CN110245744A - It is a kind of that detection method is fallen down based on multilayer perceptron - Google Patents
It is a kind of that detection method is fallen down based on multilayer perceptron Download PDFInfo
- Publication number
- CN110245744A CN110245744A CN201910455921.XA CN201910455921A CN110245744A CN 110245744 A CN110245744 A CN 110245744A CN 201910455921 A CN201910455921 A CN 201910455921A CN 110245744 A CN110245744 A CN 110245744A
- Authority
- CN
- China
- Prior art keywords
- multilayer perceptron
- layer
- axis
- detection method
- acceleration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 30
- 230000001133 acceleration Effects 0.000 claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000007689 inspection Methods 0.000 claims abstract description 4
- 210000002569 neuron Anatomy 0.000 claims description 28
- 238000012549 training Methods 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 230000000644 propagated effect Effects 0.000 claims description 3
- 210000002364 input neuron Anatomy 0.000 claims description 2
- 230000008447 perception Effects 0.000 claims 1
- 230000006399 behavior Effects 0.000 abstract description 17
- 238000010801 machine learning Methods 0.000 abstract description 3
- 230000032683 aging Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 208000027418 Wounds and injury Diseases 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 208000014674 injury Diseases 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 206010061623 Adverse drug reaction Diseases 0.000 description 1
- 208000034656 Contusions Diseases 0.000 description 1
- 208000030453 Drug-Related Side Effects and Adverse reaction Diseases 0.000 description 1
- 206010024453 Ligament sprain Diseases 0.000 description 1
- 208000010040 Sprains and Strains Diseases 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000009519 contusion Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 210000003127 knee Anatomy 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1112—Global tracking of patients, e.g. by using GPS
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1116—Determining posture transitions
- A61B5/1117—Fall detection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient ; user input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
It is a kind of that detection method is fallen down based on multilayer perceptron, method includes the following steps: the 3-axis acceleration and angular velocity data of (1) acquisition human body carry out, and data are pre-processed, and extract feature;(2) multilayer perceptron model is constructed, and multilayer perceptron model is trained using the feature extracted in step (1), can correctly classify and human body daily behavior and fall down behavior;(3) using the multilayer perceptron model inspection falling over of human body established in step (2), warning message is uploaded to server once detecting and falling down.The present invention improves the accuracy rate for falling down detection by being introduced into multilayer perceptron and its BP algorithm in machine learning in falling down detection.
Description
Technical field
It is a kind of detection method to be fallen down based on multilayer perceptron the present invention relates to detection technique is fallen down.
Background technique
With the economic rapid development with science and technology, people's lives level and medical and health conditions are increasingly improved, per capita the longevity
Life and is cannot be mentioned in the same breath decades ago there has also been earth-shaking variation.At the same time, the problem of an aging population is also increasingly aobvious
It is existing.As a generation of baby boom gradually steps into old age, this aging trend will be also amplified therewith.The year two thousand forty is expected,
It will be more than 65 years old by the people for having 23%.Aging of population allows burden on society to aggravate, and social security system is faced adverse conditions.
The elderly due to the factors such as physical function decline, various diseases, drug side-effect influence so that occurring unexpected
Probability substantial increase, and falling down is wherein most commonly seen accident, once fall down it is light if sprain contusion, it is heavy then life can be threatened
Life safety.And most of old man can be in uncared-for state when falling down, and this results in it to be unable to get timely rescuing
It controls, to miss optimal therapic opportunity, causes more major injury even threat to life.
Therefore, if can design it is a kind of fall down detection method for the elderly, perceived in time simultaneously after the elderly falls down
Alarm, so that it may be injury caused by reducing accidental falls.However detection method of falling down at this stage is all easy wrong report, fails to report,
The accuracy rate of method is not high enough, cannot be applied well.
In conclusion how to obtain a kind of accuracy rate it is high fall down detection method, still lack effective scheme.
Summary of the invention
To solve the above-mentioned problems, it reduces the wrong report fallen down in detection, fail to report problem, the present invention provides one kind based on more
Layer perceptron falls down detection method, by being introduced into multilayer perceptron and its BP algorithm in machine learning in falling down detection,
Improve the accuracy rate for falling down detection.
The technical solution adopted by the present invention to solve the technical problems is:
It is a kind of that detection method is fallen down based on multilayer perceptron, method includes the following steps:
(1) 3-axis acceleration and angular velocity data for acquiring human body carry out, and pre-process to data, and extract spy
Sign;
(2) multilayer perceptron model is constructed, and multilayer perceptron model is instructed using the feature extracted in step (1)
Practice, can correctly classify and human body daily behavior and fall down behavior;
(3) using the multilayer perceptron model inspection falling over of human body established in step (2), report once detecting and falling down
Alert information is uploaded to server.
Further, in the step (1), the equipment use for acquiring human body acceleration and angular speed is that motion sensor is
MPU6050, the CPU for acquiring equipment is MT6260MA, the realization for data processing and communication function.
Further, in the step (1), the equipment for acquiring human body acceleration and angular speed is worn on waist, and described three
Left and right directions is X-axis when axle acceleration is person upright, and front-rear direction is Y-axis, and up and down direction is Z axis, three axis angular rate
The angular speed that respectively rotates around X-axis, the angular speed rotated around Y-axis and the angular speed rotated about the z axis.
Further, in the step (1), to the preprocessing process of data are as follows: the summation for calculating 3-axis acceleration accelerates
Spend vector;Summation vector acceleration is continuously integrated, speed signal is obtained;Speed signal will be obtained continuously to be accumulated
Get position signal.
In the step (1), characteristic procedure is extracted are as follows: count respectively to acceleration, angular speed, speed, position signal is obtained
Their maximum value, minimum value, average value, variance, standard deviation and range are calculated, 24 features are amounted to.
In the step (2), construct multilayer perceptron model the step of are as follows:
(2.1) using obtained feature a part as training sample, another part is as test sample;
(2.2) training sample is used, the input layer of multilayer perceptron, hidden layer, output layer are constructed;
(2.3) test sample is used, the multilayer perceptron model of building is tested.
In the step (2.2), the neuron number of output layer is the Characteristic Number 24 extracted, the neuron number of hidden layer
It needs to select suitable number as the case may be, the neuron number of output layer is the class number 2 to be divided into, and respectively falls down row
For with daily behavior.
In the step (2.2), training multilayer perceptron model the step of are as follows:
(2.2.1) constructs 24 neurons of an input layer, 15 neurons of hidden layer, the multilayer of two neurons of output layer
Perceptron;
(2.2.2) using the neuron of activation primitive activation hidden layer and output layer, the activation primitive is Sigmoid letter
Number;
(2.2.3) generates the weight coefficient and amount of bias of each layer neuron at random;
(2.2.4) updates the weight coefficient and amount of bias of each layer using training sample by BP algorithm.
In the step (2.2.3), weight coefficient and amount of bias are only two parameters for needing to adjust in multilayer perceptron.
In the step (2.3), the mode of test multilayer perceptron model is to use test sample as input nerve
Member, the weight coefficient generated using training and amount of bias execute propagated forward algorithm, the data exported according to output layer neuron
Determine that it is the behavior of falling down or daily behavior.
It further include GPS positioning data in addition to falling down information into the warning message of server transmission in the step (3),
Position, fast and easy treatment are fallen down for positioning old man.GPS data is acquired, using the GPS positioning core of MT3336
Piece.
The invention has the benefit that being calculated by the multilayer perceptron and its BP being introduced into falling down detection in machine learning
Method improves the accuracy rate for falling down detection.
Detailed description of the invention
Fig. 1 is to fall down detection method flow chart based on multilayer perceptron.
Fig. 2 is data acquisition and sending device structural block diagram.
Fig. 3 is building multilayer perceptron model flow figure.
Fig. 4 is multilayer perceptron model structure schematic diagram.
Specific embodiment
The present invention is described further with reference to the accompanying drawing.
Referring to Fig.1~Fig. 4, it is a kind of that detection method is fallen down based on multilayer perceptron.By introducing machine in falling down detection
Multilayer perceptron and its BP algorithm in device study, improve the accuracy rate for falling down detection.
As shown in Figure 1, a kind of fall down detection method based on multilayer perceptron, method includes the following steps:
(1) 3-axis acceleration and angular velocity data for acquiring human body carry out, and pre-process to data, and extract spy
Sign;
(2) multilayer perceptron model is constructed, and multilayer perceptron model is instructed using the feature extracted in step (1)
Practice, can correctly classify and human body daily behavior and fall down behavior;
(3) using the multilayer perceptron model inspection falling over of human body established in step (2), report once detecting and falling down
Alert information is uploaded to server.
In the present embodiment, in the step (1), the equipment use for acquiring human body acceleration and angular speed is motion-sensing
Device is MPU6050, and the CPU for acquiring equipment is that MT6260MA is illustrated in figure 2 for the realization of data processing and communication function
The structural block diagram of the equipment.
In the present embodiment, in the step (1), the equipment for acquiring human body acceleration and angular speed is worn on waist, institute
Stating left and right directions when 3-axis acceleration is person upright is X-axis, and front-rear direction is Y-axis, and up and down direction is Z axis, three shaft angle
Speed is respectively angular speed rotate around X-axis, the angular speed rotated around Y-axis and the angular speed rotated about the z axis.
In the present embodiment, in the step (1), to the preprocessing process of data are as follows: calculate the summation of 3-axis acceleration
Vector acceleration, formula are as follows:
In above formula, x, y, z is respectively the angular speed of X-axis, Y-axis, Z-direction.
Summation vector acceleration is continuously integrated, speed signal is obtained;Speed signal will be obtained to carry out continuously
Integral obtains position signal.
In the present embodiment, in the step (1), characteristic procedure is extracted are as follows: to obtaining acceleration, angular speed, speed, position
Confidence number calculates separately their maximum value, minimum value, average value, variance, standard deviation and range, amounts to 24 features.
As shown in figure 3,
In the present embodiment, in the step (2), construct multilayer perceptron model the step of are as follows:
(2.1) using obtained feature a part as training sample, another part is as test sample;In the present embodiment
In, daily behavior have standing, walking, running, jump, sit on chair, sit on the ground, lie on the ground, lie it is first-class to bed;It falls down
Behavior includes directly falling down forward, and bending knee is fallen down forward, falls down by wall, directly falls down backward, to the left, to the right backward, to recoil
Ground is fallen down, and is directly fallen down to the left, to the right;
(2.2) training sample is used, the input layer of multilayer perceptron, hidden layer, output layer are constructed;
(2.3) test sample is used, the multilayer perceptron model of building is tested.
In the present embodiment, in the step (2.2), the neuron number of output layer is the Characteristic Number 24 extracted, hidden
The neuron number of layer needs to select suitable number (being selected as 15 in the present embodiment), the nerve of output layer as the case may be
First number is the class number 2 to be divided into, and respectively falls down behavior and daily behavior.
In the present embodiment, in the step (2.2), training multilayer perceptron model the step of are as follows:
(2.2.1) constructs 24 neurons of an input layer, 15 neurons of hidden layer, the multilayer of two neurons of output layer
Perceptron;
(2.2.2) using the neuron of activation primitive activation hidden layer and output layer, the activation primitive is Sigmoid letter
Number;Sigmoid function formula is as follows:
(2.2.3) generates the weight coefficient and amount of bias of each layer neuron at random;
(2.2.4) updates the weight coefficient and amount of bias of each layer using training sample by BP algorithm.
In the formula illustrated later, lowercase alphabet indicating amount, overstriking lowercase indicates that vector, capitalization indicate
Matrix.
As shown in figure 4, the information of input is [x1, x2, x3].For layer l, L is usedlIndicate all neurons of this layer, it is defeated
It is out yl, wherein the output of j-th of node isThe input of the node isConnect the l layers of power with (l-1) layer
Weight matrix is Wl, the weight of i-th of node to l j-th of node of layer of upper one layer ((l-1) layer) is It is
The biasing of l j-th of node of layer.
Each layer of output is represented by following formula in multilayer perceptron:
Wherein f () indicates activation primitive, is in the present embodiment Sigmoid function.
In the present embodiment, the weight coefficient in the step (2.2.3) and amount of bias are only two need in multilayer perceptron
The parameter to be adjusted.
In the present embodiment, in the step (2.2.4), input sample is x=[x1, x2..., xn], label t;It is right
In layer l, L is usedlIndicate all neurons of this layer, output is yl, wherein the output of j-th of node isThe node
Input isThe weight matrix for connecting l layers and (l-1) layer is Wl, i-th of node of one layer upper ((l-1) layer) to
The weight of l j-th of node of layer is For the biasing of l j-th of node of layer.The last layer (output layer) of network
For kth layer.
The parameter of BP algorithm more new formula are as follows:
Wherein E is loss function:
Wherein δlIt is error to the change rate of input:
Wherein, since the present embodiment uses Sigmoid function, so
f′(ul)=yl(1-yl) (8)
In the present embodiment, the mode of test multilayer perceptron model is to be made using test sample in the step (2.3)
To input neuron, the weight coefficient and amount of bias generated using training executes propagated forward algorithm (using formula (3)), according to
The data of output layer neuron output determine that it is the behavior of falling down or daily behavior.
In the present embodiment, it in the step (3), into the warning message of server transmission, is also wrapped in addition to falling down information
GPS positioning data are included, fall down position, fast and easy treatment for positioning old man.As shown in Fig. 2, acquisition GPS data, is adopted
It is the GPS positioning chip of MT3336.
The present embodiment falls down detection method based on multilayer perceptron, and present invention use is by acceleration and angular velocity data
Input layer of derivative 24 characteristic values as multilayer sensor constructs multilayer sensor by multilayer sensor and its BP algorithm
Model can effectively classify to human body daily behavior with the behavior of falling down.It is acquired by MPU6050, wearing is worked as in MT6260 processing
It can be monitored in real time whether human body is fallen down when the equipment, information and GPS positioning information hair will be fallen down once falling down
Server is given, household is notified by server, gives treatment to the old man fallen down in time.
Claims (7)
1. a kind of fall down detection method based on multilayer perceptron, which is characterized in that method includes the following steps:
(1) 3-axis acceleration and angular velocity data for acquiring human body carry out, and pre-process to data, and extract feature;
(2) multilayer perceptron model is constructed, and multilayer perceptron model is trained using the feature extracted in step (1),
It can correctly classify and human body daily behavior and fall down behavior;
(3) using the multilayer perceptron model inspection falling over of human body established in step (2), alarm signal once detecting and falling down
Breath is uploaded to server.
2. a kind of as described in claim 1 fall down detection method based on multilayer perceptron, it is characterised in that: the step
(1) in, the equipment use of acquisition human body acceleration and angular speed is that motion sensor is MPU6050, and the CPU for acquiring equipment is
MT6260MA, the realization for data processing and communication function;Acquisition human body acceleration and angular speed sets in the step (1)
Standby to be worn on waist, left and right directions is X-axis when the 3-axis acceleration is person upright, and front-rear direction is Y-axis, and up and down direction is
Z axis, three axis angular rate are respectively that angular speed rotate around X-axis, angular speed rotate around Y-axis and the angle rotated about the z axis are fast
Degree.
3. a kind of as claimed in claim 1 or 2 fall down detection method based on multilayer perceptron, it is characterised in that: the step
Suddenly in (1), to the pretreated process of data are as follows: calculate the summation vector acceleration of 3-axis acceleration;By summation acceleration to
Amount is continuously integrated, and speed signal is obtained;It will obtain the continuous integral of speed signal progress and obtain position signal;The step
Suddenly characteristic procedure is extracted in (1) are as follows: to obtain acceleration, angular speed, speed, position signal and calculate separately they maximum value,
Minimum value, average value, variance, standard deviation and range amount to 24 features.
4. a kind of as described in one of claims 1 to 3 falls down detection method based on multilayer perceptron, it is characterised in that: institute
The step of stating in step (2), constructing multilayer perceptron model are as follows:
(2.1) using obtained feature a part as training sample, another part is as test sample;
(2.2) training sample is used, the input layer of multilayer perceptron, hidden layer, output layer are constructed;
(2.3) test sample is used, the multilayer perceptron model of building is tested;
In the step (2.2), the neuron number of output layer is the Characteristic Number 24 extracted, and the neuron number of hidden layer needs
Select suitable number as the case may be, the neuron number of output layer is the class number 2 to be divided into, respectively fall down behavior with
Daily behavior.
5. a kind of as claimed in claim 4 fall down detection method based on multilayer perceptron, it is characterised in that: the step
(2.2) in the step of training multilayer perceptron model are as follows:
(2.2.1) constructs 24 neurons of an input layer, 15 neurons of hidden layer, the Multilayer Perception of two neurons of output layer
Device;
(2.2.2) using the neuron of activation primitive activation hidden layer and output layer, the activation primitive is Sigmoid function;
(2.2.3) generates the weight coefficient and amount of bias of each layer neuron at random;
(2.2.4) updates the weight coefficient and amount of bias of each layer using training sample by BP algorithm;
Weight coefficient and amount of bias in the step (2.2.3) are only two parameters for needing to adjust in multilayer perceptron.
6. a kind of as claimed in claim 4 fall down detection method based on multilayer perceptron, it is characterised in that: the step
(2.3) in, the mode of test multilayer perceptron model is to use test sample as input neuron, is generated using training
Weight coefficient and amount of bias execute propagated forward algorithm, determine that it is the behavior of falling down also according to the data that output layer neuron exports
It is daily behavior.
7. a kind of as described in one of claims 1 to 3 falls down detection method based on multilayer perceptron, it is characterised in that: institute
It states in step (3), further includes GPS positioning data in addition to falling down information in the warning message sent to server, it is old for positioning
People falls down position, acquires GPS data using the GPS positioning chip of MT3336.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910455921.XA CN110245744A (en) | 2019-05-29 | 2019-05-29 | It is a kind of that detection method is fallen down based on multilayer perceptron |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910455921.XA CN110245744A (en) | 2019-05-29 | 2019-05-29 | It is a kind of that detection method is fallen down based on multilayer perceptron |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110245744A true CN110245744A (en) | 2019-09-17 |
Family
ID=67885407
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910455921.XA Pending CN110245744A (en) | 2019-05-29 | 2019-05-29 | It is a kind of that detection method is fallen down based on multilayer perceptron |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110245744A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110889330A (en) * | 2019-10-30 | 2020-03-17 | 西北工业大学 | BP neural network-based old people tumbling detection method and system |
CN113255527A (en) * | 2021-05-28 | 2021-08-13 | 汉谷云智(武汉)科技有限公司 | Method and equipment for monitoring operation normative during concrete unloading process |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103976739A (en) * | 2014-05-04 | 2014-08-13 | 宁波麦思电子科技有限公司 | Wearing type dynamic real-time fall detection method and device |
CN106539587A (en) * | 2016-12-08 | 2017-03-29 | 浙江大学 | A kind of fall risk assessment and monitoring system and appraisal procedure based on sensor of doing more physical exercises |
CN107019501A (en) * | 2017-05-05 | 2017-08-08 | 山东师范大学 | Detection method and system are fallen down based on genetic algorithm and the long-range of probabilistic neural network |
CN108338791A (en) * | 2018-02-09 | 2018-07-31 | 张立海 | The detection device and detection method of unstable motion data |
CN108549900A (en) * | 2018-03-07 | 2018-09-18 | 浙江大学 | Tumble detection method for human body based on mobile device wearing position |
-
2019
- 2019-05-29 CN CN201910455921.XA patent/CN110245744A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103976739A (en) * | 2014-05-04 | 2014-08-13 | 宁波麦思电子科技有限公司 | Wearing type dynamic real-time fall detection method and device |
CN106539587A (en) * | 2016-12-08 | 2017-03-29 | 浙江大学 | A kind of fall risk assessment and monitoring system and appraisal procedure based on sensor of doing more physical exercises |
CN107019501A (en) * | 2017-05-05 | 2017-08-08 | 山东师范大学 | Detection method and system are fallen down based on genetic algorithm and the long-range of probabilistic neural network |
CN108338791A (en) * | 2018-02-09 | 2018-07-31 | 张立海 | The detection device and detection method of unstable motion data |
CN108549900A (en) * | 2018-03-07 | 2018-09-18 | 浙江大学 | Tumble detection method for human body based on mobile device wearing position |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110889330A (en) * | 2019-10-30 | 2020-03-17 | 西北工业大学 | BP neural network-based old people tumbling detection method and system |
CN113255527A (en) * | 2021-05-28 | 2021-08-13 | 汉谷云智(武汉)科技有限公司 | Method and equipment for monitoring operation normative during concrete unloading process |
CN113255527B (en) * | 2021-05-28 | 2021-10-08 | 汉谷云智(武汉)科技有限公司 | Method and equipment for monitoring operation normative during concrete unloading process |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10234936B2 (en) | Smart wearable devices and methods with attention level and workload sensing | |
Jefiza et al. | Fall detection based on accelerometer and gyroscope using back propagation | |
JP2023002595A (en) | Systems and methods of tracking patient movement | |
CN105125221B (en) | Detecting system and method are fallen down in cloud service in real time | |
Zhang et al. | Activity monitoring using a smart phone's accelerometer with hierarchical classification | |
Pannurat et al. | A hybrid temporal reasoning framework for fall monitoring | |
US8560267B2 (en) | Identifying one or more activities of an animate or inanimate object | |
CN104361321B (en) | A kind of method for judging the elderly and falling down behavior and balance ability | |
Boissy et al. | User-based motion sensing and fuzzy logic for automated fall detection in older adults | |
CN109009145A (en) | A kind of tumble judgment method based on wearable device | |
CN110245744A (en) | It is a kind of that detection method is fallen down based on multilayer perceptron | |
Zhao et al. | Recognition of human fall events based on single tri-axial gyroscope | |
Zurbuchen et al. | A comparison of machine learning algorithms for fall detection using wearable sensors | |
CN109310351A (en) | For characterizing the assessment system and method for the heart rate of object | |
Gao et al. | A comparison of classifiers for activity recognition using multiple accelerometer-based sensors | |
CN110464315A (en) | It is a kind of merge multisensor the elderly fall down prediction technique and device | |
CN110415825A (en) | A kind of old man's safe condition intelligent evaluation method and system based on machine learning | |
Suriani et al. | Optimal accelerometer placement for fall detection of rehabilitation patients | |
Lv et al. | Information collection system for fall detection of stroke patients under cascade algorithm in the context of multi‐modal information fusion and e‐health | |
Bisio et al. | Towards IoT-based eHealth services: A smart prototype system for home rehabilitation | |
Zhang et al. | The prediction and error correction of physiological sign during exercise using Bayesian combined predictor and naive Bayesian classifier | |
Zhang et al. | Context-aware fall detection using a bayesian network | |
CN106781245A (en) | Inmate based on wearable device has a fist fight the method and system of early warning | |
Pereira et al. | Fall detection on ambient assisted living using a wireless sensor network | |
CN115462782A (en) | Human body falling dynamic monitoring method and system based on multi-dimensional characteristic parameters |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190917 |