Non-contact alarm device triggered by brain waves and alarm detection method thereof

Abstract

The present invention proposes a non-contact alarm device triggered by brain waves and an alarm detection method thereof, which solves the technical problem of non-contact alarm. All components in the receiving and detecting unit are installed in the headband, and all components in the processing unit are installed in the box. The sending unit is the user's smart phone. The processing unit and the headband, and the processing unit and the user are interconnected through a set of Bluetooth. The receiving and detecting unit naturally collects brain waves, the processing unit performs calculations, and the mobile phone sends a distress message to the outside world. The method includes wearing a headband to collect brain wave signals in real time; no alarm in a stable state; using a startle detection algorithm to judge an emergency state; an alarm is generated after the emergency state is judged; and reset after the emergency state alarm. The startle and blinking algorithm is the core algorithm of the present invention to solve the non-contact alarm. The present invention is a non-contact alarm, which does not require manual intervention for natural alarm, has a low false alarm rate, is portable, and has strong concealment, and is suitable for personal safety and self-defense alarm.

Description

Technical field:

[0001] The present invention belongs to the field of intelligent electronic technology, and in particular refers to a technology applied to personal safety protection alarm, mainly involving non-contact alarm, specifically a non-contact alarm device triggered by brain waves and an alarm detection method thereof, which are used for personal safety self-defense alarm. Background technology:

[0002] Currently, portable alarm devices on the market can be divided into four categories: contact objects, inductive objects, mobile phone apps, and smart alarms. Most of them are activated by active triggering. In an emergency, they are subject to various conditions and cannot be actively intervened, resulting in their inability to perform set functions. Object alarms such as anti-wolf alarms are equivalent to small emergency speakers. They are mostly made into small ornaments. By touching the switch in dangerous situations, they can emit a buzzer of up to 120 decibels. They cannot provide positioning information and emergency help alarm functions. Inductive objects use passive infrared detectors to detect approaching human bodies through infrared sensing and trigger alarms. Smart products such as the "PaPa Smart Self-Defense Alarm" can turn on the GPS positioning function and recording function by the user pressing the switch to collect location information and voice information. After the feedback confirms the alarm operation, the recording and location information are sent to the mobile phone through the Bluetooth module, and then the information is edited and sent to the stored contacts through the mobile phone APP. If the feedback cancels the alarm operation, the above operation is stopped. Mobile phone apps are more modern than physical objects. They can trigger the alarm by shaking the phone, plugging the headphone jack, etc., and send information to emergency contacts or public security departments based on the location.

[0003] In the patent design of the patent number CN206991514U, "Portable Intelligent Security Active Alarm", a portable alarm is disclosed, including a housing, a PCB board, a buzzer, a power supply and a plug. The PCB board is provided with a socket, and correspondingly, the housing is provided with a jack. The above-mentioned plug passes through the jack and is pluggable and arranged on the socket. When the plug is pulled out, the buzzer and the LED alarm light are turned on, the buzzer emits a high-decibel alarm sound, and the LED alarm light keeps flashing to scare away suspects with illegal intentions. When the plug is inserted, the buzzer and the LED alarm light are turned off.

[0004] In the article "Development of a non-contact apnea alarm device" published by Hu Ye, Li Chuantao, Qi Fugui, Wang Shuaijie, Zhang Hua, Wang Jianqi, and Lu Guohua in China Medical Equipment, 2015, 30(05):18-20, a non-contact apnea alarm device was designed. The device mainly uses a bio-radar sensor to detect the respiratory signal non-contactly, pre-processes the signal through a hardware circuit, and finally connects the pre-processed signal to a single-chip microcomputer for comprehensive apnea judgment and alarm. There is no wireless communication module in the system, so the alarm signal cannot be transmitted over a long distance by wireless means, which is not suitable for remote wireless monitoring and cannot be used to seek help from the outside in a critical situation. In addition, the device cannot eliminate the interference caused by movement, and the user's use state is limited to a resting state, and cannot be used in an active state.

[0005] The mainstream alarm products on the market are mainly object alarms. Since the triggering method is active, it is not easy to perform alarm operation when personal freedom is restricted. In addition, the false alarm rate is high because alarm triggering devices such as pull rings are easy to fall off during daily activities and there is no feedback link in the alarm link. In addition, since there is no system connected to the Internet, the alarm method is single, and it is impossible to get external help in time in remote areas, which can easily delay the opportunity to alarm.

[0006] In order to protect personal safety in emergencies, an alarm device is needed that does not require manual intervention and is not easily discovered. Summary of the invention:

[0007] In view of the deficiencies, defects and needs of the prior art, the present invention proposes a portable non-contact alarm device triggered by brain waves, which does not require manual contact, has a low misjudgment rate and an alarm detection method thereof.

[0008] The present invention is a non-contact alarm device triggered by brain waves, which is provided with an alarm signal receiving and detecting unit, an alarm signal processing unit and an alarm signal sending unit in sequence according to the alarm signal transmission process; referred to as the receiving and detecting unit, the processing unit and the alarm unit; it is characterized in that the entity equipment of the non-contact alarm device triggered by brain waves is divided into an alarm headband, an alarm signal processing unit and an alarm signal sending unit; the alarm signal receiving and detecting unit comprises a TGAM brain wave processing module, a micro main control board, a vibration module, a Bluetooth module and a bioelectric acquisition electrode, wherein the bioelectric acquisition electrode comprises an ear clip type potential sensor and a forehead potential sensor, all of which are installed and fixed on the alarm headband The alarm headband is designed as a double-layer headband with an interlayer, and all components of the alarm signal receiving and detecting unit are installed in the interlayer of the headband, and an ear clip-type potential sensor is led out on the left side of the headband; the forehead potential sensor collects bioelectric signals, and the TGAM brain wave processing module converts the bioelectric signals into brain wave data. The brain wave data includes a raw wave value and an EEG power value, and the blink value can be calculated through the raw wave value; the TGAM brain wave processing module transmits the user's brain wave data to the alarm signal processing unit through the Bluetooth module, and after being processed by the alarm signal processing unit, it is fed back to the Bluetooth module to send vibration command information to the micro main control board to control the vibration module to vibrate; the alarm signal processing The processing unit includes a main control board, a GPS positioning module, a GPS signal receiving antenna and at least two Bluetooth modules, the two Bluetooth modules are respectively called the first Bluetooth module and the second Bluetooth module, and the main control board is used as the core to form a brain wave operation processing circuit, all of which are installed in a square box; the Bluetooth module in the alarm signal receiving and detecting unit is set to a paired state with the first Bluetooth module in the alarm signal processing unit; the second Bluetooth module in the alarm signal processing unit is set to an unpaired state and can be paired with any smart phone; the alarm signal processing unit receives the brain wave signal from the alarm signal receiving and detecting unit through the first Bluetooth module, and returns the vibration command information to the alarm signal through the same Bluetooth module The receiving and detecting unit, at the same time, the second Bluetooth module transmits the alarm command information generated by the main control board operation to the alarm signal sending unit for alarm. The alarm signal sending unit is a smart phone, and the inherent Bluetooth on the smart phone is paired and connected with the second Bluetooth module in the alarm signal processing unit to receive the alarm command information from the alarm signal processing unit; the user's smart phone is installed with a specific application, and the user's designated specific smart phone also needs to have a specific application installed, and the application contains the user's current location information, current time, and distress text message; the user needs to specify the specific smart phone number that receives the alarm information in advance to issue an alarm; it is enough to alert people.

[0009] The present invention is also an alarm detection method for a non-contact alarm device triggered by brain waves, which is used on the non-contact alarm device triggered by brain waves and is characterized in that it includes the following steps:

[0010] (1) Wearing the alarm headband to collect brain wave signals in real time: The user wears the alarm headband correctly and the alarm headband is in normal working condition. In the working state, the user is required to place the sensor contact end in the middle of the forehead, adjust the headband to the appropriate tightness, remove the hair and other obstructions, and achieve full contact. At the same time, the ear clip contact end also needs to be in full contact with the skin at the earlobe. When the data stream Signal value in the TGAM brain wave processing module is 0, it is confirmed that the user has worn it correctly and the alarm headband is in normal working condition. The user installs a specific application on his mobile phone and a specified specific smartphone.

[0011] (2) No alarm signal when in stable state: If no emergency occurs, the user's current brain waves will not trigger the alarm process. The entire system will not trigger any operation, the smartphone software will not send an alarm message, and the alarm signal receiving and detection unit will monitor the user's brain waves in real time.

[0012] (3) Use the Euclidean distance startle detection algorithm to make abnormal judgments in sudden situations: When a special situation occurs to the user, the brain wave data will generate a burst signal. The burst bioelectric signal is converted into brain wave data by the TGAM brain wave processing module. The brain wave data is sent to the first Bluetooth module and the main control board in the alarm signal processing unit through the Bluetooth module in the alarm signal receiving and detecting unit. The main control board calculates the startle detection algorithm based on the Euclidean distance and verifies the brain wave data to monitor whether the user is currently in a dangerous state. If the brain wave data converted from the burst bioelectric signal is received, it is in a dangerous state. The main control board will generate vibration command information, which is sent to the micro main control board in the alarm signal receiving and detecting unit through the first Bluetooth module in the alarm signal receiving and detecting unit. After receiving the vibration command information, the micro main control board activates the vibration module to remind the user that the user is currently in the alarm preparation stage.

[0013] (4) Determine whether to alarm in an emergency state: The main control board runs a double down-sampling blink detection algorithm on the brain wave data received from the first Bluetooth module to calculate the user's current blinking frequency, and uses the blinking frequency to determine whether an alarm is necessary. If the user's current blinking frequency reaches the set threshold, the main control board will send an alarm signal to the alarm signal sending unit, i.e., the smartphone, through the second Bluetooth module in the processing unit, and at the same time send a vibration signal to the micro main control board in the receiving and detection unit through the first Bluetooth module again, so that the micro main control board activates the vibration module to remind the user that the alarm signal has been issued. If the user's blinking frequency does not reach the set threshold, the main control board in the processing unit will determine it as a false detection and will not run the subsequent alarm program. If the user is indeed in a dangerous state but does not receive a vibration reminder, the user can directly activate the alarm program by blinking quickly.

[0014] (5) Device alarm process in emergency state: When the main control board in the processing unit determines that the current state requires an alarm, the main control board sends an alarm instruction message to the user's smartphone through the second Bluetooth module. After the smartphone receives the alarm instruction message based on its own inherent Bluetooth, it automatically sends a distress message containing the user's current location and time to the specific smartphone number specified in advance by the user through a specific smartphone application.

[0015] (6) Reset process after alarm completion: After a complete alarm process is completed, the user needs to enter a specific smartphone application to manually cancel the current alarm status, and then the specific smartphone application will stop sharing location information and the entire alarm device will be reset.

[0016] In order to allow people to directly raise an alarm in an emergency without complicated actions and manual operations, the present invention provides a non-contact alarm device triggered by brain waves and an alarm detection method thereof.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] Alarm based on sudden signals without manual contact: In most criminal cases involving infringement of the personal and property safety of victims, the victims are suddenly restricted in their personal freedom and cannot use their hands or other limbs to trigger the alarm device. Therefore, an alarm device that does not require manual intervention is needed to send alarm information to the outside world. This device does not require manual triggering and can directly alarm based on the blinking frequency of the eyes, greatly improving the success rate of the alarm.

[0019] Low false alarm rate: The startle detection algorithm based on Euclidean distance determines whether the user is in a dangerous situation. The algorithm has been verified to be able to detect the brain wave data of most people when they encounter an emergency. Secondly, the double-downsampling blink detection algorithm quickly calculates the user's current blinking frequency. The user needs to blink before the program can send an alarm message to the outside world. The combination of active and passive alarm processes can maximize the guarantee of being able to efficiently detect emergencies while preventing false alarms.

[0020] Portability: The device of the present invention includes a headband device and a data processing module, with a total weight of no more than 500g, which is equivalent to the weight of less than a bottle of water. In daily use, you only need to wear the headband device and carry the data processing module with you. Compared with other brain wave processing devices, the present invention is easy to carry, and the volume and weight are greatly optimized, which is basically the same as the portability of portable electronic devices such as mobile phones and sports bracelets.

[0021] Strong concealment: The existing alarm devices on the market have a relatively obvious appearance design, which is clearly distinguished from daily carry-on items or clothes, and can be easily noticed by suspects, who then damage or isolate the alarm device when committing a crime. The appearance of the device of the present invention is a headband, and many women currently choose various headbands as a match when going out, so the appearance of the present invention is more in line with daily wearing habits, is not easily discovered by suspects, and can help users promptly transmit distress information to the outside world. Description of the drawings:

[0022] Figure 1 It is a concise block diagram of the overall structure of the present invention;

[0023] Figure 2 It is a block diagram of the alarm signal receiving and detecting unit of the present invention;

[0024] Figure 3 It is a block diagram of the alarm signal processing unit of the present invention;

[0025] Figure 4 is a 3D perspective view of Embodiment 5 of the present invention, wherein the left side is an alarm headband and the right side is an alarm signal processing unit;

[0026] Figure 5 It is a simulated diagram of wearing the alarm headband of the present invention;

[0027] Figure 6 It is a flowchart of the alarm method of the present invention;

[0028] Figure 7 It is a rawdata waveform diagram of a single blink received by the alarm headband of the present invention;

[0029] Figure 8 It is a variance fluctuation diagram of the present invention after the double down-sampling blink detection algorithm is used when the user blinks once;

[0030] Fig. 9 It is a rawdata waveform diagram of the present invention when the user blinks quickly;

[0031] Fig.10 It is a variance fluctuation diagram of the present invention after the double down-sampling blink detection algorithm is used when the user blinks quickly;

[0032] Fig.11 is a scatter plot of the gamma brain wave values ​​detected by the present invention when being frightened;

[0033] Fig.12 It is a scatter plot of the gamma brain wave values ​​detected by the present invention without being frightened. Specific implementation method:

[0034] The present invention is described in detail below with reference to the accompanying drawings and examples:

[0035] Example 1

[0036] In the prior art, the self-defense alarm device needs to manually trigger the alarm process, but in most cases of infringement on personal and property safety, the victim is basically suddenly restricted in personal freedom and cannot manually activate the alarm device, resulting in the inability to transmit the alarm information to the outside world, and the victim is very helpless. The Didi taxi driver assaulted a passenger in 2018 inspired the inventor of this invention, and proposed a variety of ideas and solutions. After many experiments and improvements, a set of non-contact alarm devices and methods triggered by brain waves were gradually formed.

[0037] The present invention is a non-contact alarm device triggered by brain waves, which is provided with an alarm signal receiving and detecting unit, an alarm signal processing unit and an alarm signal sending unit in sequence according to the alarm signal transmission process. They are referred to as the receiving and detecting unit, the processing unit and the alarm unit in short. Figure 1, the physical equipment of the non-contact alarm device triggered by brain waves is divided into an alarm headband, an alarm signal processing unit and an alarm signal sending unit. The alarm signal receiving and detecting unit of the present invention comprises a TGAM brain wave processing module, a micro main control board, a vibration module, a Bluetooth module and a bioelectric acquisition electrode, wherein the bioelectric acquisition electrode comprises an ear clip type potential sensor and a forehead potential sensor, all of which are installed and fixed in the alarm headband; the alarm headband is designed as a double-layer headband with an interlayer, and all components of the alarm signal receiving and detecting unit are installed in the headband interlayer, and the ear clip type potential sensor is led out on the left side of the headband. The forehead potential sensor collects the user's bioelectric signal and transmits it to the TGAM brain wave processing module, which converts the bioelectric signal into brain wave data, and the brain wave data comprises a raw data value and an EEG power value, namely, an EEG Power value, and the user's blink value can be obtained by calculating the raw data value. The TGAM brain wave processing module transmits the user's brain wave data to the alarm signal processing unit through the Bluetooth module, and after being processed by the alarm signal processing unit, it is fed back to the Bluetooth module to send vibration command information to the micro main control board to control the vibration module vibration. The alarm signal processing unit of the present invention includes a main control board, a GPS positioning module, a GPS signal receiving antenna and at least two Bluetooth modules, the two Bluetooth modules are respectively called the first Bluetooth module and the second Bluetooth module, if there are more Bluetooth modules, their names should be arranged in sequence. The brain wave operation processing circuit is formed with the main control board as the core, and all components of the processing unit are installed in a square box. The Bluetooth module in the alarm signal receiving and detecting unit is set to a pairing state with the first Bluetooth module in the alarm signal processing unit, and the user does not need to manually connect again. After the two modules are connected, data will be automatically transmitted. After the entire device is powered on, the Bluetooth module in the receiving and detecting unit will automatically connect to the first Bluetooth module in the processing unit. The second Bluetooth module in the alarm signal processing unit is set to an unpaired state in the overall alarm device, but when in use, the user should pair it with a specific smart phone specified by the user. After pairing, the second Bluetooth module can send alarm command information to the specific smart phone in an emergency. The alarm signal processing unit receives the brain wave signal from the receiving and detecting unit through the first Bluetooth module, and returns the vibration command information to the receiving and detecting unit through the same Bluetooth module. At the same time, the second Bluetooth module transmits the alarm command information generated by the operation of the main control board to the sending unit, and the sending unit sends an alarm to a specific smart phone. All components of the processing unit are in an opaque white square box, because it is convenient to carry all components in the square box and not easy to scatter. The alarm signal sending unit of the present invention is a smart phone, because the smart phone sends text messages quickly and conveniently. The inherent Bluetooth on the smart phone is paired and connected with the second Bluetooth module in the processing unit to receive the alarm command information from the processing unit.The user's smartphone has a specific application installed on it, and the user's designated specific smartphone also needs to have a specific application installed on it. The application contains the user's current location information, current time, and SOS SMS. The location information is included in the alarm instruction information, and the user's current location information is collected by the GPS module in the processing unit. Because this GPS module has an independent antenna, it is more accurate and faster than mobile phone positioning. The user needs to specify in advance the specific smartphone number that receives the alarm information, such as the smartphone number of relatives, friends, etc., to make an alarm. It is enough to make people alert.

[0038] Compared with the existing manually activated alarm device, the present invention does not need manual activation and has the advantages of portability and strong concealment, etc. A new device is added to the alarm system that can successfully alarm without the offender noticing.

[0039] In most criminal cases involving infringement of the personal and property safety of the victim, the victim's personal freedom is suddenly restricted and he cannot use his hands or other limbs to trigger the alarm device. Usually, the outside world cannot know that the victim is in danger. Even if the victim shouts for help, it has little effect in many cases. The present invention provides a technical solution for alarming based on electrooculography and sudden signals without manual contact. The alarm information can be directly sent to the outside world through the blinking frequency of the eyes, which greatly improves the success rate of the alarm.

[0040] The device of the present invention includes a headband device and a data processing module, with a total weight of no more than 500g, which is equivalent to the weight of less than a bottle of water. In daily use, you only need to wear the headband device and carry the data processing module with you. Compared with other brain wave processing devices, the present invention is easy to carry, small in size, and light to carry, and is basically the same as the portability of portable electronic devices such as mobile phones and sports bracelets.

[0041] The existing alarm devices on the market have a relatively obvious appearance design, which is clearly distinguished from daily carry-on items or clothes, and is easily noticed by criminal suspects, who then damage or isolate the alarm device when committing a crime. The appearance of the device of the present invention is a headband, and many women currently choose a variety of styles of headbands as a match when going out, so the appearance of the present invention is more in line with daily wearing habits, is not easily discovered by criminal suspects, and can help users promptly transmit distress information to the outside world.

[0042] Example 2

[0043] The overall structure of the non-contact alarm device triggered by brain waves is the same as that of Example 1, see Figure 2The core component of the alarm signal receiving and detecting unit in the present invention is the TGAM brain wave processing module. The bioelectric signal collected by the forehead potential sensor is first transmitted to the TGAM brain wave processing module. The bioelectric signal is converted into brain wave data in the TGAM brain wave processing module. The brain wave data is transmitted to the Bluetooth module. Because the Bluetooth module of the alarm signal receiving and detecting unit and the first Bluetooth module of the alarm signal processing unit are set to a pairing state, the Bluetooth module of the alarm signal receiving and detecting unit and the first Bluetooth module of the alarm signal processing unit will automatically connect after power is turned on. The brain wave data is automatically sent to the first Bluetooth module of the alarm signal processing unit by the Bluetooth module. The first Bluetooth module transmits the brain wave data to the main control board of the alarm signal processing unit. The main control board runs the startle detection algorithm based on the Euclidean distance and the double down-sampling blink detection algorithm on the received brain wave data. The main control board will generate vibration command information and feedback it to the Bluetooth module of the alarm signal receiving and detecting unit through the first Bluetooth module. The motion command information is transmitted to the micro main control board, and the micro main control board will control the vibration module to vibrate according to the motion command information. The TGAM brainwave processing module collects electrical signals from the human brain through sensors, performs denoising, amplification, and A / D conversion preprocessing, and outputs raw wave data and EEG characteristic values; the bioelectric collection electrodes include a forehead sensor and an ear clip potential sensor, which uses the potential at the earlobe as a reference potential to collect the user's brainwave bioelectric signals.

[0044] The bioelectric signal from the user's earlobe is received and transmitted along the wire to the TGAM brain wave processing module. The TGAM brain wave processing module outputs the user's brain wave data and transmits it to the alarm signal processing unit via the Bluetooth module. The same Bluetooth module in the alarm signal receiving and detecting unit receives the instruction information from the alarm signal processing unit to control the vibration module to enable, so that the vibration module vibrates to provide feedback and remind the user.

[0045] In other words, see Figure 2 The present invention collects the bioelectric signal at the earlobe of the user and transmits it to the TGAM brain wave processing module along the wire. The TGAM brain wave processing module outputs the user's brain wave data and transmits the user's brain wave data to the alarm signal processing unit through the Bluetooth module. The same Bluetooth module in the alarm signal receiving and detecting unit receives the instruction information from the alarm signal processing unit to control the vibration module to enable, so that the vibration module vibrates to remind the user.

[0046] By receiving and analyzing the user's brain waves, the user's blinking frequency and current emotional state are extracted. If the user's blinking frequency exceeds the threshold or the user is emotionally frightened and the user's blinking frequency exceeds the threshold, an alarm will be sounded through the mobile phone buzzer or a distress text message will be sent to a specific smartphone number through the mobile phone APP.

[0047] The entire alarm process does not require manual intervention, and only needs to be controlled by blinking to send the alarm signal. Especially in an emergency, the user can send alarm information to the outside world through this device without any obvious physical operation, which not only transmits information to the outside world, but also ensures their own safety to the greatest extent in an emergency. And the entire alarm process is very fast, ensuring that the alarm information is sent to the outside world before the user's alarm intention is discovered. Reduce the risk of the user being detected by the suspect.

[0048] Example 3

[0049] The overall structure of the non-contact alarm device triggered by brain waves is the same as that of Embodiment 1-2. The alarm signal processing unit of the present invention is an alarm signal processing unit that receives and processes brain wave signals and sends an alarm command signal to the alarm signal sending unit.

[0050] In this example, the alarm signal processing unit includes an Arduino Mega 2560 main control board, a GPS positioning module, a GPS signal receiving antenna, an HC-05 Bluetooth module, and an HC-08 Bluetooth module. The GPS positioning module, the HC-05 Bluetooth module, and the HC-08 Bluetooth module are respectively connected to the Arduino Mega 2560 main control board through wires, and the GPS signal receiving antenna is connected to the GPS positioning module through wires. The GPS positioning module, the HC-05 Bluetooth module, and the HC-08 Bluetooth module each occupy a set of serial ports of the Arduino Mega 2560 main control board. The HC-05 Bluetooth module is used to receive the brain wave data sent by the alarm headband and send vibration command information to the alarm headband, and the HC-08 Bluetooth module is used to send alarm command information to the smartphone and receive reset information sent by the smartphone.

[0051] The alarm signal processing unit of the present invention takes the main control board as the core. The transmitting end of the first Bluetooth module is connected to the receiving end of the third serial port of the main control board, and the receiving end of the first Bluetooth module is connected to the transmitting end of the third serial port. The transmitting end of the second Bluetooth module is connected to the receiving end of the second serial port of the main control board, and the receiving end of the second Bluetooth module is connected to the transmitting end of the second serial port. The transmitting end of the GPS module is connected to the receiving end of the first serial port of the main control board, and the receiving end of the GPS module is connected to the transmitting end of the first serial port, forming a brain wave calculation processing circuit with the main control board as the core.

[0052] Compared with common smart electronic products on the market, the device of the present invention has low power consumption, strong endurance, low cost, and multiple functions are automatically operated without manual debugging. The alarm detection method has been optimized based on the computing power and storage space of existing equipment to ensure that alarm detection can be performed even under low computing power. If the GPS signal disappears, the GPS coordinates closest to the point where the signal disappears will be recorded and the time of loss of connection will be sent, to maximize the accuracy of the alarm distress information.

[0053] Example 4

[0054] The overall structure of the non-contact alarm device triggered by brain waves is the same as that of Examples 1-3. The alarm signal and sending unit in the present invention is the user's smartphone. Existing user smartphones come with apps and buzzers. The emergency contact information needs to be pre-stored in the smartphone App. The emergency contact is a designated specific smartphone. When the user encounters danger, the user's location information will be sent to the emergency contact in the form of a text message. At the same time, the App can set whether the user's smartphone buzzer will sound an alarm. Three working modes are set according to user needs.

[0055] (1) Working mode: warning working mode

[0056] Using a smartphone buzzer to sound a warning can attract the attention of others as quickly as possible and get rescued in time. It can also have a deterrent effect on the suspect and create a split-second escape space.

[0057] (2) Working mode 2 SMS alarm mode

[0058] Use the notification function in the APP and mobile phone text messages to send distress messages to pre-set emergency contacts. The distress messages contain the GPS coordinates of the person in need, the time the distress signal was sent, and the name of the person in need, striving to provide the most detailed information to the person in need so that the person in need can get help in time.

[0059] (3) Working mode: Three-double alarm mode

[0060] At the same time, the smartphone buzzer sound alarm and SMS alarm are used to get help from nearby people and notify emergency contacts of emergency events, doubly protecting the life safety of the person seeking help.

[0061] Using mobile phone APP for emergency alarm is highly operational and convenient for subsequent updates and upgrades of products.

[0062] Example 5

[0063] The present invention is also an alarm detection method for a non-contact alarm device triggered by brain waves, which is used on the above-mentioned non-contact alarm device triggered by brain waves. The overall structure of the non-contact alarm device triggered by brain waves is the same as that of embodiments 1-4, see Figure 6 , including the following steps:

[0064] (1) Wearing the alarm headband to collect brain wave signals in real time: The user wears the alarm headband correctly and the alarm headband is in normal working state, that is, the alarm signal receiving and detection unit is in working state. The user is required to place the sensor contact end in the middle of the forehead, adjust the headband to a suitable tightness, remove obstructions such as hair, and achieve full contact. At the same time, the ear clip contact end also needs to be in full contact with the skin at the earlobe. See the simulated wearing diagram for details. Figure 5 , is a simulation diagram produced by the present invention during the experiment. When the signal value of the data stream Signal in the TGAM brain wave processing module is 0, it is confirmed that the user has worn it correctly and the alarm headband is in normal working condition. The user installs a specific application on his mobile phone and a specified specific smart phone.

[0065] (2) No alarm signal when in stable state: If no emergency occurs, the user's current brain waves will not trigger the alarm process. The entire system will not trigger any operation, the smartphone software will not send an alarm message, and the alarm signal receiving and detection unit will monitor the user's brain waves in real time.

[0066] (3) Use the startle detection algorithm of Euclidean distance to judge abnormalities in sudden state: When a special situation occurs to the user, the EEG data will generate a burst signal. The burst signal is equivalent to the EEG signal in the stable state and can be regarded as an abnormal point. This burst signal is collected by the sensor and is called a burst bioelectric signal. The sudden bioelectric signal is transmitted to the TGAM brain wave processing module, and converted into brain wave data by the TGAM brain wave processing module. The brain wave data is sent to the first Bluetooth module and in the alarm signal processing unit through the Bluetooth module in the alarm signal receiving and detecting unit. The brain wave data will be transmitted to the main control board in the alarm signal processing unit. If the main control board in the alarm signal processing unit calculates the startle detection algorithm based on Euclidean distance and verifies the brain wave data, it will monitor whether the user is currently in a dangerous state. If the brain wave data converted from the sudden bioelectric signal is received, it is in a dangerous state. The main control board will generate vibration command information, which is sent to the micro main control board in the alarm signal receiving and detecting unit through the first Bluetooth module in the alarm signal receiving and detecting unit. After receiving the vibration command information, the micro main control board activates the vibration module to remind the user that the user is currently in the alarm preparation stage.

[0067] (4) Determine whether to alarm in an emergency state: After this, the main control board in the processing unit runs a double down-sampling blink detection algorithm on the brain wave data received from the first Bluetooth module to calculate the user's current blinking frequency, and uses the blinking frequency to determine whether an alarm is necessary. If the user's current blinking frequency reaches the set threshold, the main control board in the processing unit will send an alarm signal to the alarm signal sending unit, i.e., the smart phone, through the second Bluetooth module in the processing unit, and at the same time send a vibration signal to the micro main control board in the receiving and detection unit through the first Bluetooth module again, so that the micro main control board activates the vibration module to remind the user that the alarm signal has been issued. If the user's blinking frequency does not reach the set threshold, the main control board in the processing unit will determine it as a false detection and will not run the subsequent alarm program. If the user is indeed in a dangerous state but does not receive the vibration reminder alarm signal, the processing unit does not detect that the user is currently in a dangerous state, then the user can directly activate the alarm program by blinking quickly, and the processing flow of the alarm program is the same as above.

[0068] (5) Device alarm process in emergency state: When the main control board in the processing unit determines that the current state requires an alarm, the main control board sends an alarm instruction message to the user's smartphone through the second Bluetooth module. After the smartphone receives the alarm instruction message based on its own inherent Bluetooth, it automatically sends a distress message containing the user's current location and time to the specific smartphone number specified in advance by the user through a specific smartphone application. At the same time, the specific smartphone application sends an application message reminder to the specific smartphone specified by the user and continuously shares the location information.

[0069] (6) Reset process after alarm completion: After a complete alarm process is completed, the user needs to enter a specific smartphone application to manually cancel the current alarm status, and then the specific smartphone application will stop sharing location information, and the entire alarm device will be reset. When an alarm is required, the present invention will restart step 2.

[0070] The present invention ensures to the maximum extent that all emergency situations can be detected and prevents false alarms through the double insurance of passive judgment of the startle detection algorithm based on Euclidean distance and active decision-making of the user's blinking operation.

[0071] In most criminal cases involving infringement of the personal and property safety of the victim, the victim is suddenly restricted in his personal freedom and cannot use his hands or other limbs to trigger the alarm device. Therefore, an alarm device that does not require manual intervention is needed to send alarm information to the outside world. This device does not require manual triggering and can directly perform alarm behavior through the blinking frequency of the eyes, greatly improving the success rate of the alarm.

[0072] Example 6

[0073] The non-contact alarm device triggered by brain waves and the alarm detection method thereof are the same as those in Embodiments 1-5. The startle detection algorithm based on Euclidean distance described in step 3 of the alarm detection method is used to detect whether the user is currently in a startled state. The specific algorithm flow is as follows:

[0074] 3.1 Extract fixed-band EEG data: Collect EEG data from the alarm signal receiving and detection unit, extract the intermediate frequency gamma value (intermediate frequency gamma (41-49.75 Hz)) and low frequency gamma value (low frequency gamma (31-39.75 Hz)), and regard them as a two-dimensional point set A1.

[0075] 3.2 Outlier judgment: Calculate the Euclidean distance between the corresponding two points in the point set A1, store it as a one-dimensional matrix X1, sort the one-dimensional matrix X1 in descending order, and regard the first 3% of the data in the sequence in X1 as outliers.

[0076] 3.3 Frightened state judgment: The two-dimensional point numbers of the abnormal points are stored as a one-dimensional matrix X2, and the two-dimensional point numbers in the first quadrant of the one-dimensional matrix X2 are regarded as the user is currently in a frightened state.

[0077] 3.4 Feedback of the startle detection method: Then receive the blink frequency from the double downsampling blink algorithm. If the blink frequency is higher than the threshold, it is confirmed that the user is in a startled state, and the current threshold is stored as a reference value. If the blink frequency is lower than the threshold, it is confirmed that the user is not in a startled state, and the current maximum value of the Euclidean distance, that is, the maximum value of the one-dimensional matrix X1, is stored as a reference value. When the algorithm is performed again, if the calculated threshold is higher than the reference value, the current threshold is used as the actual threshold; if the calculated threshold is lower than or equal to the reference value, the reference value is used as the actual threshold.

[0078] Example 7

[0079] The non-contact alarm device triggered by brain waves and the alarm detection method thereof are the same as those in Embodiments 1-6. The double down-sampling blink detection method described in step 4 of the alarm detection method is used to detect the blink frequency of the user within a unit time. Here, the unit time is defined as 3 seconds. The specific algorithm flow is as follows:

[0080] 4.1 First downsampling: Collect the EEG data from the alarm signal receiving and detection unit, extract the rawdata value from it, and perform the first downsampling, that is, find its peak matrix Y1 in the rawdata data matrix X1. The peak value is defined as: Let the current down index be n. If the slope from position n-1 to n is positive, and the slope from position n to n+1 is negative, then define the value X at the current position n. 1n is the peak value.

[0081] 4.2 Second downsampling: Continue to obtain the peak matrix Y2 in the peak matrix Y1 according to the method defined above, which is the second downsampling.

[0082] 4.3 Calculate variance using sliding window: Use sliding window method to calculate variance of Y2, that is, use 4 as window size in Y2, calculate variance of every 4 data, and store the final value as one-dimensional matrix X2.

[0083] 4.4 Blink judgment and calculation of blink frequency: Find the peak matrix Y3 in X2 and set up a rule. 3n Higher than 1×10 4 If the distance between the previous peak and the current peak is greater than or equal to 3 data points, it is considered that the peak Y is generated. 3n By counting the length of the peak matrix Y3 in a unit time, the number of times the user blinks and the blinking frequency in a unit time can be calculated.

[0084] The present invention determines whether the user is in a dangerous situation based on the startle detection algorithm of Euclidean distance. The algorithm has been verified to be able to detect the brain wave data of most people when they encounter emergency situations. Secondly, the user's current blinking frequency is quickly calculated based on the double downsampling blink detection algorithm. The user needs to blink before the program can send an alarm message to the outside world. Through the active and passive combination of the alarm process, it is guaranteed to the maximum extent that the emergency situation can be detected efficiently while preventing false alarms.

[0085] The following describes in detail the connection and operation process of the device of the present invention to further illustrate the present invention:

[0086] Example 8

[0087] The non-contact alarm device triggered by brain waves and its alarm detection method are the same as those in Example 1-7.

[0088] The alarm headband, i.e., the alarm signal receiving and detecting unit, of the present invention is equipped with a micro main control board, a Bluetooth module, a vibration module, a power supply module, and the like.

[0089] The micro main control board can be selected from: Arduino NANO, STM32F1, STM32F4, STM32F7, Raspberry Pi Zero, Raspberry Pi ZeroW, etc. The alarm signal receiving and detection unit in this example uses Arduino NANO as the micro main control board because this micro main control board has low power consumption, small size and is easy to program.

[0090] The Bluetooth modules available are: HC-05, HC-08, HC-02, HC-06, etc. In this example, the HC-05 Bluetooth module is selected because it has a high transmission rate and can meet the characteristics of large data transmission in this example.

[0091] The power supply module can be selected according to the voltage and size: CR2032 button battery, CR2025 button battery, LR44 button battery, etc. In this example, CR2032 button battery and its corresponding battery box are selected according to the actual situation.

[0092] In this example, any vibration module can be selected as long as it meets the characteristics of small size, high vibration frequency, and low power consumption.

[0093] In general, the device of the present invention includes a headband device and a data processing module, with a total weight of no more than 500g, which is equivalent to the weight of less than a bottle of water. In daily use, you only need to wear the headband device and carry the data processing module with you. After turning on the switch, Bluetooth will automatically connect without other manual operations. Compared with other brain wave processing devices, the present invention is easier to carry and convenient to operate; compared with other self-defense alarm devices, the present invention has high accuracy and a wide range of applications, and the alarm method is more timely and effective. In addition, the appearance of the present invention is closer to life and has high concealment, which can invisibly protect the user's life safety.

[0094] The following is a description of the specific connection and debugging:

[0095] Pin 1 of the HC-05 Bluetooth module is connected to the TX1 pin of the Arduino NANO micro-control board, and pin 2 of the module is connected to the T jack of the TGAM module. Pin 12 of the module is connected to the 3.3V pin of the Arduino NANO micro-control board, and pin 13 is connected to the GND pin on the opposite side of the 3.3V pin of the micro-control board. The other pins of the module are left floating.

[0096] The bioelectric acquisition electrodes are connected to the TGAM brain wave processing module. In this example, the bioelectric acquisition electrodes include an ear clip potential sensor and a forehead potential sensor.

[0097] Connect the "+" pin of the vibration module to the 3.3V pin of the Arduino NANO micro-control board, and the "-" pin to the GND pin on the opposite side of the 3.3V pin. Connect the S pin to the D3 pin of the micro-control board.

[0098] Set this Bluetooth module to "slave mode" with a baud rate of 57600.

[0099] After the power is turned on, the red LED light PWR on the micro main control board is on, indicating that the main control board is powered on successfully. After the HC-05 Bluetooth module is powered on, its blue LED will flash quickly, indicating that the module is in normal working condition. If the HC-05 Bluetooth module on the portable device is also turned on at this time, you need to wait for the flashing frequency of the blue LED of the module to change. If the flashing mode of the blue LED of the module changes to two flashes and one stop, it means that the module has been successfully connected to the HC-05 Bluetooth module of the alarm head.

[0100] The portable device of the present invention, namely the alarm signal processing unit, is equipped with a main control board, a GPS module, a GPS antenna, an HC-05 Bluetooth module, an HC-08 Bluetooth module, a lithium battery and the like.

[0101] The main control board can be: Arduino Mega2560, STM32F104, STM32F107, STM32F407, Raspberry Pi 3B+, etc. The alarm signal processing unit in this example uses Arduino Mega2560 as the main control board because this main control board has low power consumption, small size, multiple input and output interfaces, is easy to connect to multiple devices and is easy to program.

[0102] The first and second Bluetooth modules can be selected from: HC-05, HC-08, HC-02, HC-06, HC-42, etc. In this example, the first Bluetooth module uses the HC-05 Bluetooth module because the first Bluetooth module is compatible with the Bluetooth module in the alarm headband. In this example, the second Bluetooth module uses the HC-08 Bluetooth module because the second Bluetooth module needs to be compatible with electronic devices such as smart phones, and pairing and connection are relatively easy.

[0103] The GPS modules available are: ATGM336H-5N-31, NEO-7N, NEO-6M, GY-NEO6MV2, etc. The GPS in this example uses the ATGM336H-5N-31GPS module because the module has low power consumption, accurate positioning, and is compatible with multiple navigation systems in the world.

[0104] GPS antenna and lithium battery can be selected or customized according to actual conditions. In this example, IPX dual-mode antenna and 6V900mAh lithium battery are selected.

[0105] The following is a description of the specific connection and debugging:

[0106] The 1st pin of the HC-05 Bluetooth module in the alarm signal processing unit is connected to the 15th jack of the Arduino Mega2560 main control board, and the 2nd pin of the module is connected to the 14th jack of the Arduino Mega2560 main control board. The 12th pin of the module is connected to the 3.3V jack of the Arduino Mega2560 main control board, and the 13th pin is connected to the GND jack on the right side of the 3.3V jack of the Arduino Mega2560 main control board. The other pins of the module are left floating.

[0107] Pin 1 of the HC-08 Bluetooth module is connected to jack 17 of the Arduino Mega2560 main control board, and pin 2 of the module is connected to jack 16 of the Arduino Mega2560 main control board. Pin 12 of the module is connected to the 3.3V jack of the Arduino Mega2560 main control board, and pin 13 is connected to the GND jack on the right side of the 3.3V jack of the Arduino Mega2560 main control board. The other pins of the module are left floating.

[0108] The GPS antenna is connected to the IPX port of the ATGM336H-5N-31GPS module through a signal shielded cable.

[0109] The VCC pin of the ATGM336H-5N-31GPS module is connected to the 5V jack of the Arduino Mega2560 main control board, the GND pin of the module is connected to the GND pin on the left side of the 13th pin of the Arduino Mega2560 main control board, the TX pin of the module is connected to the 19th jack of the Arduino Mega2560 main control board, and the RX pin of the module is connected to the 18th pin of the Arduino Mega2560 main control board. The PPS pin of the module is left floating.

[0110] Connect the positive pole of the 6V 900mAh lithium battery to the Vin jack of the Arduino Mega2560 main control board, and the negative pole to the GND jack on the left side of the Vin jack of the Arduino Mega2560 main control board.

[0111] Set the HC-05 Bluetooth module to "host mode" with a baud rate of 57600; set the HC-08 Bluetooth module to "slave mode" with a baud rate of 9600. Use the default settings for the ATGM336H-5N-31GPS module.

[0112] The normal connection of Bluetooth is shown as follows: After the power is turned on, the yellow LED of the GPS module flashes, indicating that the positioning is successful and the positioning information is being sent to the Arduino Mega2560 main control board. The blue LED flashing mode of the HC-05 Bluetooth module changes to two flashes and one stop, indicating that the module has been successfully connected to the Bluetooth module of the alarm headband. The blue LED of the HC-08 module will become permanently on, indicating that the smartphone has been successfully connected to the HC-08 Bluetooth module. After all installations are completed, the 3D perspective view of the alarm signal receiving and detection unit and the alarm signal processing unit can be seen in Figure 4 , Figure 4 It is a 3D perspective view of this embodiment, wherein the left side is the alarm headband and the right side is the alarm signal processing unit.

[0113] After confirming that the system is working properly, the following describes the entire working process:

[0114] The user wears the alarm headband. The correct way to wear it is: fix the metal electrode in front of the headband to the center of the user's forehead and ensure that the magnetic pole contacts the skin with the largest area, and clip the left ear clip of the headband to the left earlobe. After wearing the headband correctly, TGAM will read the user's brain wave data in real time. The TGAM data can be parsed and stored in the Arduino Mega2560 main control board through the TGAM communication code in the Arduino Mega2560 main control board. At this time, the device is in normal state.

[0115] When the Arduino Mega2560 main control board calculates that the user is currently in an abnormal state through the startle detection algorithm based on Euclidean distance, the Arduino NANO will activate the vibration module to remind the user to operate. If the user blinks quickly, and the Arduino Mega2560 main control board calculates that the user's blinking frequency is higher than the established threshold through the double downsampling blink detection algorithm, it is said that the device enters the alarm state at this time. The Arduino Mega2560 main control board will send instructions to the mobile phone APP via Bluetooth, allowing it to use the APP to send a distress text message to the emergency contact, and choose to activate or not activate the mobile phone buzzer according to the settings. After sending the alarm information, the Arduino NANO will activate the vibration module to give the user an alarm feedback, and then the user manually enters the mobile phone APP to cancel the alarm state, and the device returns to normal.

[0116] If the Arduino Mega2560 main control board does not calculate that the user is currently in an abnormal state through the startle detection algorithm based on Euclidean distance, but the user blinks quickly, and the Arduino Mega2560 main control board calculates that the user's blinking frequency is higher than the established threshold through the double downsampling blink detection algorithm, the device will also enter the alarm state. After completing the above alarm operation, the user manually enters the mobile phone APP to cancel the alarm state and return to the normal state.

[0117] If the Arduino Mega2560 main control board calculates that the user is currently in an abnormal state through the startle detection algorithm based on Euclidean distance, but the user does not perform a rapid blinking operation and the Arduino Mega2560 main control board calculates that the user's blinking frequency is not higher than a predetermined threshold through the double downsampling blink detection algorithm, the device is said to enter a calibration state, and the Arduino Mega2560 main control board will record the maximum value of the Euclidean distance in the current startle detection algorithm based on Euclidean distance as a reference value, update the minimum threshold of the startle detection algorithm based on Euclidean distance to the reference value, and then restore the device to a normal state.

[0118] The TGAM brainwave processing module transmits data in the form of a serial data stream, sending approximately 513 data packets per second. The data packets sent are large and small. A small data packet contains a rawdata value, and a large data packet includes a signal strength value, a concentration value, a relaxation value and 8 EEG Power values; the 8 EEG Power values ​​are output in the following order: delta (0.5-2.75Hz), theta (3.5-6.75Hz), low-frequency alpha (7.5-9.25Hz), high-frequency alpha (10-11.75), low-frequency beta (13-16.75Hz), high-frequency beta (18-29.75), low-frequency gamma (31-39.75Hz), and medium-frequency gamma (41-49.75Hz); Under normal circumstances, the TGAM brainwave processing module will transmit all the above data once per second; the data packet is transmitted to the main control board in the alarm signal processing unit via the onboard Bluetooth for further processing.

[0119] The accuracy of the present invention is further described below through experiments and data:

[0120] Example 9

[0121] The non-contact alarm device triggered by brain waves and its alarm detection method are the same as those in Example 1-8.

[0122] Experiment 1: Blink Detection

[0123] Experimental setup: Five adult males and females with normal blinking function were used as experimental subjects. In a resting state, their EEG data were collected within a unit time (3 seconds) for normal blinking, rapid blinking, and no blinking.

[0124] Experimental results: see Figure 7 , Figure 8 , Fig. 9 and Fig.10 .

[0125] Figure 7 It is a rawdata waveform diagram of a single blink, where the horizontal axis is time and the vertical axis is the rawdata value, which is the rawdata value information extracted from the alarm signal processing unit of the present invention. Figure 7 It can be seen that a narrow pulse is generated at 3000ms, which is a sudden change of the rawdata value caused by the blinking activity collected at 3000ms during the test process of the present invention.

[0126] Figure 8This is a variance fluctuation diagram of the present invention after the double downsampling blink detection algorithm is used for a single blink. The horizontal axis is time and the vertical axis is the variance value. This is a variance fluctuation diagram of the signal collected from the alarm signal processing unit of the present invention after the double downsampling blink detection algorithm is used. Figure 8 In the figure, the solid dots are the detected blink points, and the dotted line is the threshold. The algorithm calculates that one blink occurs within this time period.

[0127] Fig. 9 It is a rawdata waveform diagram of rapid blinking of the present invention, the horizontal axis is time, the vertical axis is rawdata value, which is the rawdata value information extracted from the alarm signal processing unit of the present invention, Fig. 9 It can be seen that there are a total of 10 strong fluctuations in the value during this period, which are caused by the fluctuations in the rawdata value caused by ten blinks.

[0128] Fig.10 It is the variance fluctuation diagram of the present invention after the double downsampling blink detection algorithm is used during rapid blinking. Fig.10 In the figure, the solid dots are the detected blink points, and the dotted line is the threshold. As can be seen from the figure, after the algorithm calculation, all 10 strong fluctuations were detected, which means that there were 10 blinks in total during this time period.

[0129] Experiments show that in the case of multiple groups of experiments, the double downsampling blink detection method of the present invention shows universality. Among 136 data points, 125 blink value points are correctly detected, and the detection accuracy is 91.91%, which has engineering practical value.

[0130] Example 10

[0131] The non-contact alarm device triggered by brain waves and its alarm detection method are the same as those in Example 1-8.

[0132] Experiment 2: Startle Detection

[0133] Experimental setup: 40 adults with normal brain function were used as experimental subjects. They were asked to play a computer game called "Scary Maze". The game requires players to use the mouse to control the cursor movement, but are not allowed to touch the specified wall edges. In the third level, the game requires players to pass through a very small gap. Players must concentrate in order to pass the level, but when they are about to pass the level, the game will automatically pop up a horror picture, and the players will be frightened to varying degrees. This experiment will record the brain wave data of the experimental subjects while they are playing the game, and observe through appearance. When the experimental subjects show a frightened state, the current brain wave data value will be marked and recorded as an abnormal brain wave data value.

[0134] Experimental results:

[0135] Fig.11This is a scatter plot of the gamma EEG wave values ​​when being frightened.

[0136] Fig.11 The solid circles in the middle are the detected outliers, which are consistent with the marked values.

[0137] Fig.12 It is a scatter plot of the gamma EEG values ​​when not frightened.

[0138] Table 1 is an outlier detection table obtained by analyzing the data of 40 experimental subjects using the startle detection algorithm based on Euclidean distance.

[0139] Table 1

[0140]

[0141]

[0142] Of the 40 subjects in Table 1, 13 subjects were not frightened or were not obviously frightened, and 27 subjects were frightened. In the entire data verification, 18 subjects’ data accurately judged the marked points or accurately found that there were no marked points; 16 subjects’ data judged that there were multiple outliers, but there were marked points in the outliers; 6 subjects’ data judged incorrectly. The overall accuracy rate was 75%.

[0143] The existing alarm devices on the market have a relatively obvious appearance design, such as a pull-type alarm, which is usually designed as an ellipsoidal alarm body equipped with a pull ring, which is clearly distinguished from daily carry-on items or clothes, and is easily noticed by criminal suspects, so that the alarm device is damaged or isolated when committing a crime. All unit components of the receiving and detecting unit of the present invention are installed in the headband, and the appearance is a headband. Many women currently choose various headbands as a match when going out. The appearance of the present invention is more in line with daily wearing habits, is not easy to be discovered by criminal suspects, and can help users transmit distress information to the outside world in time. More importantly, the present invention directly collects the natural brain waves of the human body through electrodes, and the Bluetooth of the present invention is interconnected with the Bluetooth of the smart phone without manual intervention, so that the alarm information is naturally transmitted to the outside world, and the amplitude of the alarm action is reduced to a minimum, and the concealment is strong. In addition, the present invention also has the advantage of accuracy. The present invention adopts a startle detection algorithm based on Euclidean distance to determine whether the user is in a dangerous situation. It has been verified that the present invention can detect the brain wave data of most people when they encounter an emergency. Secondly, the user's current blinking frequency is quickly calculated based on the double downsampling blink detection algorithm. The user needs to blink before the program can send an alarm message to the outside world. Through the active and passive combination of the alarm process, it is guaranteed to the maximum extent that the emergency situation can be detected efficiently while preventing false alarms.

[0144] In short, the present invention proposes a non-contact alarm device triggered by brain waves and its alarm detection method, which solves the technical problem of non-contact alarm, belongs to the field of intelligent electronic technology, and especially refers to the technology applied to personal safety protection alarm. The TGAM brain wave processing module, micro main control board, vibration module, and Bluetooth module as the receiving and detecting unit are all installed in the headband, and the main control board, GPS positioning module, GPS signal receiving antenna and two Bluetooth modules of the processing unit are installed in the square box. The sending unit is the user's smart phone. The processing unit and the headband are interconnected through a set of Bluetooth, and the processing unit and the user's smart phone are interconnected through a set of Bluetooth. The main function of the receiving and detecting unit is to collect brain waves, the main function of the processing unit is to process and compare the user's brain wave data, and the main function of the sending unit is to send a text message for help to the outside world. The method includes: wearing an alarm headband to collect brain wave signals in real time; the device has no alarm signal in a stable state; using the startle detection algorithm of the Euclidean distance to make an abnormal judgment in an emergency state; judging whether to alarm in an emergency state; the device alarm process in an emergency state; and the reset process after the alarm is completed. The startle algorithm and the blinking algorithm are the core algorithms of the present invention to solve the non-contact alarm. The present invention is a contactless alarm, has the advantages of being able to alarm without manual intervention, low false alarm rate, portability, strong concealment, etc., and is suitable for personal safety and self-defense alarms.

Claims

1. A non-contact alarm device triggered by brain waves, which is provided with an alarm signal receiving and detecting unit, an alarm signal processing unit and an alarm signal sending unit in sequence according to the alarm signal transmission process; referred to as the receiving and detecting unit, the processing unit and the alarm unit; characterized in that: The physical equipment of the non-contact alarm device triggered by brain waves is divided into an alarm headband, an alarm signal processing unit and an alarm signal sending unit; the alarm signal receiving and detecting unit comprises a TGAM brain wave processing module, a micro main control board, a vibration module, a Bluetooth module and a bioelectric acquisition electrode, wherein the bioelectric acquisition electrode comprises an ear clip type potential sensor and a forehead potential sensor, all of which are installed and fixed in the alarm headband; the alarm headband is designed as a double-layer headband with an interlayer, all components of the alarm signal receiving and detecting unit are installed in the interlayer of the headband, and the ear clip type potential sensor is led out on the left side of the headband; the forehead potential sensor is used for collecting The TGAM brainwave processing module collects bioelectric signals and converts the bioelectric signals into brainwave data. The brainwave data includes a raw wave value and a brainwave power value. The blink value can be calculated from the raw wave value. The TGAM brainwave processing module transmits the user's brainwave data to the alarm signal processing unit through the Bluetooth module. After being processed by the alarm signal processing unit, the alarm signal processing unit feeds back the data to the Bluetooth module to send vibration command information to the micro main control board to control the vibration module to vibrate. The alarm signal processing unit includes a main control board, a GPS positioning module, a GPS signal receiving antenna and at least two Bluetooth modules. The two Bluetooth modules are respectively called The first Bluetooth module and the second Bluetooth module form a brain wave operation and processing circuit with the main control board as the core, and are all installed in a square box; the Bluetooth module in the alarm signal receiving and detecting unit is set to a paired state with the first Bluetooth module in the alarm signal processing unit; the second Bluetooth module in the alarm signal processing unit is set to an unpaired state and can be paired with any smart phone; the alarm signal processing unit receives the brain wave signal from the alarm signal receiving and detecting unit through the first Bluetooth module, and returns the vibration command information to the alarm signal receiving and detecting unit through the same Bluetooth module, and at the same time, the second Bluetooth module transmits the alarm command information generated by the main control board operation to the alarm signal sending unit for alarming, the alarm signal sending unit is a smart phone, the inherent Bluetooth on the smart phone is paired and connected with the second Bluetooth module in the alarm signal processing unit, and receives the alarm command information from the alarm signal processing unit; a specific application is installed on the user's smart phone, and a specific application must also be installed on the user's designated specific smart phone, and the application contains the user's current location information, current time, and emergency text message; the user needs to specify the specific smart phone number that receives the alarm information in advance to alarm; it is enough to alert people; The core component of the alarm signal receiving and detecting unit is the TGAM brain wave processing module. The bioelectric signal collected by the forehead potential sensor is first transmitted to the TGAM brain wave processing module. The bioelectric signal is converted into brain wave data in the TGAM brain wave processing module. The brain wave data is transmitted to the Bluetooth module. The brain wave data is sent by the Bluetooth module to the first Bluetooth module of the alarm signal processing unit. The first Bluetooth module transmits the brain wave data to the main control board of the alarm signal processing unit. The main control board runs the startle detection algorithm based on Euclidean distance and the double down-sampling blink detection algorithm on the received brain wave data, generates a command signal, and feeds it back to the Bluetooth module of the alarm signal receiving and detecting unit through the first Bluetooth module. The alarm signal processing unit is a portable processing unit that receives and processes brain wave signals and sends an alarm command signal to the alarm signal sending unit; The transmitting end of the first Bluetooth module is connected to the receiving end of the third serial port of the main control board correspondingly, and the receiving end of the first Bluetooth module is connected to the transmitting end of the third serial port correspondingly; The transmitting end of the second Bluetooth module is connected to the receiving end of the second serial port of the main control board, and the receiving end of the second Bluetooth module is connected to the transmitting end of the second serial port; The transmitting end of the GPS module is connected to the receiving end of the first serial port of the main control board, and the receiving end of the GPS module is connected to the transmitting end of the first serial port, forming a brain wave calculation processing circuit with the main control board as the core.

2. The non-contact alarm device triggered by brain waves according to claim 1, characterized in that: The alarm signal sending unit is the user's smartphone, which comes with an app and a buzzer. Three working modes can be set according to user needs; (1) Working mode: warning working mode Use smartphone buzzers to give audible warnings to act as a deterrent; (2) Working mode 2 SMS alarm mode Use text messages to alert the outside world and report the incident secretly; (3) Working mode: Three-double alarm mode At the same time, the smartphone buzzer sound alarm and SMS alarm are used.

3. An alarm detection method for a non-contact alarm device triggered by brain waves, used on any non-contact alarm device triggered by brain waves as claimed in any one of claims 1 to 2, characterized in that: The steps include: (1) Wearing an alarm headband to collect brain wave signals in real time: The user wears the alarm headband correctly and the alarm headband is in normal working condition. In the working state, the user is required to place the sensor contact end in the middle of the forehead, adjust the headband to a suitable tightness, remove hair and other obstructions, and achieve full contact. At the same time, the ear clip contact end also needs to be in full contact with the skin at the earlobe. When the data stream Signal value in the TGAM brain wave processing module is 0, it is confirmed that the user has worn it correctly and the alarm headband is in normal working condition; the user installs a specific application on his mobile phone and a specified specific smartphone; (2) The device has no alarm signal in a stable state: If no emergency occurs, the user's current brain waves will not trigger the alarm process; the entire system will not trigger any operation, the smartphone software will not send an alarm message, and the alarm signal receiving and detection unit will monitor the user's brain waves in real time; (3) In an emergency state, the startle detection algorithm based on Euclidean distance is used to make abnormal judgments: When a special situation occurs to the user, the brain wave data will generate a burst signal. The burst signal is the bioelectric signal converted into brain wave data by the TGAM brain wave processing module. The brain wave data is sent to the first Bluetooth module and the main control board in the alarm signal processing unit through the Bluetooth module in the alarm signal receiving and detecting unit. The main control board calculates the startle detection algorithm based on Euclidean distance and verifies the brain wave data to monitor whether the user is currently in a dangerous state. If the received bioelectric signal is converted into brain wave data, it is in a dangerous state. The main control board will generate vibration command information, which is sent to the micro main control board in the alarm signal receiving and detecting unit through the first Bluetooth module in the alarm signal receiving and detecting unit. After receiving the vibration command information, the micro main control board activates the vibration module to remind the user that it is currently in the alarm preparation stage; (4) Determine whether to alarm in an emergency state: The main control board runs a double down-sampling blink detection algorithm on the brain wave data received from the first Bluetooth module to calculate the user's current blinking frequency, and uses the blinking frequency to determine whether an alarm is necessary; if the user's current blinking frequency reaches the set threshold, the main control board will send an alarm signal to the alarm signal sending unit, i.e., the smartphone, through the second Bluetooth module in the processing unit, and at the same time send a vibration signal to the micro main control board in the receiving and detection unit through the first Bluetooth module again, so that the micro main control board activates the vibration module to remind the user that the alarm signal has been issued; if the user's blinking frequency does not reach the set threshold, the main control board in the processing unit will determine that it is a false detection and will not run the subsequent alarm program; if the user is indeed in a dangerous state but does not receive a vibration reminder, the user can directly activate the alarm program by blinking quickly, and the processing flow of the alarm program is the same as above; (5) Device alarm process in emergency state: When the main control board in the processing unit determines that the current state requires an alarm, the main control board sends an alarm instruction message to the user's smartphone through the second Bluetooth module; after the smartphone receives the alarm instruction message based on its own inherent Bluetooth, it automatically sends a distress message containing the user's current location and time to the specific smartphone number specified in advance by the user through a specific smartphone application; (6) Reset process after alarm completion: After a complete alarm process is completed, the user needs to enter a specific smartphone application to manually cancel the current alarm status, and then the specific smartphone application will stop sharing location information and the entire alarm device will be reset.

4. The alarm detection method of the non-contact alarm device triggered by brain waves according to claim 3, characterized in that: The startle detection algorithm based on Euclidean distance described in step 3 is used to detect whether the user is currently in a startled state. The specific algorithm flow is as follows: 3.1 Extracting fixed-band brain wave data: Collect brain wave data from the alarm signal receiving and detection unit, extract the Middleγ value and Lowγ value from it, and regard it as a two-dimensional point set A1; 3.2 Outlier judgment: Calculate the Euclidean distance between two corresponding points in the point set A1, store it as a one-dimensional matrix X1, sort the one-dimensional matrix X1 in descending order, and regard the first 3% of the data in the sequence in X1 as outliers; 3.3 Frightened state judgment: the two-dimensional point numbers of the abnormal points are stored as a one-dimensional matrix X2, and the two-dimensional point numbers in the first quadrant of the one-dimensional matrix X2 are regarded as the user is currently in a frightened state; 3.4 Feedback of startle detection method: Then the blink frequency from the double down-sampling blink algorithm is received. If the blink frequency is higher than the threshold, it is confirmed that the user is in a startled state, and the current threshold is stored as a reference value; If the blinking frequency is lower than the threshold, it is confirmed that the user is not in a frightened state, and the current maximum value of the Euclidean distance, that is, the maximum value of the one-dimensional matrix X1, is stored as a reference value; when the algorithm is performed again, if the calculated threshold is higher than the reference value, the current threshold is used as the actual threshold; if the calculated threshold is lower than or equal to the reference value, the reference value is used as the actual threshold.

5. The alarm detection method of the non-contact alarm device triggered by brain waves according to claim 3, characterized in that: The double down-sampling blink detection algorithm described in step (4) is used to detect the blink frequency of the user within a unit time. Here, the unit time is defined as 3 seconds. The specific algorithm flow is as follows: 4.1 First downsampling: Collect brain wave data from the alarm signal receiving and detection unit, extract the rawdata value from it, and perform the first downsampling, that is, find its peak matrix Y1 in the rawdata data matrix X1. The peak value is defined as: let the current down index be n, if the slope from position n-1 to n is positive, and the slope from position n to n+1 is negative, that is, define the value X1n at the current position n as the peak value; 4.2 Second downsampling: Continue to obtain the peak matrix Y2 in the peak matrix Y1 according to the above definition method, that is, the second downsampling; 4.3 Sliding window variance calculation: The sliding window method is used to calculate the variance of Y2, that is, the window size is 4 in Y2, and the variance is calculated for every 4 data, and the final value is stored as a one-dimensional matrix X2; 4.4 Blink judgment and calculation of blink frequency: Find the peak matrix Y3 in X2 and set up rules. If the peak value Y3n is higher than 1*10^4 and the distance between the previous peak and the current peak is greater than or equal to 3 data points, it is determined that the user blinked when the peak value Y3n was generated; by counting the length of the peak matrix Y3 in unit time, the number of times and frequency of the user's blinks in unit time can be calculated.

Images