The content of the invention
According to an aspect of the invention, there is provided a kind of intelligence tumble monitor device, including：Alarm modules, signal are adopted
Collect module and signal processing module, wherein,
The signal acquisition module includes 3-axis acceleration sensor, baroceptor and pressure sensor, for gathering
Human body attitude real time data, pressure altitude data and pressure data；
The signal processing module includes fall detection unit, and the fall detection unit is arranged to according to the acquisition
Human body attitude real time data, pressure altitude data and pressure data are analyzed and processed, and to judge human body tumble state, work as judgement
The first signal is exported when human body is fallen to alarm modules；
The alarm modules are according to first signal generation and export warning information.
The monitor device of the present invention carries out fall detection by gathering 3-axis acceleration, height and pressure data, and synthesis is examined
Consider the behavioural characteristic of human body and other environmental characteristics, the accuracy higher of fall detection.
In some embodiments, the signal acquisition module is additionally operable to harvester wearing information, the signal processing
Module may also include device and wear detection unit, and described device wears detection unit and is arranged to be divided according to device wearing information
Analysis is handled, output wearing state mark；The fall detection unit is identified according to the wearing state of the output, and detection human body falls
State and to the alarm modules export the first signal；Wherein, described device wearing information includes 3-axis acceleration signal, wears
Wear combination more than one of site pressure signal, temperature signal and human biological signal or both.Thus, it is possible to
It detects whether wearing direction and/or the wearing position of device are correct, just carries out fall detection when device wears correct, realize
It avoids wearing the bad of incorrect caused wrong report, further improves the accuracy rate of fall detection.
In some embodiments, the signal acquisition module further includes temperature sensor and/or human biological electricity sensing
Device, described device wear acceleration signal, the pressure sensing that detection unit is gathered according to the 3-axis acceleration sensor
The human body of the pressure signal of device acquisition, the temperature signal of temperature sensor acquisition and human-body biological electric transducer acquisition
Combinations more than one of bioelectrical signals or both, is analyzed and processed, output wearing state mark.Thus, it is possible to
Further detection device whether there is the situation of wearing, realizes that avoiding device does not have to cause the bad of wrong report under wear condition, improves dress
Put the accuracy rate of wearing and fall detection.
In some embodiments, described device further includes locating module and wireless communication module, and the locating module is adopted
Collect the location information of the monitor device, the alarm modules send the alarm signal containing location information by wireless communication module
It ceases to remote terminal.Thus, it is possible to when falling risk occurs, guardian is notified in time, and location information is supplied to monitoring
People is effectively timely helped with obtaining.
In some embodiments, the signal processing module further includes state detection unit after tumble, is arranged to basis
The data of the signal acquisition module acquisition, the body state after detection tumble.When judging that human body releases tumble state, to institute
It states alarm modules and sends secondary signal, the alarm modules are sent by wireless communication module and departed from according to the secondary signal
The information of tumble is to remote terminal.Thus, it is possible to further detect the disposition after user falls, fall to depart from user
Guardian is notified in time after state, is brought conveniently for guardian, is effectively improved user experience.Also, police is released in time
Report, it is possible to reduce the consumption of device in alarm reduces the power consumption of device.
In some embodiments, the alarm modules are broadcast including carrying out emergency voice according to the first signal enabling loud speaker
It reports and is stopped playing loud speaker distress signal according to secondary signal.Thus, it is possible to realize while notifying to cry for help to guardian,
Emergency is sent to people around, so that user is succoured in time, is out of danger, increases safety guarantee.
In some embodiments, which further includes human-computer interaction module, is arranged to through touch-screen, button or voice
User's input is received, typing user basic information is inputted according to user or starts the alarm modules and is cried for help or released emergency
Processing.User can be by touch-screen typing user information, device to be facilitated to be carried out according to user information at analysis as a result,
Reason can also realize that user is cried for help or released wrong report emergency by button, can be effective when detection wrong report occurs for device
Improve user experience, it is efficient and convenient.
According to another aspect of the present invention, a kind of processing method of intelligent tumble monitor device, the monitoring are additionally provided
Device includes alarm modules, signal acquisition module and signal processing module, the treating method comprises：
Signal acquisition module gathers user behavior information data in real time, exports to signal processing module, the user behavior
Information data includes 3-axis acceleration data, pressure altitude data and pressure data；
Signal processing module is analyzed and processed according to the user behavior information data, judges human body tumble state, when
When judging that human body is fallen, the first signal of output to alarm modules；
Alarm modules are according to first signal generation and export warning information.
The method of the present invention carries out fall detection by gathering 3-axis acceleration, height and pressure data, considers people
The behavioural characteristic of body and other environmental characteristics, the accuracy higher of fall detection.
In some embodiments, described information processing module is carried out according to the user behavior information data at analysis
Reason, judges that human body tumble state includes：
A, the acceleration information of 3-axis acceleration sensor acquisition, baroceptor in intervals is stored to gather
Pressure altitude data and the pressure data of pressure sensor acquisition；
B, the data for whether occurring great change in the intervals are judged according to the acceleration information of storage, according to huge
Become the time point of data, the trunk direction in a period of time interior a period of time with after great change before calculating great change
Before and after the sum of 3-axis acceleration variation in a period of time after acceleration average AY_1 and AY_2, great change ACC_SUM, great change
Human body pressure altitude difference H2-H1 and great change before and after the sum of the pressure differential of body side and opposite side ∑ | Pi1-Pi2 |；
C, given threshold TH3, TH4, TH5 judges whether to meet condition H2-H1>TH3、ACC_SUM<TH4 and ∑ | Pi1-
Pi2|>TH5, and AY_1, close to gravity acceleration g, AY_2 is close to 0, if it is satisfied, then judging that human body is fallen.
The attitudes vibration for judging the tumble of human body by 3-axis acceleration as a result, and lying low judges to fill by altitude information
Height change from the ground is put, judges that the wearing site compression dynamics of one week changes by pressure data, user's appearance can be combined
State and behavioural characteristic, more accurately detect whether human body falls.
In some embodiments, the step b includes：
According to the 3-axis acceleration data in the intervals of the storage, to the 3-axis acceleration number gathered every time
According to, its 3-axis acceleration amplitude ACC is calculated, judges whether ACC is more than the threshold value TH1 set, when more than given threshold, judgement
The time point of this time data acquisition is the time point for occurring data great change；
According to the 3-axis acceleration data of the storage, the acceleration information variable quantity in great change surrounding time section is calculated
ACC_CHG, judges whether the acceleration information variable quantity ACC_CHG is more than given threshold, if greater than given threshold, then sentences
Disconnected great change surrounding time section is the moment fallen；
According to the 3-axis acceleration data of the storage, adding for the trunk direction in a period of time before great change is calculated
One after the acceleration average AY_2 in the trunk direction in a period of time after speed data average AY_1, great change, great change
The front and rear pendant of human body pressure altitude value H2-H1 and great change before and after the sum of 3-axis acceleration variation in the section time ACC_SUM, great change
Wear the sum of body side and the opposite side pressure differential at device position ∑ | Pi1-Pi2 |.
Thus, it is possible to according to the big ups and downs of data, the time that human body is fallen is judged, obtain front and rear data of falling
It is compared, to determine whether human body falls.
In some embodiments, the method may also include：Signal acquisition module real-time acquisition device wearing information number
According to output to signal processing module, the signal processing module is analyzed and processed according to device wearing information data, judges to fill
Put wearing state, output device wearing state mark；Signal processing module reads the wearing state mark and is judged, works as pendant
When wearing status indicator correctly to wear, human body fall detection is carried out, and exports the first signal to alarm modules.It is detecting as a result,
When correctly being worn to device, human body fall detection is just carried out, it can be to avoid because device is not worn or wear incorrect or slides
When lead to the problem of wrong report, further improve the accuracy rate of detection.Moreover, only user is just carried out when device wears correct
The detection that the acquisition of behavioural information data and human body are fallen, can reduce unnecessary data processing operation, improve efficiency, save
In some embodiments, the signal acquisition module further includes temperature sensor and/or human biological electricity sensing
Device, described device wearing information data include temperature data, human biological signal, wearing site pressure data and three axis and accelerate
Combination more than one of degrees of data or both.
In some embodiments, the signal processing module carries out following A to D according to described device wearing information data
Combination more than one of analyzing and processing or both is identified with generation and output device wearing state：
A, temperature of the temperature signal T1 and equipment close to human body one side exposed to one side in air of temperature sensor is read
Signal T2 is spent, when the temperature difference for judging T1 and T2 is more than given threshold, wearing state is set to be identified as correct wearing, is otherwise set
Put wearing state be identified as wear it is wrong；
B, the human biological signal of human-body biological electric transducer output is read, is judging human biological signal for high electricity
Usually, wearing state is set to be identified as correct wearing, wearing state is otherwise set to be identified as wearing wrong；
C, according to typing acquisition of information user waistline value L, the pressure sensor space D being distributed according to matrix form, which calculates, to be generated
The number N=L/D of the sensor of pressure reads pressure signal P 1 ... the PN of N number of pressure sensor, is judging N number of pressure sensing
When the pressure signal of device is equal to given threshold, wearing state is set to be identified as correct wearing, otherwise wearing state is set to identify
It is wrong to wear；
D, the acceleration value on three direction of principal axis of 3-axis acceleration sensor is read, is judging to represent trunk direction
Acceleration value be g and when other two acceleration values are 0, wearing state is set to be identified as correct wearing, otherwise sets and wears
Status indicator is wrong to wear；
Wherein, when carrying out the combinatory analysis processing of more than the two, it is judged as with mistaking, that is, setting under either type
To wear wrong mark.
Judge whether human body has worn device by temperature and/or human biological signal thus, it is possible to realize, pass through pressure
Whether power and acceleration judge whether the position and direction of human body wearable device are correct, so as to fulfill being worn to device and wearing is
No correct detection avoids reporting by mistake due to device problem.
In some embodiments, described device further includes locating module and wireless communication module, and the method further includes：
The locating module gathers the location information of the monitor device, and the alarm modules send alarm signal by wireless communication module
For breath to remote terminal, the warning information includes the location information obtained from the locating module and emergency content.Thus, it is possible to
Location information and emergency content of falling are sent to guardian in time, timely and effectively to be succoured.
In some embodiments, this method may also include：
When human body is in tumble state, the signal acquisition module continuous collecting 3-axis acceleration data, pressure altitude
Data and pressure data simultaneously store；
Signal processing module is divided according to the 3-axis acceleration data, pressure altitude data and pressure data of real-time update
Analysis is handled, and judges that human body releases the state fallen, and when judging that human body releases tumble, secondary signal, institute are sent to alarm modules
Alarm modules are stated according to secondary signal, is sent by wireless communication module and departs from the information fallen to remote terminal.
User Status is continued to monitor after tumble thus, it is possible to realize, so that after clearing, notice is supervised in time
People is protected, is saved time and the nervous anxiety of reduction guardian to guardian, more hommization, user experience is more preferable.
In some embodiments, the signal processing module is high according to the 3-axis acceleration data of real-time update, air pressure
Degrees of data and pressure data are analyzed and processed, and judge that human body releases the state fallen and includes：
Calculate the acceleration average AY_3 of the axis in the representative trunk direction after falling in a period of time；
Judge acceleration average AY_3, pressure altitude value H3 and pressure value P 13, P23 ..., whether PN3 meets condition：|
AY_3-g|<TH6、H3-H2>TH7 and P13=P23=...=PN3, wherein, TH6 and the threshold value that TH7 is setting such as meet item
Part then judges that human body releases tumble state.
Whether as a result, by the acceleration average in trunk direction close to g, whether device height becomes larger and wears
It wears whether site pressure data become uniform, and judges whether human body stands up from tumble state, so as to fall in human body releasing
When, guardian is notified at the first time.
In some embodiments, this method may also include：It receives and is inputted by the signal of button, broadcasting/pause is raised one's voice
Device alarm and transmission distress signals/releasing distress signals to remote terminal.Thus, it is possible to realize when reporting by mistake, used
The active process at family, to reduce the adverse consequences that wrong report is brought.
The invention will now be described in further detail with reference to the accompanying drawings.
Fig. 1 schematically shows a kind of appearance of the intelligent human-body tumble monitor device of embodiment according to the present invention
Structure.As shown in Figure 1, the device includes band body 1 and buckle 2, band body 1 is fixedly connected with 2 one end of buckle, and the other end can be waist
It is fastened during band, connects and fasten mode with common waistband.Button 3 is provided on buckle 2, user can be by pressing lower button 3
Carry out human-computer interaction.The pressure sensor 5 of monitor device is uniformly distributed along band body 1, and remaining function module (such as signal acquisition
Other sensors, signal processing module, locating module, wireless communication module of module etc.) all it is built in the integrated core in buckle 2
On piece 4.In practical applications, user can realize daily monitoring to user by wearing waistband shown in FIG. 1.Due to
Waistband belongs to the daily dressing requirement of most people, easy to carry, without any subsidiary sense, is not easy to forget, very convenient.
What Fig. 2 schematically showed the present apparatus is built in the frame structure in vivo each module.It as shown in Fig. 2, should
Device includes signal processing module 20, locating module 21, signal acquisition module 22, wireless communication module 23 and alarm modules 24.
Locating module 21 is realized using GPS or the positioning methods such as the Big Dipper or mobile base station, for providing the geographical location information of user.Nothing
Line communication module 23 is that GSM communication units or bluetooth communication unit etc. can be led to mobile terminal device wirelessly
The chip or module of news, the data interaction being used to implement between remote terminal (such as mobile phone, computer, IPad device ends).
Alarm modules 24 are arranged to, when receiving the first signal (such as distress signal) corresponding with tumble, pass through wireless communication module 23
Corresponding warning information (such as distress signals) is sent to remote terminal, to notify guardian user occurs to fall to need to succour, phase
The distress signals answered may include locating module 21 geographical location information obtained and content of specifically crying for help.Signal acquisition module 22
For gathering user behavior information data in real time, signal processing module 20 is supplied to carry out human body fall detection analysis.Signal is adopted
Collection module 22 is mainly realized by multiple sensors, includes but not limited to 3-axis acceleration sensor, baroceptor and pressure
Sensor, 3-axis acceleration sensor for gathering human body attitude data, for gathering pressure altitude data press by baroceptor
Force snesor is used to gather pressure data.Signal processing module 20 is the microprocessors such as MCU.Wherein, signal processing module 20
Including fall detection unit 202, fall detection unit 202 is used for the human body attitude data gathered according to signal acquisition module 22
(acceleration information AX, AY, AZ for including three axial directions), pressure altitude data and pressure data carry out fall detection, work as judgement
When falling, distress signal (i.e. the first signal) is exported to alarm modules 24, is led to distress signals with starting alarm modules 24
Cross the remote terminal equipment that wireless communication module 23 is sent to guardian.When being fallen due to human body, 3-axis acceleration can be
Falling down moment has of short duration " great change ", then for some time can relatively " static ", and pressure altitude data have before and after tumble
Certain difference in height, pressure data the one side of human body ground proximity with away from the one side on ground can it is different according to stressing conditions and
Certain difference, fall detection unit 202 can be carried out the detection and sentence whether human body fall according to these three data
It is disconnected.
Fig. 9 schematically shows the 3-axis acceleration data line chart that a kind of human body of typical case is fallen.The
Acceleration information line chart when one section 90 is human body normal stand, second interval 91 are rolled over for the acceleration information of state of weightlessness
Line chart, 3rd interval 92 are acceleration information line chart when falling, and the 4th section 93 is in a period of time after tumble
Acceleration information line chart.As shown in figure 9, when human body is fallen, when striking ground, 3-axis acceleration signal can go out
Very violent one piece of data is now fluctuated, is human body collision ground if the 3rd interval 92 of acute variation occurs for signal in Fig. 9
Moment, the present invention be referred to as " great change " section.Acceleration information after analysis tumble occurs in a period of time, normal condition
Under, one section of opposing stationary section is had after tumble, static section (the 4th section) 93 as shown in Figure 9, in the 4th section 93
Interior, the acceleration for representing the Y-axis in trunk direction is essentially close to 0 (because human body is by uprightly becoming for the posture that lies low), this
Invention is referred to as " static " section.
In use, signal acquisition module 22 gathers user behavior data (including human body attitude data, air pressure in real time
Altitude information and wearing site pressure data), and pass through FIFO (First In First Out, first in first out) form storage one
Human body attitude data (i.e. 3-axis acceleration data), pressure altitude data and pressure data in section time (such as 4 seconds).It falls and examines
Unit 202 is surveyed according to the 3-axis acceleration data of storage, analyses whether occur fluctuating whether very violent one piece of data (goes out
Existing " great change "), be specially：The section given threshold TH1 of " great change " is occurring, by calculating the 3-axis acceleration number gathered every time
3-axis acceleration amplitude ACC is obtained according to the vectorial mould of AX, AY, AZ, that is, is had
Judge whether the amplitude ACC of the 3-axis acceleration data gathered every time is more than the threshold value TH1 of setting, when 3-axis acceleration amplitude is big
When the threshold value of setting, the time point for judging this gathered data is the time that " great change " occurs.According to generation " great change "
Time point reads the 3-axis acceleration in the representative trunk direction of (such as 1 second) storage in a period of time before " great change " occurs
Data AY according to the acceleration value AY of the Y-axis in the representative trunk direction of acquisition, calculates the equal of the Y-axis before " great change " occurs
Value(n is the number of the acceleration information gathered in a period of time before " great change ").Meanwhile it records
The pressure altitude data H1 gathered for the last time before " great change " and pressure data (P11, P21 ... ..., PN1) occurs.In order to further
Confirm whether human body falls in great change, generation " great change " can be recorded according to the gathered data of storage before and after moment in the period
Acceleration information AX, AY, the AZ of (before such as great change after 0.1s- great changes in 0.1s) are calculated according to the acceleration information of record and accelerated
Variable quantity ACC_CHG is spent, calculation formula is：
(n is the number of the acceleration information gathered in the transitory period of time).Given threshold TH2 such as could be provided as 2g, judge to calculate
Whether the acceleration change amount ACC_CHG of obtained " great change " moment meets ACC_AHG>TH2 meets explanation in this moment
Period human body is fallen, and human body then records into the quiescent time section after great change " static after this tumble moment
3-axis acceleration data AX, AY, the AZ of (in a period of time i.e. after great change after 0.1s) in section ", pressure altitude value H2 with
And the pressure value (P12, P22 ... ..., PN2) of wearing site.The representative trunk in the section is calculated according to the data of record
The average of the acceleration of the Y-axis in directionWith in the section 3-axis acceleration variation the sum ofN is to be gathered in static section
The number of acceleration information.Setting device height threshold TH3 from the ground, stationary state acceleration rate threshold TH4 and pressure differential threshold
Value TH5 judges whether the difference in height of the pressure altitude value H2 in pressure altitude value H1 and static section before great change meets and is more than
Whether the threshold value TH3 of setting, i.e., meet H2-H1>Whether the sum of TH3, the acceleration change in static section ACC_SUM, which are less than, sets
Whether fixed threshold value TH4, i.e., meet ACC_SUM<TH4.The judgment means wearing site pressure condition of one week simultaneously, if exist
The sum of the pressure value of the pressure sensor of one side and the pressure differential of pressure value of opposite side are more than the threshold value TH5 of setting, i.e. ∑ |
Pi1-Pi2|>TH5.And judge whether the acceleration average AY_1 of the Y direction before great change (illustrates to use close to gravity acceleration g
Family is standing state), whether the acceleration average AY_2 of static section Y direction is close to 0 (user is lying status).Wherein,
TH3 can be arranged to waist to the height such as 80cm of ankle according to human body information data, if it is satisfied, then explanation before great change and
Static section, the height of human body is from uprightly becoming bending or lying status；TH4 is the tranquil period after falling, this when
Phase acceleration change amount is very small, could be provided as smaller value, such as levels off to 0.1g (g is acceleration of gravity).If it is satisfied,
Then within a period of time after human height is lower by height, human body remains static explanation, and after generation of normally falling,
The situation can all occur before tumble posture is departed from；Land when TH5 falls according to human body one side pressure value with away from ground one
The sum of difference of pressure value of side is configured, and human body is illustrated if meeting after height changes and enters stationary state,
There is one side to land.It is that human body is fallen thus, it is possible to judge to meet simultaneously when three conditions, then human body is fallen shape
State mark such as FALL_DOWN_FLAG is arranged to TRUE, while sends distress signal (such as character " 1 ") to alarm modules 24, so as to
Start alarm, into help-asking mode.The geographical location information that alarm modules 24 are provided according to locating module 21, generation include geography
The distress signals of location information and emergency content are sent to the remote terminal of guardian by wireless communication module 23, are led to
Know, to obtain relief.The fall detection mode that the embodiment provides, it is necessary to detect the variation of pressure altitude data, accelerate simultaneously
Degree variation and one week pressure data of wearing site change, can be than more comprehensively considering the behavioural characteristic and data of user, phase
To the detection mode of single acceleration or angle change, Detection accuracy higher of the invention is more effective, so that user exists
Emergency request can be sent after falling at the first time, is succoured.
Simultaneously, it is contemplated that in addition to the accuracy rate in terms of algorithm, device wear condition is also influence Detection accuracy one
A key factor causes the solution of wrong report invention also provides device under no wear condition or under the incorrect situation of wearing
Certainly scheme.As shown in Fig. 2, device is further included in signal processing module 20 wears detection unit 201.Device wears detection unit
The 201 device wearing information data for being arranged to be gathered according to signal acquisition module are (including 3-axis acceleration data and wearing site
The pressure data of one week) it is analyzed and processed, output wearing state control signal to fall detection unit 202.Fall detection list
The wearing state control signal of 202 output of member carries out fall detection, correct when being worn in device, gathers user behavior
Data carry out analysis detection, and distress signal is exported to alarm modules 24 when falling.
When in use, after user's starter, signal acquisition module 21 gathers pressure data P1, P1 ... ... incessantly,
PN.Device wear detection unit 201 reduced pressure data P1, P1 ... ..., PN, if human body do not wear or wear it is elastic not
It meets the requirements, such as too loosening or wearing site inaccuracy etc., each data of pressure sensor just have larger difference, such as
Fruit is correctly worn, and the value of pressure data should substantially meet P1=P2=...=PN=P, and wherein P is correctly to be worn on waist
Elasticity pressure value during portion.It is distributed in it should be noted that pressure sensor is matrix form on waistband, number N=L/
D, wherein, L is the waistline information (being obtained according to the essential information of user's typing) of user, and D is the pressure sensing of matrix form distribution
The spacing of device.If the pressure data that signal acquisition module 21 gathers meets P1=P2=..., and=PN=P can be determined that use
The elasticity and wearing position that family is worn are correct, then wearing state mark WARE_FLAG are arranged to TRUE, be otherwise provided as
Preferably, after user's starter, signal acquisition module 21 while constantly acquisition 3-axis acceleration data AX, AY,
AZ.Whether device wears detection unit 201 can correct according to 3-axis acceleration data detection device wearing orientation simultaneously.Due to,
Under normal circumstances during person upright, correct wearing mode should only have an axis (i.e. the axis of trunk vertical direction) acceleration
Angle value is g (i.e. acceleration of gravity), and other two axle acceleration values are 0.Assuming that Y-axis represents trunk side when human body is stood
To, then device wearing detection unit 201 judges whether AX, AY, AZ of acquisition meet AY=g and AX=AZ=0, if it is satisfied,
It is in the right direction then to judge that user wears, wearing state is set to identify WARE_FLAG=TRUE, is otherwise provided as FALSE.
Preferably, signal acquisition module 21 further includes temperature sensor.Wherein, since temperature sensor has direction in itself
Property (towards human body side and towards outside), the temperature sensor of the present embodiment is arranged to two, and the direction of one is provided towards
The one side of human body, for gathering human body temperature, another direction is provided towards the one side of air, for gathering environment temperature
Degree after two temperature sensors set direction, is integrated in 4 enterprising trip temperature of integrated chip acquisition shown in FIG. 1.Device starts
Afterwards, the temperature data T1 and device of 21 continuous harvester of signal acquisition module close to human body one side are exposed to one side in air
Temperature data T2, if human body does not have wearable device, should meet substantially T1=T2 (allow have a certain range error, as T1 with
The differential of T2 is bordering on such as 0.5 ° of the threshold value of setting), if in the case of wearable device, the temperature difference, that is, T1 and T2 of both sides should have one
Fixed amplitude (0.5 ° of the threshold value such as larger than set).Device wears detection unit 201 according to the temperature data of acquisition, compare T1 and
The temperature difference of T2, you can whether decision maker is worn.If worn, set wearing state and be identified as TRUE, be otherwise provided as
Preferably, signal acquisition module 21 may also include human-body biological electric transducer.After device starts, signal acquisition module
Whether 21 human biological signals that constantly acquisition human-body biological electric transducer exports are high electricity according to human biological signal
It is flat, judge whether human body has worn device.If human biological signal is high level, wearing state is set to be identified as
TRUE is otherwise provided as FALSE.
In practical applications, device wear detection unit 201 can only according to above pressure data, 3-axis acceleration,
The one of which of temperature data and human biological signal carry out device whether the detection worn or whether correctly worn, can also
The combination of arbitrary two therein of simultaneous selection or more is detected, and the detection data of selection are more, and the accuracy rate of detection is higher.
Wherein, when the combination of wherein arbitrary two or more is selected to be detected, as long as the testing result of any of which mode is wearing
It is wrong, wearing state will be set to be identified as FALSE.Such as, the combination of four can be carried out at the same time to be detected, including first passing through
Whether temperature data detection device is worn, if passing through human body (as being all 37 degree) when human body and external environment temperature difference very little
Whether bioelectric signal measuring apparatus is worn, and reduced pressure data judge whether wearing position is correct if wearing, if just
Really judge whether wearing direction is correct further according to 3-axis acceleration.If four is all correct, it is determined as that device is worn correctly, if
It puts wearing state and is identified as TRUE, be otherwise provided as FALSE, and continue data acquisition.Fall detection unit 202, which is read, wears
Wear the value of status indicator, when for TRUE when, pass through signal acquisition module 21 and gather user behavior information data and carry out fall detection.
Preferably, when carrying out wearing detection, it can also after initialization or detect with mistaking, pass through voice
Correct usual method is played, user is instructed to be worn.
As shown in Fig. 2, the present apparatus can also include human-computer interaction module 25.Human-computer interaction module 25 can be touch-screen,
Sound identification module or button are arranged to receive user's input, carry out data input or start alarm modules according to user instruction
24 carry out rescue alarm or release rescue alarm.Such as a key is carried out by touch-screen typing user basic information or by button
Alarm, when also disclosure satisfy that influence testing result the problems such as because of sample rate deficiency and algorithm discrimination, user can be fallen in time
It alarms, it is very efficient and convenient.
Preferably, met the needs of users in order to more hommization, the present invention may further be provided to move back automatically
Go out the function of alarm of falling, rest with to meet user after tumble and voluntarily climbed for a period of time or the feelings such as other modes are stood up
, it is necessary to inform guardian in time and automatically exit from the demand of alarm mode under condition.
As shown in Fig. 2, signal processing module 20 further includes state detection unit 203 after tumble, it is arranged to occur in human body
After tumble, 21 continuous collecting 3-axis acceleration data (AX, AY, AZ) of signal acquisition module, pressure altitude data (H) and wearing portion
Position pressure data (P1, P2 ... ..., PN), and acceleration information, altitude information and the pressure of (in such as 4 seconds) were stored in a period of time
Force data, analysis represent the Y-axis average in trunk directionPressure altitude value H3
The condition stood is switched to by tumble so that whether pressure value (P13, P23 ... ..., PN3) meets.Specifically, given threshold TH6 and
TH7 judges whether to meet | AY_3-g |<TH6、H3-H2>TH7 and P13=P23=...=PN3 if all met, sentence
It is set to user voluntarily to stand, it is FALSE to set tumble status indicator, and sends corresponding secondary signal and (ask as released
Rescue signal) to alarm modules 24, alarm modules 24 are sent by wireless communication module 23 departs from the information of tumble state to long-range
Terminal, guardian user to be reminded to have been detached from tumble state.Wherein, TH6 be represent trunk direction acceleration it is equal
The threshold value of difference between acceleration value g (i.e. acceleration of gravity) under value AY_3 and standing state can be set as close to 0,
AY_3 closer to gravity acceleration g (i.e. | AY_3-g |<TH6 is smaller during TH6), show human body closer to normal standing shape
State；TH7 is difference in height threshold value, represents under current state the height of device and difference and standing shape between the height of device during tumble
The degree of closeness of height, could be provided as 80cm during state, can also be configured according to the height information of user, H3-H2>TH7
Device height H2 high when representing present apparatus height H3 than falling, illustrates human body away from ground；And P13=P23=...
=PN3 shows that the pressure of device wearing site has tended to balance, i.e., the larger point of no pressure-bearing illustrates user not
It is to land state in one side.As a result, by judging whether to meet decision condition | AY_3-g |<TH6、H3-H2>TH7 and P13=
P23=...=PN3, you can judge whether user has voluntarily stood, so as to be notified in time when user departs from tumble pattern
Guardian can effectively promote user experience, very convenient.Simultaneously as in warning stage, device needs not stop using wireless
Communication module, human-computer interaction module and alarm modules, the power consumption of device can be higher, and automatically detects user and release tumble shape
State backed off after random alarm can effectively reduce the power consumption of device.
Optionally, alarm modules 25 can also be loud speaker playing device, when starting help-asking mode, pass through channel radio
When interrogating module 23 to guardian's transmission distress signals, while start loud speaker and play voice help-asking signal, to be rescued in time
It helps；And when automatically exiting from help-asking mode, it is sent by wireless communication module 23 to guardian and departs from tumble status information and stop
Only play the voice help-asking signal of loud speaker.
Intelligent human-body tumble monitor device provided by the invention can be worn on user's waist, be used as waistband, very
It is convenient.Moreover, the device of the invention carries out human body tumble inspection by 3-axis acceleration data, pressure altitude data and pressure data
It surveys, more meets the behavioural characteristic of user, accuracy higher.Meanwhile inventive arrangement provides device wearing detection function, energy
It enough avoids not wearing because of device or reports by mistake bad when wearing incorrect, further improve the accuracy of fall detection, with
Promptly and accurately the tumble distress signals and location information of user are sent to guardian.The device of the invention can also carry simultaneously
For the automatic detection after tumble, it can continue to detect user behavior state after user falls, after user stands, will fall in time
The information released is sent to guardian, and it is convenient (as saved the time of guardian, reducing the essence of guardian to be brought to guardian
Refreshing tension pressure etc.).The device of the invention can also be handed over by the information of the realizations such as touch-screen, button, speech recognition and user
Mutually, it is user-friendly, can meet the needs of user cries for help by button when emergency situation occurs or reports by mistake.
Fig. 3 schematically shows the processing method (work of the intelligent human-body tumble monitor device of an embodiment of the present invention
Make method).As shown in figure 3, this method includes：
Step S301：Signal acquisition module gathers user behavior information data.
Signal acquisition module gathers user's human body attitude data (including 3-axis acceleration value by 3-axis acceleration sensor
AX, AY, AZ), user's pressure altitude data (H) are gathered by ambient pressure sensor, gathering user by pressure sensor wears
Wear the position pressure data of one week (P1, P2 ..., PN).User's attitude data can be used for the standing for judging user or flat
It lies state, pressure altitude data can be used for the height of judgment means from the ground, and the wearing site pressure data of one week can be used
Land the pressure condition of side and the side that do not land when judging that user lands.
Step S302：Signal processing module detects whether human body falls according to the user behavior information data of acquisition.
Signal processing module is analyzed according to the user behavior information data collected, judges whether human body falls
, when detecting that human body is fallen, step S303 is carried out, if not having to detect that human body is fallen, is continued
The data acquisition of step S301.Fig. 4 schematically shows the method flow of human body fall detection.As shown in figure 4, this method bag
Step S401：Acquisition human body attitude behavioral data, pressure altitude data and pressure data in real time.
Signal acquisition module carries out data acquisition in real time, main to gather human body attitude real time data (i.e. 3-axis acceleration number
According to AX, AY, AZ), ambient pressure height (H) data and wearing site pressure data (P1, P2 ..., PN).And using FIFO
Data in (First In First Out, first in first out) pattern storage a period of time, acceleration information, gas such as in 4 seconds
Press altitude information and pressure data.
Step S402：Judge whether " great change " of generation data.
Signal processing module accelerates data according to three axis, analyses whether the data of " great change " occur, because working as human body
During tumble, when striking ground, 3-axis acceleration signal, which is present with, fluctuates very violent one piece of data (before for details, reference can be made to
The narration of texts and pictures 9), in this given threshold TH1, according to 3-axis acceleration data AX, AY, the vectorial mould of AZ three, which calculates, obtains three
Axle acceleration amplitude ACC (calculation formula is referring to narration above), judges whether ACC>TH1 if meeting condition, illustrates
" great change " occurs at this time, then carries out step S403, if being unsatisfactory for condition, continues step S401.
Step S403：Before record variation height value (H1), Y-axis acceleration average (AY_1) and pressure value (P11,
P21 ..., PN1).
According to the data that each sensor of signal acquisition module gathers, the representative trunk before generation " great change " is recorded
The Y-axis mean data AY_1 (calculation formula is referring to narration above) in direction, pressure altitude data H1, pressure data (P11,
P21 ... ..., PN1).
Step S404：Calculate the acceleration change value ACC_CHG of " great change " moment.
According to the data that each sensor of signal acquisition module gathers, record occur " great change " before and after moment the period (as before
0.1s after 0.1s-) in acceleration information variable quantity ACC_CHG (calculation formula referring to above describe).
Step S405：Judge whether acceleration change value ACC_CHG is more than given threshold TH2.
Because the moment of contacting to earth normally fallen is very of short duration, while total variation when contacting to earth can be very big, given threshold
TH2 judges whether to meet condition ACC_CHG>TH2 illustrates that human body has contacted to earth tumble at this time, after tumble if meeting
Human normal can enter static section, then carry out step S406, otherwise carry out step S401 and persistently carry out data acquisition.
Step S406：Record variation after height value (H2), Y-axis acceleration average (AY_2), pressure value (P12,
P22 ..., PN2) and " calmness " segment acceleration change summation ACC_SUM.
According to the narration above to the line chart of Fig. 9, analysis is fallen occur after 3-axis acceleration in a period of time, detection
Going out the static section after falling, (Y-axis that trunk direction is represented in the section is essentially close to 0, can pass through Y-axis in the section
Acceleration whether level off to and 0 judged.), calculate the sum of 3-axis acceleration variation ACC_SUM (calculation formula in the section
Referring to narration above), while record the human body pressure altitude value H2 of the period and the pressure value at wearable device position
(P12, P22 ... ..., PN2).
Step S407：Whether satisfaction tumble decision condition.
Given threshold TH3, TH4, TH5 (value refer to above describe) judge the pressure altitude value H1 before great change and static
Whether the difference in height of the pressure altitude value H2 in section meets the threshold value TH3 more than setting, i.e., whether meets H2-H1>TH3, it is quiet
Only whether the sum of acceleration change in section ACC_SUM (calculation formula is referring to narration above) is less than the threshold value TH4 of setting, i.e.,
Whether ACC_SUM is met<TH4, while the judgment means wearing site pressure condition of one week, if there are the pressure sensings of one side
The pressure value of device is more than the threshold value TH5 of setting, i.e. ∑ | Pi1-Pi2 |>TH5.And judge the Y-axis acceleration average AY_1 before great change
Whether close to gravity acceleration g, whether the Y-axis acceleration average in static section is close to 0, if meeting H2-H1 simultaneously>TH3、
ACC_SUM<TH4 and ∑ | Pi1-Pi2 |>TH5, and AY_1 then carries out step S408, is otherwise walked close to g, AY_2 close to 0
Step S408：Judgement is fallen, and it is TRUE to set tumble status indicator, into tumble emergency state.
If meeting conditions above simultaneously, judge that human body has occurred that tumble (specifically seing above narration), then set
Tumble status indicator FALL_DOWN_FLAG=TRUE, while the first signal such as distress signal is sent to alarm modules, to enter
Tumble emergency state.
Pass through above step, you can the direction change for judging the standing of human body by 3-axis acceleration and lying low passes through gas
The height change of altitude information judgment means from the ground is pressed, human body wearing site (present invention is waist) is judged by pressure data
The compression situation variation of one week, so as to detect whether human body falls, meets human body behavioural characteristic, the accuracy rate of detection is more
It is high.
Step S303：Signal processing module sends distress signal to alarm modules, and obtains user position by locating module
Signal processing module sends distress signal (such as character " 1 ") to alarm modules, while is obtained and used by locating module
The geographical location information at family.
Step S304：Alarm modules carry out emergency response processing.
It, will be in the geographical location information and emergency of user by wireless communication module after alarm modules receive distress signal
Hold the remote terminal equipment for being sent to monitoring, guardian is notified, timely to be succoured.
Fig. 5 schematically shows the processing method of the intelligent human-body tumble monitor device of another embodiment of the present invention.
As shown in figure 5, the difference of the embodiment and Fig. 3 illustrated embodiments is, current embodiment require that first detection device whether
It wears correctly, if wearing the detection whether correct just progress human body falls.It is specific as follows：
Step S501：Signal acquisition module harvester wearing information data.
Signal acquisition module can be 3-axis acceleration sensor, pressure sensor, temperature sensor and human biological electricity
One of sensor or the combination of two or more can gather 3-axis acceleration data by 3-axis acceleration sensor
AX, AY, AZ gather wearing site pressure data P1, the P2 of one week by pressure sensor ..., PN passes through temperature sensing
Device gathers the temperature data T1 close to the human body one side and temperature data T2 exposed to air one side, is sensed by human biological electricity
Device gathers human biological signal.Signal acquisition module harvester wearing information data, can be in more than sensing data
One or a variety of combinations, the scheme of the present embodiment preferred three combination is described in detail.The combination
The accuracy of detection can be improved.
Step S502：Whether signal processing module is correctly worn according to the device wearing information data detection device of acquisition.
Whether signal processing module is correctly worn according to the data of acquisition, detection device.Fig. 6 schematically shows device
The detection method whether correctly worn, as shown in fig. 6, this method includes：
Step S601：It opens equipment and initializes.
The power supply of user's opening device, waiting device carry out the initialization of data automatically, and the state variable of device is assigned
Wearing state mark WARE_FLAG is such as initialized as FALSE, by tumble status indicator FALL_DOWN_FLAG assignment by initial value
For FALSE etc..
Step S602：Real-time collecting temperature, pressure and acceleration information, and carry out voice guidance wearing to user.
Signal acquisition module gathers 3-axis acceleration data AX, AY, AZ in real time, the wearing site pressure data P1 of one week,
P2 ..., PN, the temperature data T1 close to human body one side and the temperature data T2 exposed to air one side, while pass through language
Sound plays usual method and carries out wearing guidance to user.
Step S603：Judge human body one side temperature T1 compared with the temperature T2 of air one side, if T1-T2>TH.
If human body does not have wearable device, theoretically T1=T2, there is 0.5 ° or so of error in practice, if human body is worn
It wears, then one side is that air themperature, one side are that human body temperature may result in both sides and temperature difference occurs, has thus reached detection device
The purpose whether worn.Based on there are the fact that the temperature difference, whether given threshold TH, comparison T1 and T2 meets temperature difference and is more than setting
Threshold value, if greater than step S604 is then carried out, otherwise persistently carry out the data acquisition of step S602.
It, as the presently preferred embodiments, can be with it should be noted that due to there is a situation where that ambient temperature is close with human body temperature
Increase human-body biological electric transducer in signal acquisition module, further detected, specially gather human biological signal,
Determine whether high level, if the output of human-body biological electric transducer is high level, illustrate that human body has worn device, it can be into
The pressure detecting of row step S604, otherwise persistently carries out data acquisition.In practical applications, human-body biological electric transducer can also
As the alternative solution whether temperature sensor judgment means are worn, i.e., temperature sensor is replaced with into human biological electricity sensing
Device, carries out the judgement of human biological signal, and the present invention is not limited combination.
Step S604：Judge whether the value of each pressure sensor meets P1=P2=......=PN>0.
If human body is correctly worn, the human body waist pressure value of one week meets pressure equal and equal to elasticity when moderate
Force value P, that is, have P1=P2=......=PN=P>0, judge whether to meet the condition can judge device whether wearing portion
Position is correct and elasticity is suitable (for details, reference can be made to and describe above).If it is satisfied, then carrying out step S605, otherwise persistently walked
The data acquisition of rapid S602.
Step S605：Judge whether 3-axis acceleration value meets AY=g and AX=AZ=0 under normal stand situation.
Under normal circumstances, human body stand when, correct wearing mode should only have an axle acceleration value for g (gravity and
Speed) other two axis should be 0, if Y-axis represents trunk direction, i.e. AY=g and AX=AZ=0 when human body is stood, if full
Wearing that is sufficient then illustrating device is in the right direction, then carries out step S607, otherwise carries out step S606.
Step S606：Anisotropy is worn by voice prompt user.
Voice prompt is played, user is reminded to wear anisotropy, and continues the data acquisition of step S602.
Step S607：Judge to wear correctly, wearing correct status is set to be identified as TRUE, into fall detection state.
If meeting the decision condition of temperature, pressure and 3-axis acceleration simultaneously, illustrate that device has been worn, and wear
Position and direction are all correct, will wear correct status mark WARE_FLAG at this time and are arranged to TRUE, carry out step S503's afterwards
The data acquisition of fall detection and judge whether human body falls, otherwise without fall detection.Thus, it is possible to it avoids because of dress
Wrong report when not wearing or wearing incorrect is put, improves fall detection and the accuracy rate for alarm of crying for help.
Step S503：Signal acquisition module gathers user behavior information data.
Step S504：Speech play usual method.
Step S505：Signal processing module detects whether human body falls according to the user behavior information data of acquisition.
Step S506：Signal processing module sends distress signal to alarm modules, and obtains user position by locating module
Step S507：Alarm modules carry out emergency response processing.
The realization of step S503 to step S507 can refer to rapid S301 to step S304 above.Pass through the embodiment, you can
It realizes under the premise of judging to wear correctly, then carries out fall detection, the efficiency and accuracy rate of detection can be improved.
Fig. 7 schematically shows the processing method of the intelligent human-body tumble monitor device of another embodiment of the present invention.
As shown in fig. 7, the difference of the embodiment and Fig. 5 illustrated embodiments is, the present embodiment is detecting that human body falls
And after entering help-asking mode, acquisition user behavior data is may proceed to, whether human body releases the detection of tumble state after being fallen,
And when detecting that human body releases tumble state, tumble help-asking mode is automatically exited from, it is provided conveniently for user and guardian.Specifically
Step S701：Signal acquisition module harvester wearing information data.
Step S702：Whether signal processing module is correctly worn according to the device wearing information data detection device of acquisition.
Step S703：Signal acquisition module gathers user behavior information data.
Step S704：Speech play usual method.
Step S705：Signal processing module detects whether human body falls according to the user behavior information data of acquisition.
Step S706：Signal processing module sends distress signal to alarm modules, and obtains user position by locating module
Step S707：Alarm modules carry out emergency response processing.
Step S708：Signal acquisition module continuous collecting user behavior information data.
Step S709：Signal processing module detects whether human body releases tumble shape according to the user behavior information data of acquisition
Wherein, step S701 to step S707 is the same as step S501 to step S507.Difference is, when carrying out tumble alarm
Afterwards, it is necessary to continue the data acquisition of step S708, the data of acquisition include 3-axis acceleration data, pressure altitude data and
The pressure data of wearing site, and need to carry out the data of acquisition the releasing fall detection of step S709, so as in user
After rest after the automatic or standing that depends on others' help, guardian user can be notified to have released tumble state in time, so as to
The nervous anxiety of guardian is reduced, saves the time of guardian, to provide more intelligent better user service.
Fig. 8 schematically shows that signal processing module detects whether human body solves according to the user behavior information data of acquisition
Except the method for tumble state.As shown in figure 8, this method includes：
Step S801：Judge whether tumble status indicator is TRUE.
The value of tumble status indicator FALL_DOWN_FLAG is read, TRUE is determined whether, if TRUE then illustrates
It through falling, then carries out step S802 and persistently carries out data acquisition, otherwise illustrate that user there is no falling, carries out step
Step S802：Acquisition human body attitude behavioral data, pressure altitude data and pressure data in real time.
Signal acquisition module continuous collecting 3-axis acceleration data (AX, AY, AZ), pressure-altitude sensor continuous collecting are high
Degrees of data (H), pressure sensor gather pressure data (P1, P2 ... ..., PN) in real time, and when storing one section using form of FIFO
Between such as the data gathered in 4 seconds.
Step S803：Exit detection.
Step S804：Record fall after height value (H3), Y-axis acceleration average (AY_3) and pressure value (P13,
P23 ..., PN3).
For signal processing module according to the 3-axis acceleration data of storage, the Y-axis acceleration being calculated in the period is equal
Value AY_3 (calculation formula referring to above describe), and record the pressure altitude data H3 gathered in real time and pressure data (P13,
P23 ..., PN3).
Step S805：Judge whether to meet the decision condition for releasing tumble state.
Given threshold TH6 and TH7, judge whether to meet condition | AY_3-g |<TH6、H3-H2>TH7, P13=P23=...
=PN3, if met | AY_3-g |<TH6 illustrates that trunk perpendicular to ground, that is, is stood.Meet H3-H2>TH7 is then said
It is high when bright current device height is than falling.Meeting P13=P23=...=PN3 then can determine whether equipment current pressure state with falling
Preceding pressure state comparison has become uniform, is no longer the big situation that lands of one side stress.Wherein, the value tool of threshold value TH6, TH7
Body referring to describing above.If meeting decision condition, step S806 is carried out, step S802 is otherwise carried out and continues gathered data.
Step S806：Judge that it is FALSE to set tumble status indicator from tumble mode release, output is released to fall and asked
If meeting decision condition, it can determine that user stands from tumble state, tumble status indicator be set at this time
It for FALSE, and is sent to alarm modules and releases tumble signal (such as character " 0 "), to stop the alarm of emergency.
Step S710：Signal processing module sends to alarm modules and releases tumble signal, and alarm modules carry out releasing emergency
Alarm modules send to have released and fall by wireless communication module according to the releasing tumble signal received to guardian
, the information of emergency is stopped, meanwhile, alarm modules can also stop carrying out voice help-asking by mute speaker.
Preferably, in order to avoid when detecting mistake, reporting by mistake, the personal safety of user, this hair are preferably ensured
Operating method in bright monitor device can also include：The signal that user is received by button inputs, broadcasting/pause loud speaker
Alarm and transmission emergency/releasing distress signals to remote terminal.One button is set such as on device, if user is of short duration to press one
Under, then signal processing module receives the input of user, to alarm modules send distress signal, alarm modules play loud speaker and
The distress signals comprising location information are sent to remote terminal to guardian.If user's long-press button, signal processing module
Receive user input, to alarm modules send release distress signal, alarm modules then stop play loud speaker voice help-asking and
The information cleared is sent to guardian.
With the method for the invention it is achieved that by 3-axis acceleration sensor, baroceptor and pressure sensor into
The acquisition of pedestrian's body behavioural information data guards the detection of human body tumble state with but by human body behavioural information data, accurately
Rate higher meets the monitoring demand of the tumble to old man and patient.Meanwhile method of the invention additionally provide device wear correctly with
Detection that is no and automatically exiting from tumble pattern, can avoid the wrong report brought due to device wears problem, also can judge to fall
Continue the situation of detection user afterwards, reach timely notifier processes during standing, more intelligence is convenient, and the accuracy rate of detection is more
It is high.
Above-described is only some embodiments of the present invention.For those of ordinary skill in the art, not
On the premise of departing from the invention design, various modifications and improvements can be made, these belong to the protection model of the present invention