a falling state and the person is in a staying state; a first determination unit, for acquiring the current real-time radar signal again based on the radar device arranged in the toilet when the person situation information indicates that the person is in a falling state, and determining whether the person status in the toilet meets a first preset condition based on the current real-time radar signal, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; a second determination unit, for When the personnel situation information indicates that the personnel is in a resident state, the current real-time radar signal is obtained again based on the radar device arranged in the bathroom, and according to the current real-time radar signal, it is determined whether the status of the personnel in the bathroom meets the second preset condition, wherein the second preset condition at least includes: no action state and exceeding the preset time; an alarm unit is used to make voice inquiries and voice collection to the personnel in the bathroom through the voice device arranged in the bathroom when the status of the personnel in the bathroom meets the first preset condition or the second preset condition, and obtain the voice data obtained by the voice collection, and determine whether the personnel in the bathroom is in a dangerous state based on the recorded voice data, and generate an alarm message when the personnel in the bathroom is in a dangerous state.

[0009] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0010] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0011] The toilet safety monitoring method, device, computer equipment and storage medium provided by the present invention can realize real-time monitoring and automatic alarm of toilet users without infringing privacy by integrating millimeter wave radar technology and voice recognition technology. This integration method not only improves the effectiveness of monitoring, but also greatly enhances the intelligence level of the system and user experience.

[0012] Specifically, the safety monitoring system of the present invention obtains the user's motion status and posture information in real time through the millimeter wave radar equipment installed in the bathroom. The millimeter wave radar captures the user's limb movements, the amplitude of the micro-movement of the chest during breathing, and the slight changes in the heartbeat by emitting and receiving electromagnetic waves, so as to accurately determine whether the user has fallen or has been stationary for a long time and other abnormal situations. Based on this information, the system can determine whether the preset danger conditions are met, such as whether there is a continuous struggle after falling or a long period of inability to get up, and whether the user is completely stationary for more than a preset time in the resident state.

[0013] When an abnormal state is detected, the system will actively make voice inquiries through the voice equipment installed in the bathroom. This process further confirms whether the user is in danger through the collected voice data. If the voice data indicates that the user is indeed in danger, the system will generate an alarm message and upload the data to the server through the wireless communication module. The server then transmits the alarm information to the mobile terminal, so that relevant personnel can view the alarm information and historical records in real time and take necessary rescue measures in time.

[0014] In summary, the present invention solves the problems of insufficient privacy protection and incomplete real-time monitoring in the prior art by introducing advanced millimeter-wave radar technology and voice recognition technology, and realizes comprehensive monitoring and timely alarm for toilet users. This innovation not only improves the life safety of the elderly and people with limited mobility, but also improves the intelligence and convenience of the equipment, bringing new solutions to the field of toilet safety monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the steps of a toilet safety monitoring method according to one embodiment of the present invention; Figure 2 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 1 ; Figure 3 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 2 ; Figure 4 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 3 ; Figure 5 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 4 ; Figure 6 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 5 ; Figure 7 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 6 ; Figure 8 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 7 ; Fig. 9 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 8 ; Fig.10 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 9 ; Fig.11 This is a schematic diagram of a toilet safety monitoring method according to an embodiment of the present invention. Figure 10 ; Fig.12 This is a structural block diagram of a toilet safety monitoring device in one embodiment of the present invention; Fig.13 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0016] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] Reference Figure 1 The embodiment of the present invention provides a toilet safety monitoring method, comprising the following steps: S1, based on the radar equipment arranged in the toilet, obtaining the current real-time radar signal, and determining the posture information of the person in the toilet based on the current real-time radar signal, wherein the person situation information at least includes: the person is in a falling state and the person is in a staying state; S2, when the personnel situation information indicates that the person is in a falling state, obtaining a current real-time radar signal based on the radar device arranged in the toilet again, and determining whether the status of the person in the toilet meets a first preset condition according to the current real-time radar signal, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; S3, when the personnel situation information indicates that the personnel is in a stationary state, obtaining a current real-time radar signal based on a radar device arranged in the toilet again, and determining whether the status of the personnel in the toilet meets a second preset condition according to the current real-time radar signal, wherein the second preset condition at least includes: no action state and exceeding a preset time; S4, when the status of the person in the bathroom meets the first preset condition or the second preset condition, voice inquiries and voice recordings are made to the person in the bathroom through the voice equipment installed in the bathroom, and voice data obtained by the voice recording is obtained, and based on the recorded voice data, it is determined whether the person in the bathroom is in a dangerous state, and when the person in the bathroom is in a dangerous state, an alarm message is generated.

[0019] In other words, the core is to use radar equipment to replace traditional cameras and manual triggering devices, so as to provide all-round real-time monitoring without infringing the privacy of users. Radar equipment can obtain the posture information of people in the bathroom by emitting electromagnetic waves and receiving reflected signals. This method can not only accurately detect the location information of people, but also identify the specific posture of people, such as whether they are in a falling state or a stationary state. This recognition capability directly solves the privacy problem caused by camera monitoring, because radar signals do not capture specific image information, but analyze information based on electromagnetic wave reflection, which greatly reduces the risk of privacy leakage.

[0020] Furthermore, the technical solution achieves accurate monitoring of the status of people in the bathroom by acquiring and analyzing real-time radar signals multiple times. For example, when the system detects that a person is in a state of falling, it will further monitor whether the person meets the first preset condition, that is, a state of continuous struggle and / or a state of not getting up, and the time exceeds the preset threshold. This design enables the system to not only detect whether a person has fallen, but also to judge the severity and urgency of the fall. If the system determines that a person is in a state of continuous struggle, this may mean that the person tried to stand up but failed, and the system will immediately sound an alarm; if the system detects a state of not getting up, it means that the person may be in a more serious predicament, such as a coma or a fracture, and the system will also sound an alarm.

[0021] In addition, the technical solution also takes into account the status of people staying in the bathroom. By acquiring the real-time radar signal again and analyzing whether it meets the second preset condition (no movement and exceeding the preset time), the system can determine whether the user may be in danger due to physical discomfort. For example, if an elderly person does not move in the bathroom for a long time, it may be because he feels dizzy or is about to faint. The system can issue an alarm by analyzing this state of no movement. In this way, the technical solution can not only respond to emergencies, but also warn of potential risks, effectively improving the safety in the bathroom.

[0022] In order to further improve the system's responsiveness, the technical solution also introduces voice equipment. When the system detects that the person's status meets dangerous conditions, it will ask the user through the voice device and collect their voice feedback. By analyzing the collected voice data, the system can further confirm whether the user is in a dangerous state. For example, if the user is unable to make an effective voice response after falling, the system can be more confident that the user may need emergency assistance, so as to quickly generate an alarm message and notify relevant personnel or equipment.

[0023] For example, if an elderly person slips and falls in the bathroom and is unable to get up, the system will first detect the fall through radar, and after the person has not gotten up for a set time, it will send a query through the voice device. If the elderly person is unable to respond, the system will immediately generate an alarm message to notify family members or medical institutions for rescue. This automated monitoring and alarm mechanism not only improves the response speed of emergencies, but also ensures that help can be provided in a timely manner without relying on the user's active operation.

[0024] In general, this technical solution provides comprehensive monitoring and alarm functions for the status of users in the toilet by combining radar equipment and voice equipment while ensuring privacy. It effectively solves the problems of privacy infringement, active operation dependence and insufficient monitoring coverage in the existing technology, and provides more reliable protection for the safety of toilet use for the elderly and people with limited mobility.

[0025] In one example, before obtaining the current real-time radar signal based on a radar device arranged in the bathroom, and determining the posture information of the person in the bathroom based on the current real-time radar signal, the safety monitoring method also includes: obtaining the current real-time radar signal based on the radar device arranged in the bathroom; filtering and processing the current real-time radar signal moving signal through a preset radar interference background noise filtering model to obtain the motion energy intensity in the bathroom; comparing the motion energy intensity in the bathroom with the energy model threshold range triggered by the person entering the bathroom; if the motion energy intensity in the bathroom meets the threshold range, it is determined that there is a person in the bathroom, the lighting device arranged in the bathroom is controlled to be turned on, and the step of "obtaining the current real-time radar signal based on the radar device arranged in the bathroom, and determining the posture information of the person in the bathroom based on the current real-time radar signal" is executed; if the motion energy intensity in the bathroom meets the threshold range, it is determined that there is no person in the bathroom, and the lighting device arranged in the bathroom is controlled to be turned off.

[0026] In this technical solution, the automatic opening and closing process of lighting equipment and the personnel detection process are added, further optimizing the safety monitoring function in the bathroom. This improvement not only improves the intelligence of the system, but also enhances the convenience of users and energy saving effect.

[0027] First, regarding the on / off process of lighting equipment, the solution uses radar equipment to detect the intensity of motion energy in the bathroom. When a person enters the bathroom, the radar equipment will immediately obtain the current real-time radar signal and filter and process these signals through the preset radar interference background noise filter model. The purpose of this process is to exclude environmental noise and interference signals to ensure that only valid motion signals are extracted. After screening and processing, the system will calculate the intensity of motion energy in the bathroom.

[0028] Next, the system compares the acquired motion energy intensity with the preset energy model threshold range triggered when a person enters the bathroom. If the detected motion energy intensity meets the threshold range, it means that someone has entered the bathroom, and the system will automatically turn on the lighting equipment to provide the user with the required light. This intelligent lighting control not only improves user convenience, but also avoids unnecessary energy consumption because the lighting equipment is only turned on when someone is there.

[0029] Secondly, regarding the personnel detection process, the system uses the same radar signal screening and energy intensity calculation method to accurately determine whether there is someone in the bathroom. Figure 2As shown in the figure, if the motion energy intensity detected by the system does not reach the preset threshold range, it means that there is no one in the bathroom, and the system will automatically turn off the lighting. This automated detection and control process not only effectively avoids energy waste when no one is in the bathroom for a long time, but also ensures that the lighting is always on when someone is using the bathroom, providing a safe and comfortable environment for users.

[0030] In general, this design concept not only improves the intelligence level of the bathroom, but also enhances the accuracy of safety monitoring. Through automated lighting control and personnel detection, the system can better protect the privacy of users, provide efficient energy management, and ensure timely response in dangerous situations. For example, when an elderly person enters the bathroom at night, the system can turn on the lighting immediately after detecting his entry to prevent him from falling in the dark. After confirming that the bathroom is empty, the system can quickly turn off the lighting to save energy. This process design integrates security, privacy protection, energy saving and intelligence, providing more complete life support for the elderly and people with limited mobility.

[0031] In one example, if Figure 3 As shown, based on the radar equipment installed in the bathroom, the current real-time radar signal is obtained; the current real-time radar signal is screened and processed by the preset radar interference background noise filtering model to obtain the motion energy intensity in the bathroom. The method also includes: a simple process of powering on the bathroom fall monitoring device, a physical button trigger, until the device is normally connected to the server, and the LED light status and other behavior status data of the device are reported normally.

[0032] That is, the device is powered on and starts working. The user presses and holds the case for 5 seconds and writes the server information to connect to the server. If the connection fails, the user continues to press and hold the case for 5 seconds and writes the server information. If the connection is successful, the server receives the data, and the network indicator light changes from flashing to off. After that, if there is someone, the light will turn on, and if there is no one, the light will turn off. In this process, data will be reported to the server. The specific reported content includes light status, alarm status, and human behavior status.

[0033] In one example, the current real-time radar signal moving signal is filtered and processed by a preset radar interference background noise filtering model to obtain the motion energy intensity in the bathroom, which can be achieved by the following steps: The current real-time radar signal is segmented into time windows by using the radar interference background noise filtering model, and the real-time radar signal data in each time window is divided into multiple data segments; Performing a short-time Fourier transform on the data segment to convert the time domain signal into a frequency domain signal. At the same time, by analyzing the spectrum density of the data segment, marking the signal with a frequency lower than 0.5 Hz as a background noise signal (this is because the frequency reflected in the radar signal by human micro-movement, breathing, heartbeat, etc. is usually higher, while environmental noise such as low-frequency vibration of the ventilation system is prominent in the low-frequency band); and using a band-stop filter to attenuate the low frequency band in the data segment, retaining only the frequency components above the threshold, thereby removing most of the ambient background noise (the resulting purified radar signal retains key information reflecting human motion and removes most of the irrelevant noise, ensuring the accuracy of subsequent feature extraction); Perform short-time Fourier transform on the retained data segments again to convert the time domain signal into a frequency domain signal and obtain its spectrum characteristics (at this point, frequency domain analysis helps to identify different types of movement patterns. For example, breathing and heartbeats usually appear as periodic signals between 0.1 and 2 Hz, while slight movements of limbs may appear in higher frequency bands); Acquire spectrum energy density based on the spectrum characteristics, and perform feature extraction based on the spectrum energy density to obtain multiple feature data (spectrum energy density can quantify energy contributions of different frequency components, thereby extracting key features related to human body movement); Specifically, the feature extraction process includes: determining specific frequency bands related to human motion. For example, breathing signals are usually in the frequency band of 0.1 to 0.5 Hz, while heartbeat signals are between 1 and 2 Hz. The frequency of limb movements may be higher; extracting various characteristic parameters from the spectrum energy density, such as the peak value, average value, and variance of the energy distribution of the frequency band. For example, the sum of the energy in the frequency band of 0.1 to 0.5 Hz can be calculated to evaluate the intensity of breathing activity; or the energy distribution in the frequency band of 1 to 2 Hz can be calculated to evaluate the characteristics of the heartbeat signal; and the above characteristic parameters are organized into characteristic data. These characteristic data include the energy value of each frequency band, the mean and standard deviation of the spectrum, etc. These characteristic data are helpful for further analysis and judgment of the human motion state. For example, by analyzing the energy changes in the breathing frequency band, it can be judged whether a person is breathing; by the energy intensity of the heartbeat frequency band, it can be judged whether there is a heartbeat signal.

[0034] Normalize the feature data (to ensure that different features are balanced in the calculation. For example, the energy of breathing and heartbeat signals can be normalized to the range of 0 to 1, while limb movement signals may occupy a larger range); The normalized feature data are weighted linearly combined to obtain a motion energy intensity value, and the motion energy intensity value is smoothed to eliminate short-term fluctuations to obtain a stable energy motion intensity trend line, and the presence of a person in the bathroom is judged based on the motion energy intensity trend line.

[0035] In other words, the real-time radar signal is segmented into time windows through the preset radar interference background noise filtering model. The signal data in each time window is divided into multiple data segments. This step is to decompose the signal into smaller parts for subsequent processing and analysis. Next, these data segments are short-time Fourier transformed to convert the time domain signal into a frequency domain signal. In the frequency domain, the spectral density of each data segment is analyzed, and the signal with a frequency below 0.5 Hz is marked as background noise. This is because signals such as human micro-movements, breathing, and heartbeats usually appear as higher frequencies, while low-frequency signals are often environmental noise. These low-frequency signals are processed by a band-stop filter, retaining only the frequency components above the threshold, removing most of the environmental noise, and thus obtaining a purified radar signal.

[0036] Subsequently, the cleaned data segments are short-time Fourier transformed again to obtain spectral characteristics. This step helps to identify different frequency patterns related to human movement. For example, the frequency of breathing and heartbeat signals is usually between 0.1 and 2 Hz, while limb movement may be manifested in a higher frequency band. Based on the spectral characteristics, the spectral energy density is calculated and the characteristic data related to human movement are extracted. These characteristic data are normalized to ensure the balance of each feature in subsequent calculations. For example, the energy of breathing and heartbeat may be normalized to the range of 0 to 1, while the energy range of limb movement signals may be larger.

[0037] Finally, the normalized feature data is weighted linearly combined to obtain the motion energy intensity value. In order to eliminate short-term fluctuations, the motion energy intensity value is further smoothed to obtain a stable energy intensity trend line. This trend line is used to determine whether there are people in the bathroom and their motion status. Through these steps, the motion situation in the bathroom can be accurately evaluated and the status of people can be effectively monitored.

[0038] In one example, a voice device disposed in a toilet is used to conduct a voice inquiry and voice recording of a person in the toilet, and voice data obtained by the voice recording is obtained, and based on the recorded voice data, it is determined whether the person in the toilet is in a dangerous state, and when the person in the toilet is in a dangerous state, an alarm message is generated, including: the voice device disposed in the toilet is used to conduct a voice recording of the person in the toilet to obtain first voice data, and a radar device disposed in the toilet is used to perform an abnormality detection on the person in the toilet to obtain first detection data; based on the first voice data and the first detection data, the voice device is used to generate an alarm message. The person in the bathroom plays the preset safety inquiry voice data to determine whether the person in the bathroom needs help; the voice of the person in the bathroom is recorded again based on the voice device to obtain the second voice data, and when it is determined based on the second voice data that the person in the bathroom does not need help, continue to execute the step of "recording the voice of the person in the bathroom through the voice device set in the bathroom to obtain the first voice data, and performing abnormality detection on the person in the bathroom through the radar device set in the bathroom to obtain the first detection data"; when it is determined based on the second voice data that the person in the bathroom needs help, an alarm message is generated.

[0039] That is, if Figure 4 As shown in the figure, this technical solution provides a more intelligent and comprehensive safety monitoring method by combining voice equipment and radar equipment, which is particularly suitable for the safety monitoring of the elderly and people with limited mobility in the bathroom. The beneficial effect of this solution is that it can not only actively monitor and detect the user's status, but also further confirm the user's safety status through voice interaction, minimize the possibility of false alarms, and can issue alarm information in a timely manner in critical situations.

[0040] First of all, the collaborative work of the voice device and radar device mentioned in the scheme solves some limitations of the existing technology. For example, relying solely on radar equipment to detect the user's posture information or movement status, although some obvious abnormal situations (such as falling or long periods of inactivity) can be identified, in some cases, radar detection may have errors, or the detection results may not be accurate enough due to environmental factors. In order to make up for these possible shortcomings, the scheme uses a voice device to make voice inquiries to the user when a suspected dangerous state is detected to obtain more accurate status feedback. For example, when the radar device detects that the user has been inactive for a long time, the system will issue a safety inquiry through the voice device, such as "Are you okay? Do you need help?" At this time, the user's voice feedback will be used as a basis for further judgment.

[0041] During this process, the voice device will first record the user's first voice data, while the radar device will continue to monitor the user's abnormal state and generate the first detection data. The system will analyze and compare the two data to determine whether the user is in a dangerous state. If the system identifies an abnormality in the voice data, such as a weak tone, slow speech, or inability to answer normally, this may indicate that the user is in poor physical condition, and the system will proceed to the next step of confirmation. At this time, the voice device will play the preset safety inquiry voice to further confirm whether the user needs help. If the user can respond normally and indicates that he does not need help, the system will end the current alarm process and resume monitoring. However, if the user cannot make an effective response, or both the voice data and the radar data point to a possible dangerous state, such as an ambiguous answer, obvious signs of pain in the voice, or no response for a long time, the system will immediately generate an alarm message.

[0042] The beneficial effect of this design is that it greatly improves the safety and reliability of the system. On the one hand, the continuous monitoring of the radar device ensures real-time grasp of the user's status, while the introduction of the voice device enhances the system's ability to handle complex scenarios. Through this double verification mechanism, the system can minimize the probability of false alarms and provide timely assistance in real emergencies. For example, suppose an elderly person suddenly feels unwell in the bathroom. Although he has not fallen, he may not be able to leave the bathroom or press the help button in time due to his weak body. At this time, the system's radar device may detect that he has been stationary for a long time, and the system will immediately issue an inquiry through the voice device. If the elderly person can respond normally and indicates that he is just taking a short break, the system will stop further alarm operations; but if he cannot respond or responds vaguely, the system will issue an alarm message to notify his family or caregivers to take emergency measures.

[0043] In summary, the implementation of this technical solution not only improves the safety of users in the bathroom, but also greatly enhances the intelligence of the system, and can more accurately identify various potential dangerous situations. By combining radar detection and voice interaction, the system can not only capture the user's physical state, but also further confirm the user's real needs through language communication. This multi-level safety monitoring strategy not only reduces false alarms, but also provides more timely and effective assistance in real emergencies, thereby significantly improving the quality of life and safety of the elderly and people with limited mobility in the bathroom.

[0044] In one example, based on the current real-time radar signal, determining the posture information of the person in the bathroom includes: based on the current real-time radar signal, determining the point cloud orientation information detected in the bathroom, wherein the point cloud orientation information includes: X-axis information, Y-axis information and Z-axis information; parsing the point cloud orientation to obtain the posture change information of the person in the bathroom, wherein the posture change information at least includes: person's center of gravity change information, person's lateral area change information, person's longitudinal change acceleration information and person's lateral change acceleration information; inputting the posture change information into a preset target fall posture matching neural network model, and obtaining a matching result output by the target fall posture matching neural network model, wherein the matching result includes at least any one of the following: the person is in a falling state and the person is not in a falling state; when the matching result is that the person is in a falling state, continuing to execute "again based on the setting in the bathroom The radar device obtains the current real-time radar signal, and determines whether the status of the person in the bathroom meets the first preset condition based on the current real-time radar signal"; when the matching result is that the person is not in a falling state, the micro-motion amplitude information of the person in the bathroom is determined based on the current real-time radar signal, wherein the micro-motion amplitude information includes: the millimeter-level movement amplitude information of the chest cavity for breathing and the millimeter-level movement amplitude information of the chest cavity for heartbeat; and based on the millimeter-level movement amplitude information of the chest cavity for breathing and the millimeter-level movement amplitude information of the chest cavity for heartbeat in the micro-motion amplitude information, when it is detected that the person in the bathroom has breathing and heartbeat, it is determined that the personnel situation information is that the person is in a resident state, and continue to execute the step of "again based on the radar device set in the bathroom to obtain the current real-time radar signal, and determine whether the status of the person in the bathroom meets the second preset condition based on the current real-time radar signal".

[0045] That is, this technical solution provides a highly intelligent posture monitoring system for people in the bathroom by combining real-time radar signals and neural network models, which can achieve extremely accurate identification and processing, especially in preventing and detecting fall events. The core of this solution is to obtain point cloud orientation information through real-time radar signals, and use this information to parse and determine the posture changes of people in the bathroom. Specifically, the system first receives the current real-time radar signal from the radar device, and extracts the point cloud orientation information through the signal. Point cloud orientation information usually includes X-axis, Y-axis, and Z-axis information in three-dimensional space, which reflects the specific position and posture of the person in the bathroom.

[0046] Once the point cloud orientation information is obtained, the system will parse it to generate information about the person's posture change. Posture change information covers multiple key indicators, including changes in the person's center of gravity, lateral area changes, longitudinal acceleration changes, and lateral acceleration changes. These indicators are obtained through a detailed analysis of the person's movement trajectory and body posture in space. For example, center of gravity change information can reflect whether the person is maintaining balance, while lateral area change information can indicate the degree of lateral expansion of the person. For example, when a person changes from a standing position to a sitting or lying position, the lateral area of ​​the body usually changes significantly. Longitudinal and lateral acceleration information can provide detailed data on the speed and direction of a person's movement, thereby helping the system to more comprehensively understand the person's posture changes.

[0047] Next, the system inputs these posture change information into the preset target fall posture matching neural network model for processing. The neural network model has been trained with a large amount of data and can recognize specific posture patterns, especially those related to falls. By comparing the real-time posture change information with the existing model in the neural network, the system can obtain matching results. The matching results are mainly divided into two types: one is to determine that the person is in a fall state, and the other is to determine that the person is not in a fall state. If the system determines that the person is in a fall state, it will continue to obtain real-time radar signals from the radar device and further evaluate whether the person's state meets the first preset condition based on these signals. The purpose of this step is to perform secondary confirmation of possible fall events to prevent false alarms. For example, if the detected fall state may be just a misjudgment caused by the person bending over or picking up an object, the system will further confirm whether this state continues through continuous radar signals.

[0048] However, if the matching result shows that the person is not in a falling state, the system will further determine the person's micro-motion amplitude information based on the current real-time radar signal. The micro-motion amplitude information includes two key parts: the millimeter-level movement amplitude information of the respiratory chest cavity and the millimeter-level movement amplitude information of the heartbeat chest cavity. By monitoring these tiny body movements, the system can determine whether the person's vital signs are functioning normally. For example, by monitoring the movement amplitude of the respiratory chest cavity, the system can determine whether the person is breathing normally, while the movement amplitude of the heartbeat chest cavity can reflect the beating of the heart. The accuracy of these data can usually reach the millimeter level, which is enough to capture subtle changes in vital signs.

[0049] When the system confirms that the person is still breathing and has a heartbeat in the bathroom based on these micro-motion amplitude information, it will determine that the person is currently in a resident state. This means that although the person has not fallen or other dangerous situations, they may still stay in the bathroom, such as sitting on the toilet or resting next to the sink. At this time, the system will continue to obtain real-time radar signals and determine whether the person's status meets the second preset condition based on these signals. This continuous monitoring mechanism ensures that the system can accurately judge their safety status even if the person does not move for a long time or moves weakly. For example, if an elderly person sits on the toilet to rest in the bathroom because he is tired, the system can accurately judge that he is in a resident state through the detected breathing and heartbeat signals, rather than misjudging him as falling.

[0050] Through this design, the system can not only accurately detect changes in people's posture, especially falls, but also continue to monitor people's vital signs in the case of micro-movement to ensure their safety in the bathroom. This dual detection mechanism not only improves the sensitivity and accuracy of the system, but also effectively prevents false alarms and missed alarms, providing more comprehensive safety protection. For example, if an elderly person suddenly loses balance and falls in the bathroom, the system will be able to quickly identify and issue an alarm message. But if he just stops or sits down to rest, the system will also confirm his safety status through breathing and heartbeat detection to avoid unnecessary alarms. Therefore, this technical solution realizes a more intelligent and reliable bathroom safety monitoring system, which can effectively protect the life safety of users, especially in the use scenarios of the elderly or people with limited mobility.

[0051] In another example, if it is detected that the person in the bathroom has no breathing or heartbeat, the step of "using a voice device installed in the bathroom to conduct voice inquiries and voice recording of the person in the bathroom, and obtain voice data obtained by the voice recording, and based on the recorded voice data, determine whether the person in the bathroom is in a dangerous state, and generate an alarm message if the person in the bathroom is in a dangerous state" is directly executed. However, in general, it is considered that the person has left the bathroom, that is, there is no person in the bathroom.

[0052] In one example, the posture change information is input into a preset target fall posture matching neural network model, and a matching result output by the target fall posture matching neural network model is obtained, including: inputting the posture change information into a preset target fall posture matching neural network model; determining, through the target fall posture matching neural network model, whether the point cloud center of gravity shows a downward trend higher than a preset amplitude based on the personnel center of gravity change information in the posture change information; in the case where the point cloud center of gravity shows a downward trend higher than a preset amplitude, based on the personnel longitudinal change acceleration information and the personnel lateral change acceleration information in the posture change information, judging whether the personnel longitudinal change acceleration and the personnel lateral change acceleration are generated by the falling behavior of the person in the bathroom, rather than the walking behavior of the person in the bathroom; in the case where the judgment result is that the falling behavior of the person in the bathroom is generated, rather than the walking behavior of the person in the bathroom, judging whether the lateral point cloud area of ​​the person has a sudden change trend based on the personnel lateral area change information in the posture change information; in the case where the lateral point cloud area of ​​the person has a sudden change trend, outputting a matching result that the person is in a falling state.

[0053] That is, if Figure 5 , Figure 6 and Figure 7 As shown in the figure, the core of this technical solution is to use a neural network model to accurately identify the fall state of people in the bathroom and distinguish this state from ordinary walking or other normal activities. This method analyzes the posture change information of people, gradually selects features, and finally makes a judgment, thereby improving the accuracy and reliability of the system.

[0054] First, the system inputs the posture change information obtained from the radar device into a preset target fall posture matching neural network model. This neural network model has been optimized through a large amount of training data and can identify specific posture features related to falls. The posture change information includes several key data points, such as the person's center of gravity change information, longitudinal and lateral acceleration information, and the person's lateral area change information.

[0055] Next, the neural network model will first analyze the center of gravity change information of the person in the posture change information. Specifically, the system will check whether the center of gravity in the point cloud data has a downward trend higher than the preset amplitude. A sudden drop in the center of gravity is usually an important indicator of a fall, so this step is to initially screen out possible fall events. If the model detects a significant downward trend in the center of gravity of the point cloud, the system will further analyze other posture change information.

[0056] After confirming that the center of gravity has dropped, the system will then determine whether the acceleration change is caused by the person's falling behavior or by normal walking behavior based on the longitudinal and lateral acceleration change information. This step is very critical because it can effectively distinguish between falling and normal walking. Longitudinal and lateral acceleration changes can reflect how a person moves in space. Falling behavior is usually accompanied by a large acceleration change, while walking behavior does not produce such a drastic acceleration change. Therefore, through the analysis of acceleration, the system can further confirm whether the person is really in a falling state.

[0057] If the judgment result shows that the acceleration change is indeed caused by falling behavior, the system will continue to analyze the person's lateral area change information in the posture change information. The lateral area change information reflects the degree of expansion of the person's body in the lateral direction. If a person falls, their body will usually spread out on the ground, causing a sudden change trend in the lateral area. By checking whether the lateral area has a sudden change, the system can be more confident that the person has indeed fallen, and is not in another state.

[0058] When the horizontal point cloud area shows a sudden change trend, the system will eventually output a matching result that the person is in a falling state. This matching result will trigger subsequent safety measures, such as alarming or notifying relevant personnel. This entire process ensures that the system can not only accurately identify fall events, but also effectively eliminate false alarms and avoid erroneous alarms caused by ordinary walking or other activities.

[0059] For example, if an elderly person suddenly loses balance and falls in the bathroom, the system will first detect the sharp drop in their center of gravity, and combined with the changes in longitudinal and lateral acceleration, determine that this is a fall rather than an ordinary walk. At the same time, the system will also detect the sudden increase in its lateral area. This comprehensive information will be input into the neural network model, and finally a match result of a fall will be obtained, and the system will immediately issue an alarm notification. In this way, the technical solution achieves highly accurate identification of fall events and significantly improves the safety level of people in the bathroom.

[0060] In one example, the current real-time radar signal is obtained again based on the radar equipment installed in the bathroom, and according to the current real-time radar signal, it is determined whether the status of the person in the bathroom meets the second preset condition, including: obtaining the current real-time radar signal again based on the radar equipment installed in the bathroom, and determining the limb movement amplitude information of the person in the bathroom according to the current real-time radar signal; and based on the limb movement amplitude information, detecting whether the cumulative time value of the limb movement amplitude of the person in the bathroom is lower than a preset threshold is higher than the preset threshold; if it is higher than the preset threshold, it is determined that the status of the person in the bathroom meets the second preset condition, and continue to execute the step of "when the status of the person in the bathroom meets the second preset condition, conduct voice inquiries and voice collection to the person in the bathroom through the voice equipment installed in the bathroom".

[0061] That is, this technical solution uses radar equipment to monitor people in the bathroom in real time, aiming to determine whether people are in a possible dangerous state, especially by detecting small changes in people's limb movements to judge their safety status. The working principle of this solution is based on the analysis of the amplitude information of people's limb movements, and predicting and handling possible critical situations.

[0062] Specifically, after detecting a possible risk, the system will first obtain the current real-time radar signal again through the radar device installed in the bathroom. This radar signal is used to capture and analyze the limb movement amplitude information of people in the bathroom. Limb movement amplitude information reflects the activity of people's movements, such as the movement of arms or legs. These tiny movement changes are continuous and regular under normal circumstances, but if people are in trouble, especially the elderly or those with limited mobility, limb movement may be reduced or even stopped.

[0063] Next, based on the acquired limb movement amplitude information, the system will detect whether these movement amplitudes are lower than the preset threshold for a long time. The preset threshold here is a standard set by the system to determine whether a person's limb movement is normal. For example, the system may preset that if a person's limb movement amplitude continues to be lower than a certain minimum for a certain period of time, this may indicate that the person is in an abnormal state, such as losing consciousness, being unable to move, or encountering other dangerous situations. Therefore, the system must not only detect that the limb movement amplitude is lower than the preset value, but also monitor whether this low movement state continues for more than a certain cumulative time value.

[0064] Once the system detects that the accumulated time of the limb movement amplitude continuously below the preset threshold is higher than the set threshold, it will determine that the person may be in a dangerous state and that his / her state meets the second preset condition. At this point, the system will proceed to the next step, which is to conduct voice inquiries and voice recording to the person through the voice equipment in the bathroom to further confirm the person's state.

[0065] For example, if an elderly person sits in the bathroom without obvious body movements for a long time, it may be because they feel dizzy, lose consciousness, or cannot move for other reasons. After the system detects this abnormality, it will think that the person's state may be abnormal and start the voice inquiry function to ask the person if he needs help. Through this multi-level, multi-step detection and confirmation, the system can effectively identify potential critical situations and take appropriate measures to maximize the safety of personnel.

[0066] The advantage of this solution is that it can predict some subtle dangerous situations through continuous monitoring and analysis of subtle movements, especially when the elderly or infirm use the toilet. It can issue an alarm in the early stages of the problem and provide timely rescue.

[0067] like Figure 8 As shown, the present invention is an embodiment and further provides a safety monitoring system for a bathroom, the safety monitoring system also includes: a control processor, a voice device, a radar device and a lighting device, the control processor is communicatively connected to the voice device, the radar device and the lighting device, and is used to jointly execute the steps of the method described in the above embodiment; wherein, the safety monitoring system also includes: a wireless communication module, which is communicatively connected to the control processor, and is used to upload the data processed by the control processor to the server; the server side, which is communicatively connected to the mobile terminal, is used to store the data content uploaded by the wireless communication module, and upload the alarm information in the data content to the mobile terminal; the mobile terminal is used to check real-time alarm information and historical alarm information.

[0068] For ease of understanding, another embodiment is provided for technical explanation: The content of the present invention involves the comprehensive application of millimeter wave radar technology, voice technology, lighting application, one-button call for help, gyroscope technology and Internet of Things technology to form a comprehensive fall monitoring system. The system is not only suitable for bathrooms, but can also be extended to scenes that require security monitoring, such as kitchens, living rooms, and bedrooms. The system consists of millimeter wave radar, communication modules, voice intercom, emergency call function and lighting linkage, and is connected to the server to display relevant subscription information on the mobile phone App or large screen display background.

[0069] exist Fig. 9As shown in the figure, the central controller or wireless module is connected to the voice module, lighting LED, fall detection millimeter wave radar sensor and SOS physical switch respectively. The integrated system is connected to the server to report the status information in real time. After the device is powered on, it is triggered by the physical button, and the device connects to the server and reports the LED light status and other behavior status data normally.

[0070] The functions of the device include millimeter-wave radar fall posture alarm, dynamic monitoring of human breathing and heartbeat, static residence time monitoring, and voice alarm. The information status detected by the radar is connected to the controller MCU through the serial port, and the MCU parses the radar detection attribute status. When a person enters the radar detection area, the MCU controls the light to light up to determine whether a fall, abnormal struggle, or immobility occurs. The voice system of the device will issue a safety inquiry. If a danger occurs, the device will immediately report the alarm information to the background. If a standing state is detected, the danger alarm is lifted, and the communication module reports the status information to the cloud server, and the mobile App can subscribe to the message. The device turns on the lighting when a person approaches and turns off the light when the person leaves.

[0071] The system supports offline voice function and can actively or passively perform voice interaction. When a person is in danger, he can directly call for help through the intercom. The system will respond according to the voice inquiry, and actively ask if help is needed when abnormal posture and static residence are detected during the fall, and report the alarm information to the service center. The product form includes two ways of top installation and side wall installation. The millimeter wave radar transmits fixed-frequency electromagnetic waves through the antenna and receives the reflected signal, and collects the movement amplitude and azimuth information according to the changes in frequency and angle. The output movement amplitude information of the radar covers large limb movements, centimeter-level movement of the chest cavity during breathing, and millimeter-level movement of the heartbeat. The azimuth information is divided into three directional dimensions: X, Y, and Z: X is the horizontal left and right distance between the radar and the ground, Y is the distance between the object and the radar, and Z is the height information of the object vertical to the ground.

[0072] By superimposing these dimensional information, the system can analyze and determine the person's motion state and vital signs. Based on the large range of limb movement and the orientation information of the point cloud, the system can identify the motion state (such as active or still), body posture, abnormal falls, and struggles. By the changes in the amplitude of breathing and heart rate micro-movements, the system can determine whether the person has been stationary for a long time.

[0073] The method for judging the abnormal state includes: when the millimeter-wave radar detects strong movement intensity and approaches the set distance boundary condition, the radar reports the status of someone; when a person enters the fall detection range and triggers the fall model judgment mechanism, the radar judges in real time whether the point cloud center of gravity drops, the lateral point cloud area expands, and the longitudinal and lateral accelerations are in line with the fall trend, and confirms the abnormal fall state after three rounds of judgment; when it is detected that the amplitude of micro-movements such as breathing and heartbeat persists but there is no large limb movement, the radar determines it as a stationary state and counts. If it is stationary for a long time, the stationary residence warning state is reported.

[0074] like Fig.10 and Fig.11 As shown in the figure, this product is divided into two parts: local applications and cloud applications. Local applications include millimeter wave radar detection, sleep detection, voice intercom and lighting linkage. The wireless module MCU will filter and process the human presence information, fall and stationary residence information and voice intercom information reported by the radar, and then transmit it to the cloud through the MQTT protocol. Cloud applications include data recording, alarm status push and App control. Data is transmitted to the background through standard protocols such as MQTT, Zigbee3.0, and GATT for storage and recording, and real-time linkage is achieved according to importance. Alarm status includes fall alarm, stationary residence alarm and voice call, etc., and message push, SMS push or phone notification are carried out through cloud services.

[0075] In terms of sound function, the millimeter-wave radar activates the voice function after detecting a human signal. The device provides offline voice triggering, and users can set sound effects such as "turn on the light", "turn off the light", "someone is coming" or "help". The device will send a distress alarm through a remote App push, and will respond with voice after receiving a voice trigger.

[0076] The invention has the advantages of integrating the functions of fall posture alarm, stationary stay alarm, lighting linkage and voice active alarm. The protection point includes the combination of millimeter wave radar, lighting system and voice system for bathroom monitoring, as well as millimeter wave radar fall recognition and stationary stay analysis method.

[0077] like Fig.12 As shown, the present invention is an embodiment and further provides a toilet safety monitoring device, characterized in that the safety monitoring device comprises: The acquisition unit 1 is used to acquire a current real-time radar signal based on a radar device arranged in the toilet, and determine the posture information of the person in the toilet based on the current real-time radar signal, wherein the person situation information at least includes: the person is in a falling state and the person is in a staying state; The first determination unit 2 is used for, when the personnel situation information indicates that the person is in a falling state, obtaining a current real-time radar signal based on the radar device arranged in the toilet again, and determining whether the state of the person in the toilet meets a first preset condition according to the current real-time radar signal, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; The second determination unit 3 is used for, when the personnel situation information indicates that the personnel is in a resident state, obtaining a current real-time radar signal based on a radar device arranged in the toilet again, and determining whether the status of the personnel in the toilet meets a second preset condition according to the current real-time radar signal, wherein the second preset condition at least includes: no action state and exceeding a preset time; The alarm unit 4 is used to make voice inquiries and record voices of the people in the bathroom through a voice device arranged in the bathroom when the status of the people in the bathroom meets the first preset condition or the second preset condition, and obtain voice data obtained by the voice recording, and determine whether the people in the bathroom are in a dangerous state based on the recorded voice data, and generate an alarm message when the people in the bathroom are in a dangerous state.

[0078] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0079] Reference Fig.13 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Fig.13 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0080] Those skilled in the art will understand that Fig.13 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0081] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0082] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0083] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0084] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for safety monitoring of toilets, It is characterized in that The following steps are involved: Based on the radar equipment arranged in the toilet, a current real-time radar signal is obtained, and based on the current real-time radar signal, posture information of the person in the toilet is determined, wherein the person situation information at least includes: the person is in a falling state and the person is in a staying state; In the case where the personnel situation information indicates that the person is in a falling state, the current real-time radar signal is obtained again based on the radar device arranged in the toilet, and according to the current real-time radar signal, it is determined whether the state of the person in the toilet meets the first preset condition, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; In the case where the personnel situation information indicates that the personnel is in a stationary state, the current real-time radar signal is acquired again based on the radar device arranged in the toilet, and according to the current real-time radar signal, it is determined whether the status of the personnel in the toilet meets a second preset condition, wherein the second preset condition at least includes: no action state and exceeding a preset time; When the status of the person in the bathroom meets the first preset condition or the second preset condition, a voice device arranged in the bathroom is used to conduct voice inquiries and voice recording of the person in the bathroom, and voice data obtained by the voice recording is obtained, and based on the recorded voice data, it is determined whether the person in the bathroom is in a dangerous state, and when the person in the bathroom is in a dangerous state, an alarm message is generated.

2. The safety monitoring method according to claim 1, It is characterized in that Before obtaining a current real-time radar signal based on a radar device disposed in the toilet and determining the posture information of the person in the toilet based on the current real-time radar signal, the safety monitoring method further includes: Based on the radar equipment installed in the bathroom, the current real-time radar signal is obtained; The current real-time radar signal moving signal is screened and processed by a preset radar interference background noise filtering model to obtain the motion energy intensity in the bathroom; Comparing the motion energy intensity in the toilet with the energy model threshold range triggered by a person entering the toilet; If the motion energy intensity in the toilet meets the threshold range, it is determined that there is a person in the toilet, the lighting device arranged in the toilet is controlled to be turned on, and the step of "obtaining a current real-time radar signal based on a radar device arranged in the toilet, and determining the posture information of the person in the toilet based on the current real-time radar signal" is executed; If the motion energy intensity in the toilet meets the threshold range, it is determined that there is no person in the toilet, and the lighting equipment installed in the toilet is controlled to be turned off.

3. The safety monitoring method according to claim 1, It is characterized in that Through the voice equipment arranged in the toilet, voice inquiry and voice recording are made to the person in the toilet, and voice data obtained by the voice recording is obtained, and based on the recorded voice data, whether the person in the toilet is in a dangerous state is determined, and if the person in the toilet is in a dangerous state, an alarm message is generated, including: The voice device disposed in the toilet records the voice of the person in the toilet to obtain first voice data, and the radar device disposed in the toilet detects abnormalities of the person in the toilet to obtain first detection data; Based on the first voice data and the first detection data, playing preset safety inquiry voice data to the person in the toilet through the voice device to determine whether the person in the toilet needs help; The voice of the person in the toilet is recorded again based on the voice device to obtain second voice data, and when it is determined based on the second voice data that the person in the toilet does not need help, the step of "recording the voice of the person in the toilet through the voice device arranged in the toilet to obtain first voice data, and performing abnormality detection on the person in the toilet through the radar device arranged in the toilet to obtain first detection data" is continued; When it is determined based on the second voice data that the person in the toilet needs help, an alarm message is generated.

4. The safety monitoring method according to claim 1, It is characterized in that Determining the posture information of the person in the toilet based on the current real-time radar signal includes: Based on the current real-time radar signal, determine the point cloud position information detected in the bathroom, wherein the point cloud position information includes: X-axis information, Y-axis information and Z-axis information; Analyze and process the point cloud orientation to obtain posture change information of the person in the toilet, wherein the posture change information at least includes: personnel center of gravity change information, personnel lateral area change information, personnel longitudinal change acceleration information and personnel lateral change acceleration information; The posture change information is input into a preset target falling posture matching neural network model, and a matching result output by the target falling posture matching neural network model is obtained, wherein the matching result includes at least one of the following: the person is in a falling state and the person is not in a falling state; If the matching result is that the person is in a falling state, continue to execute the step of "again obtaining the current real-time radar signal based on the radar device arranged in the toilet, and determining whether the state of the person in the toilet meets the first preset condition according to the current real-time radar signal"; When the matching result indicates that the person is not in a falling state, the micro-motion amplitude information of the person in the bathroom is determined based on the current real-time radar signal, wherein the micro-motion amplitude information includes: the millimeter-level movement amplitude information of the chest cavity for breathing and the millimeter-level movement amplitude information of the chest cavity for heartbeat; And based on the breathing chest millimeter-level movement amplitude information and the heartbeat chest millimeter-level movement amplitude information in the micro-motion amplitude information, when it is detected that the person in the bathroom is breathing and having heartbeat, it is determined that the person situation information is that the person is in a resident state, and the step of "again obtaining the current real-time radar signal based on the radar equipment arranged in the bathroom, and determining whether the status of the person in the bathroom meets the second preset condition based on the current real-time radar signal" is continued.

5. The safety monitoring method according to claim 1, It is characterized in that Inputting the posture change information into a preset target fall posture matching neural network model, and obtaining a matching result output by the target fall posture matching neural network model, including: Inputting the posture change information into a preset target falling posture matching neural network model; By matching the target falling posture with the neural network model, based on the personnel center of gravity change information in the posture change information, determining whether the center of gravity of the point cloud shows a downward trend higher than a preset amplitude; In the case where the center of gravity of the point cloud shows a downward trend higher than a preset amplitude, based on the longitudinal acceleration information and the lateral acceleration information of the person in the posture change information, it is determined whether the longitudinal acceleration and the lateral acceleration are caused by the falling behavior of the person in the toilet rather than the walking behavior of the person in the toilet; In the case where the result of the judgment is that the falling behavior of the person in the toilet is caused rather than the walking behavior of the person in the toilet, judging whether the horizontal point cloud area of ​​the person has a sudden change trend based on the horizontal area change information of the person in the posture change information; When the area of ​​the lateral point cloud of the person undergoes a sudden change trend, a matching result indicating that the person is in a falling state is output.

6. The safety monitoring method according to claim 1, It is characterized in that Acquiring a current real-time radar signal based on the radar device disposed in the toilet again, and determining whether the status of the person in the toilet meets a second preset condition according to the current real-time radar signal, including: Again, based on the radar device arranged in the toilet, a current real-time radar signal is obtained, and according to the current real-time radar signal, the body movement amplitude information of the person in the toilet is determined; Based on the limb movement amplitude information, detecting whether the cumulative time value of the limb movement amplitude of the person in the bathroom being lower than a preset threshold is higher than a preset threshold; If it is higher than the preset threshold, it is determined that the status of the person in the bathroom meets the second preset condition, and the step of "when the status of the person in the bathroom meets the second preset condition, voice inquiries and voice recording are made to the person in the bathroom through the voice equipment arranged in the bathroom" is continued.

7. A toilet safety monitoring system, It is characterized in that The safety monitoring system further includes: a control processor, a voice device, a radar device and a lighting device, wherein the control processor is in communication with the voice device, the radar device and the lighting device, and is used to jointly perform the steps of the method according to any one of claims 1 to 7; wherein the safety monitoring system further includes: A wireless communication module, connected to the control processor for uploading the data processed by the control processor to the server; The server is connected to the mobile terminal for storing the data content uploaded by the wireless communication module and uploading the alarm information in the data content to the mobile terminal; The mobile terminal is used to check real-time alarm information and historical alarm information.

8. A safety monitoring device for the bathroom, It is characterized in that The safety monitoring device comprises: an acquisition unit, configured to acquire a current real-time radar signal based on a radar device disposed in the toilet, and determine posture information of a person in the toilet based on the current real-time radar signal, wherein the person situation information at least includes: a person in a falling state and a person in a staying state; A first determining unit is used for, when the personnel situation information indicates that the person is in a falling state, obtaining a current real-time radar signal based on a radar device arranged in the toilet again, and determining whether the state of the person in the toilet meets a first preset condition according to the current real-time radar signal, wherein the first preset condition at least includes: a continuous struggling state and / or a state of not getting up, and exceeding a preset time; A second determination unit is used for, when the personnel situation information indicates that the personnel is in a resident state, obtaining a current real-time radar signal based on a radar device arranged in the toilet again, and determining whether the status of the personnel in the toilet meets a second preset condition according to the current real-time radar signal, wherein the second preset condition at least includes: no action state and exceeding a preset time; An alarm unit is used to make voice inquiries and record voices of the people in the bathroom through a voice device arranged in the bathroom when the status of the people in the bathroom meets the first preset condition or the second preset condition, and obtain voice data obtained by the voice recording, and determine whether the people in the bathroom are in a dangerous state based on the recorded voice data, and generate an alarm message when the people in the bathroom are in a dangerous state.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.