CN115035686A - Real-time falling detection method, system and medium based on channel state information - Google Patents
Real-time falling detection method, system and medium based on channel state information Download PDFInfo
- Publication number
- CN115035686A CN115035686A CN202210475689.8A CN202210475689A CN115035686A CN 115035686 A CN115035686 A CN 115035686A CN 202210475689 A CN202210475689 A CN 202210475689A CN 115035686 A CN115035686 A CN 115035686A
- Authority
- CN
- China
- Prior art keywords
- data
- time domain
- matrix
- real
- effective
- 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 51
- 230000009471 action Effects 0.000 claims abstract description 90
- 239000011159 matrix material Substances 0.000 claims abstract description 61
- 238000000034 method Methods 0.000 claims abstract description 31
- 239000012634 fragment Substances 0.000 claims abstract description 25
- 238000000605 extraction Methods 0.000 claims abstract description 24
- 238000013145 classification model Methods 0.000 claims abstract description 18
- 230000009466 transformation Effects 0.000 claims abstract description 13
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 239000000969 carrier Substances 0.000 claims abstract description 5
- 230000009467 reduction Effects 0.000 claims abstract description 4
- 230000033001 locomotion Effects 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 10
- 230000000750 progressive effect Effects 0.000 claims description 6
- 230000002250 progressing effect Effects 0.000 claims description 5
- 230000035945 sensitivity Effects 0.000 claims description 4
- 101001041669 Oryctolagus cuniculus Corticostatin 1 Proteins 0.000 abstract 2
- 230000000875 corresponding effect Effects 0.000 description 41
- 230000006870 function Effects 0.000 description 18
- 238000004364 calculation method Methods 0.000 description 8
- 239000006185 dispersion Substances 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 6
- 238000005070 sampling Methods 0.000 description 6
- 238000005452 bending Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 5
- 238000011897 real-time detection Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000010606 normalization Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000001960 triggered effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004438 eyesight Effects 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 230000016776 visual perception Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
- G08B21/0407—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis
- G08B21/043—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis detecting an emergency event, e.g. a fall
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Heart & Thoracic Surgery (AREA)
- Emergency Management (AREA)
- Gerontology & Geriatric Medicine (AREA)
- General Physics & Mathematics (AREA)
- Physiology (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Social Psychology (AREA)
- Psychology (AREA)
- Pathology (AREA)
- Psychiatry (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Quality & Reliability (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to the field of fall detection, and particularly discloses a real-time fall detection method, a system and a medium based on channel state information, wherein the method comprises the steps of acquiring CS I data streams in real time and acquiring effective carriers from the CS I data streams; sequentially carrying out matrix extraction on the effective carrier waves corresponding to each data packet, and carrying out sliding window interception on the effective carrier waves according to the amplitude disturbance judgment condition to obtain effective action fragments; when an effective action fragment is intercepted, carrying out noise reduction on an effective carrier in a current window by utilizing wavelet transformation, and filtering by utilizing a Butterworth filter to obtain an initial available signal; extracting characteristic values of the initial available signals to obtain signal characteristics; normalizing the signal characteristics to obtain a characteristic value matrix; and importing the characteristic value matrix into an SVM classification model to obtain an action classification result. The invention has high detection efficiency, high detection accuracy and less occupation of operation resources, thereby reducing the equipment cost.
Description
Technical Field
The invention relates to the field of fall detection, in particular to a real-time fall detection method, a real-time fall detection system and a real-time fall detection medium based on channel state information.
Background
With the aging process of Chinese population becoming more and more rapid, the number of empty nesters increases gradually, and the health problems of the elderly become the focus of social attention. The falling is one of the main reasons of death of old people due to injury, and the number of the falling in the room of the old people living alone is continuously increased every year; studies have shown that the medical outcome after a fall depends largely on whether the response and rescue time are timely. In clinical medicine, the longer the therapy is delayed after a fall, the greater the risk of death. Therefore, how to realize an automatic and real-time fall detection technology becomes an important requirement for health monitoring of the elderly.
Up to now, health detection means for daily life has become common and various applications have been widely developed. There are three types of development directions for techniques to detect falls: sensor-based fall detection, machine vision-based fall detection, and WiFi signal-based Channel State Information (CSI) fall detection. The fall detection technology based on the sensor has high accuracy under ideal conditions, but is easily influenced by the environment, has relatively low accuracy and cannot be widely deployed; the falling detection technology based on vision can achieve high accuracy, but is high in cost, easy to receive the influence of factors such as illumination, the position of a camera, the background and the like, and can increase the risk of privacy disclosure; the fall detection method based on the WiFi-CSI has the advantages of strong privacy protection, wide application base, no need of carrying any sensor, non-visual perception, no influence of light humidity, temperature and the like, strong expandability and the like. In recent years, the internet of things technology is rapidly developed and widely applied, the universality of WiFi is greatly improved, and the realization environment and equipment support is provided for a WiFi-based fall detection system.
However, although the existing WiFi-CSI-based fall detection schemes can determine the occurrence of falls, the existing WiFi-CSI-based fall detection schemes lack real-time detection technologies, have a large improvement space for detection timeliness, and occupy high computational resources, and when the technology is applied to products, the existing WiFi-CSI-based fall detection schemes lack real-time application technologies, and meanwhile, the detection precision, accuracy and computational speed still need to be improved, and the hardware deployment cost and energy consumption are reduced.
Disclosure of Invention
In order to overcome the above problems, the present invention provides a real-time fall detection method, system and medium based on channel state information.
The invention provides a real-time falling detection method based on channel state information, which comprises the following steps:
connecting wifi signal sending equipment by using a TCP protocol, and acquiring a data stream of CSI in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressing data packets;
analyzing the data stream to obtain effective carriers;
presetting an amplitude disturbance judgment condition; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
when an effective action fragment is intercepted, carrying out noise reduction on an effective carrier in a current window by using wavelet transformation, and carrying out filtering by using a Butterworth filter to obtain an initial available signal;
extracting characteristic values of the initial available signals to obtain signal characteristics;
normalizing the signal characteristics to obtain a characteristic value matrix;
importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
and judging whether the action classification result is a fall or not, and if so, sending alarm information.
Preferably, the signal characteristics include: the method comprises the steps of time domain mean, time domain standard deviation, time domain maximum, time domain minimum, time domain range, over-mean point number, time domain 1/4 quantiles, time domain 3/4 quantile and time domain quartile range, first large FFT, frequency corresponding to the first large FFT, third large FFT, frequency corresponding to the third large FFT, fifth large FFT, frequency corresponding to the fifth large FFT, frequency domain average value, frequency domain standard deviation, frequency domain 1/4 quantile, frequency domain 3/4 quantile, frequency domain quartile range, amplitude statistical skewness, amplitude statistical kurtosis, shape statistical mean, shape statistical standard deviation, shape statistical skewness and shape statistical kurtosis.
Preferably, the characteristic value extraction is performed on the initial available signal to obtain a signal characteristic, specifically:
performing wavelet transformation on the initial available signals again to obtain two wavelet signals, and performing characteristic value extraction on the two wavelet signals to obtain signal characteristics;
the signal characteristics comprise a time domain second mean value, a time domain second standard deviation, a time domain second range, a time domain second mean value point number and a time domain second quartile range, which are extracted from one wavelet signal; and extracting a time domain third mean value, a time domain third standard deviation, a time domain third range, a time domain third mean value point number and a time domain third quartile range from another wavelet signal.
Preferably, the device is connected with wifi signal sending equipment by using a TCP protocol, and acquires a CSI data stream in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressing data packets; the method specifically comprises the following steps:
setting at least one wifi signal sending device and m wifi signal receiving devices in a space to be tested, and simultaneously acquiring data sent by the at least one wifi signal sending device through the m wifi signal receiving devices to obtain 1 dat file; wherein m is a positive integer;
extracting the data stream of the csi from the acquired dat file; the data stream comprises a plurality of data packets which are continuously progressive, and each data packet corresponds to a1 x m x 30 subcarrier matrix.
Preferably, the analyzing the data stream obtains an effective carrier from the data stream, specifically:
at preset time intervals, the sensitivities of the 1 × m × 30 subcarriers to the motion are compared, and the subcarrier most sensitive to the motion is selected as the effective carrier.
Preferably, the preset amplitude disturbance judgment condition; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not; the method comprises the following specific steps:
presetting amplitude disturbance judgment conditions, wherein the amplitude disturbance judgment conditions comprise a segment starting condition and an amplitude disturbance threshold;
sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
intercepting a first action segment in the data stream through a sliding window according to a segment starting condition;
and detecting whether the first action segment is larger than the amplitude disturbance threshold value in real time, and if so, outputting an effective action segment.
Preferably, presetting an amplitude disturbance judgment condition; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; wherein, the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window, and the method further comprises the following steps:
acquiring data samples of different movements of a human body at different spatial positions, wherein the data samples comprise at least one fallen data sample and at least one non-fallen data sample;
and marking the data samples, and distinguishing amplitude disturbance differences of fallen data samples and non-fallen data samples to obtain amplitude disturbance judgment conditions.
The invention also provides real-time fall detection equipment based on the channel state information, which comprises the following components: the device comprises a data acquisition module, a carrier acquisition module, an action detection module, a first data module, a second data module, a third data module, a classification module and an alarm module;
the data acquisition module is used for connecting wifi signal sending equipment by using a TCP protocol and acquiring CSI data flow in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressing data packets;
the carrier acquisition module is used for analyzing the data stream and acquiring an effective carrier from the data stream;
the action detection module is used for presetting an amplitude disturbance judgment condition; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
the first data module is used for denoising the effective carrier wave in the current window by utilizing wavelet transformation and filtering by utilizing a Butterworth filter to obtain an initial available signal when an effective action segment is intercepted;
the second data module is used for extracting characteristic values of the initial available signals to obtain signal characteristics;
the third data module is used for normalizing the signal characteristics to obtain a characteristic value matrix;
the classification module is used for importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
the alarm module is used for judging whether the action classification result is falling down, and if yes, sending alarm information.
The invention provides a computer-readable storage medium, which comprises a stored computer program, wherein when the computer program runs, the device where the computer-readable storage medium is located is controlled to execute the real-time fall detection method based on the channel state information.
The invention has the beneficial effects that:
(1) the CSI data stream is obtained in real time, effective action fragments are intercepted through a sliding window according to amplitude disturbance judgment conditions and are put into an SVM classification model for analysis, a current human body state result is obtained, a system for detecting falling down in real time can be carried out, data can be analyzed in real time, and a conclusion can be obtained and output in a very short time.
(2) Real-time uninterrupted detection is carried out through a tcp protocol connecting channel, real-time rapid falling monitoring is achieved, when a small amount of packet loss occurs in a link, the tcp protocol can automatically retransmit, interference is not needed, accuracy of information is guaranteed, and detection efficiency is improved.
(3) The judgment of action generation is carried out by using the amplitude disturbance condition, action fragments can be effectively segmented, the detection efficiency is improved, the occupation of operation resources is reduced, and the equipment cost is reduced.
Preferably, a plurality of characteristic values of a time domain and a frequency domain are used for judgment, an SVM classification model is put into the SVM classification model for trial, and 36 characteristic values based on the time domain and the frequency domain are obtained as an optimal judgment basis, so that the classification accuracy is greatly improved, and the detection accuracy is improved.
Drawings
The invention will be further described with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of a method according to one embodiment of the present invention;
FIG. 2 is a diagram illustrating waveforms of wavelet transformed data according to an embodiment of the present invention;
FIG. 3 is a CSI amplitude graph of a bow according to one embodiment of the present invention;
fig. 4 is a CSI amplitude diagram when one embodiment of the invention falls;
FIG. 5 is a table of human body action frequency parameters according to one embodiment of the present invention;
fig. 6 is a schematic view of a collection site according to another embodiment of the present invention.
In the figure: 1. a transmitting antenna; 2. a receiving antenna; 3. and (4) data acquisition points.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1 to 5, as one implementation of the present invention, a real-time fall detection method based on channel state information is disclosed, which includes:
s1, connecting wifi signal sending equipment by using a TCP (transmission control protocol), and acquiring CSI data stream in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressing data packets;
s2, analyzing the data flow to obtain effective carrier waves;
s3, presetting amplitude disturbance judgment conditions; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
s4, when the effective action segment is intercepted, denoising the effective carrier wave in the current window by utilizing wavelet transformation, and filtering by utilizing a Butterworth filter to obtain an initial available signal;
s5, extracting characteristic values of the initial available signals to obtain signal characteristics;
s6, normalizing the signal characteristics to obtain a characteristic value matrix;
s7, importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
and S8, judging whether the action classification result is falling, and if yes, sending alarm information.
Preferably, the signal characteristics include: the method comprises the steps of time domain mean, time domain standard deviation, time domain maximum, time domain minimum, time domain range, number of over-mean points, time domain 1/4 quantiles, time domain 3/4 quantile and time domain quartile range, first large FFT, frequency corresponding to the first large FFT, third large FFT, frequency corresponding to the third large FFT, fifth large FFT, frequency corresponding to the fifth large FFT, frequency domain average, frequency domain standard deviation, frequency domain 1/4 quantile, frequency domain 3/4 quantile, frequency domain quartile range, amplitude statistical skewness, amplitude statistical kurtosis, shape statistical mean, shape statistical standard deviation, shape statistical skewness and shape statistical kurtosis.
Preferably, step S4 includes the following sub-steps:
s41, performing wavelet transformation on the initial available signals again to obtain two wavelet signals, and performing characteristic value extraction on the two wavelet signals to obtain signal characteristics;
the signal characteristics comprise a time domain second mean value, a time domain second standard deviation, a time domain second range, a time domain second mean value point number and a time domain second quartile range, which are extracted from one wavelet signal; and extracting a time domain third mean value, a time domain third standard deviation, a time domain third range, a time domain third mean value point number and a time domain third quartile range from another wavelet signal.
Referring to fig. 2, the first waveform from top to bottom in the figure is the waveform of the raw data (without any processing); the second waveform is the waveform of the processed data (after wavelet transform denoising, filter filtering); the third waveform is the low-frequency component signal of the second waveform, i.e., the wavelet signal s (t)1 of the two wavelet signals; the fourth waveform is a high-frequency component signal of the second waveform, i.e., the wavelet signal s (t)2 of the two wavelet signals.
Aiming at the two points, one of the innovative points of the present invention is to perform wavelet transform denoising on the original CSI signal and use a butterworth filter for filtering to obtain an initial available signal s (t), and perform feature extraction on the signal by two methods:
the first method is that the initial available signal s (t) is directly used for extracting the characteristic values of the time domain and the frequency domain to obtain 26 characteristic values in total;
in the second method, the initial usable signal s (t) is wavelet transformed again, the initial usable signal s (t) is decomposed into two wavelet signals s (t)1 and s (t)2, and the two signals are time domain feature extracted again, including the mean value, the standard deviation, the range, the number of over-mean points and the range of quartile, wherein the five feature values can only show the signal characteristics, and the total number of the five feature values is 10.
The two methods have 36 characteristic values, and both methods can perfectly show the characteristics of the signal, so that the accuracy of fall identification is improved. Wherein, the specific definition of each characteristic value is as follows:
and the time domain mean value is the average characteristic of the data signal in a time window, and the mean value of multiple sampling is used as a data true value to reflect the basic characteristics of the data true value. Mainly reflects the static characteristics of received signals and can distinguish the approximate position information of human bodies; the calculation formula is as follows:
wherein n denotes the size of a time window, a i Represents the magnitude of the ith subcarrier, which corresponds to the available function mean () in Matlab.
The time domain standard deviation is obtained by subtracting the sum of squares of the mean values of all the numbers in the data set, dividing the obtained result by the total number, and then carrying out root formation on the obtained result. The standard deviation reflects the degree of mean dispersion in the data set. Large standard deviations, indicating a large difference between these numbers and their mean values; the small standard deviations indicate that the differences between these numbers and their mean values are small. The fluctuation of the received signal can be reflected, and the change of the human body action near the receiving end is estimated; the calculation formula is as follows:
it corresponds to the functions available in Matlab: std ().
Time domain maximum and time domain minimum, respectively representing maximum and minimum values in the data set; the time domain range represents the difference between the maximum and minimum values and may represent the degree of signal variation. The time domain maximum value and the time domain minimum value can better distinguish the change of human body actions.
Wherein, the time domain maximum value: max is max (a) i ) I ∈ {1, 2,..., n }, corresponding to a function max ();
time domain minimum: min is min (a) i ) I ∈ {1, 2,.., n }, corresponding to a function min ();
time domain range: range ═ max-min |, corresponding to the function: abs (max () -min ());
the number of the over-average points refers to the number of data exceeding the average point in a window, and the calculation formula is as follows:
wherein,
is an indicator function (indicator function), and takes a value of 1 when the condition in the parenthesis holds, otherwise takes a value of 0.
The time domain 1/4 quantiles, the time domain 3/4 quantiles, the time domain quartile range, the frequency domain 1/4 quantiles, the frequency domain 3/4 quantiles and the frequency domain quartile range, wherein the time domain and the frequency domain 1/4 quantiles, 3/4 quantiles and the quartile range are the same in meaning.
The quartile parameters refer to the dispersion degree of the skewed data when the data are described by a quartile statistical description analysis method, that is, the number of all the data arranged from small to large and just arranged at the lower 1/4 position is called a lower quartile (according to percentage, namely the number at the 25% position) and a first quartile, the number arranged at the upper 1/4 position is called an upper quartile (according to percentage, namely the number at the 75% position) and a third quartile, and the interquartile distance refers to the difference between the upper quartile and the lower quartile. The upper and lower quartiles contain exactly 50% of the data between them.
The four-point difference is characterized in that:
the interquartile range only indicates the dispersion degree of the middle 50% of data, and still cannot sufficiently reflect the dispersion condition of all data. The larger the interquartile difference, the larger the dispersion degree of the middle 50% data; the smaller the interquartile difference, the smaller the dispersion of the middle 50% data; to a certain extent, the quartile difference can also reflect the representative quality of the median; quartile range is a sequential statistic, so quartile range is applicable to measure the degree of dispersion of sequencing data and quantitative data.
The frequency corresponding to the first large FFT, the frequency corresponding to the third large FFT, the frequency corresponding to the fifth large FFT and the frequency corresponding to the fifth large FFT:
and (3) converting the csi time domain signals in an action time window into the csi signal information of a frequency domain through Fast Fourier Transform (FFT), and extracting the first five maximum FFT values of the signals and the frequencies corresponding to the five maximum FFT values respectively. In the experimental process, two groups of characteristic values of the second large FFT and the third large FFT, and the fourth large FFT and the fifth large FFT in the frequency domain have completely same characteristic change values, so that only one item of the second large FFT and the third large FFT, and only one item of the fourth large FFT and the fifth large FFT are adopted.
For example, there are 8: 1, 2, 3, 4, 5, 6, 7, 8;
after FFT the following 8 numbers were obtained:
36.0000+0.0000i；-4.0000+9.6569i；-4.0000+4.0000i；
-4.0000+1.6569i；-4.0000+0.0000i；-4.0000-1.6569i；
-4.0000-4.0000i；-4.0000-9.6569i。
from these 8 numbers, it can be seen that the first number is the largest and the remaining 7 numbers are symmetric around the fifth number. This is determined by the fourier transform, so the remaining number can be only half except for the first number alone.
The following is the power spectral density, which describes the energy distribution of the data in the frequency domain; the power spectral density is divided into two characteristics of amplitude statistical characteristic and shape statistical characteristic
Firstly, the amplitude statistical characteristic is set as follows, C (i) is the frequency amplitude value of the ith window, N represents the number of windows:
frequency domain mean (amplitude statistical feature mean):
frequency domain standard deviation (amplitude statistical characteristic standard deviation):
amplitude statistical skewness:
amplitude statistics kurtosis:
the following is the statistical characteristics of the shape, let C (i) be the frequency amplitude value of the ith window, N represent the number of windows,
then several quantities of shape statistics are calculated as follows
Shape statistical mean:
shape statistical standard deviation:
shape statistical skewness:
shape statistical kurtosis:
preferably, step S1 includes the following sub-steps:
s11, arranging at least one wifi signal sending device and m wifi signal receiving devices in a space to be tested, and simultaneously obtaining data sent by the at least one wifi signal sending device through the m wifi signal receiving devices to obtain 1 dat file; wherein m is a positive integer;
s12, extracting the data stream of the csi from the acquired dat file; the data stream comprises a plurality of data packets which are continuously progressive, and each data packet corresponds to a1 x m x 30 subcarrier matrix.
Preferably, step S2 includes the following sub-steps:
and S21, comparing the sensitivity of the 1 x m x 30 sub-carriers to the motion at preset time intervals, and selecting the sub-carrier most sensitive to the motion as the effective carrier.
Preferably, step S3 includes the following sub-steps:
s31, presetting amplitude disturbance judgment conditions, wherein the amplitude disturbance judgment conditions comprise a segment starting condition and an amplitude disturbance threshold value;
s32, matrix extraction is carried out on the effective carrier waves corresponding to each data packet in sequence to obtain matrix parameters corresponding to each data packet; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
s33, intercepting a first action fragment in the data stream through a sliding window according to the fragment starting condition;
and S34, detecting whether the first action segment is larger than the amplitude disturbance threshold value in real time, and if so, outputting an effective action segment.
Preferably, the step S3 further includes the following sub-steps:
s311, collecting data samples of different actions of a human body at different spatial positions, wherein the data samples comprise at least one fallen data sample and at least one non-fallen data sample;
and S312, marking the data samples, distinguishing amplitude disturbance differences of falling data samples and non-falling data samples, and obtaining amplitude disturbance judgment conditions.
In this embodiment, the segment start condition is common to any scene, and the segment start condition at least includes an action data set for falling and bending; with 6s for one dat data file, the time interval between packets is 0.01s, i.e. the sampling frequency is 100HZ, i.e. one data file corresponds to 600 packets. When the data set is measured, whether the user falls or bends, the user is specified that the user is uniformly still and motionless in the first 2s, and the user executes the action after the first 2s, so that the user selects an amplitude interval between the 250 th packet and the 300 th packet as the start of a judgment fragment under the condition of amplitude disturbance so as to match the collected data set.
An amplitude disturbance threshold value which is a preset threshold value through manual observation; under different environments, the effects generated by the multipath effect are different, and the threshold value cannot be automatically configured once and for all, so that when the embodiment is applied to other scenes, the amplitude value disturbance threshold value needs to be reconfigured.
Referring to fig. 3 and 4, it can be seen that, when testing the motion data set, the amplitude change range is 300-400, where due to human error (the time concept of each person is not consistent), the time shift is caused, and the present embodiment uniformly determines the amplitude between the 250 th packet and the 300 th packet; it can also be seen from fig. 3 and 4 that the amplitude sudden changes occur in both the falling and bending actions, and in this embodiment, the data stream that does not have sudden changes in the judgment interval is filtered by setting the amplitude disturbance judgment condition. The setting of the amplitude disturbance threshold value is related to the environment, and when deployment is needed, relevant actions are simulated, one or more groups of data files are received in advance, the amplitude mutation range is observed, and then the threshold value is set.
In other embodiments, after the wifi device is deployed, the user is guided to do several groups of actions through voice through a preset program, so that the amplitude sudden change range is calculated through one or more collected data files according to a proportion or a preset algorithm, an amplitude disturbance threshold value is obtained, and the preset is automatically completed.
Referring to fig. 5, it can be seen that the motion range of the human body is mainly concentrated in 0 to 80Hz, so the butterworth filter is selected as the low pass filter in the embodiment.
As another embodiment of the present invention, the present invention is a WIFI-based real-time detection system in a static unmanned ideal environment, which comprises a signal sending module, i.e. a WIFI router, a data receiving module, i.e. a WIFI receiving antenna 2, a data processing and SVM classification model, and a real-time detection module, i.e. a microcomputer. There are 1 transmitting antenna 1 and 3 receiving antennas 2.
The implementation method of the embodiment comprises the following steps:
a1, environment construction, wherein the embodiment adopts a version 14.04 of an Ubuntu system, and after the system is installed on a microcomputer, a CSI tool is adopted to modify a driver and a kernel according to a deployed environment, so as to extract CSI subcarrier information from an intel 5300 network card; from the intel 5300 network card, the present embodiment can extract 30 subcarriers, and each data packet includes 30 subcarriers;
a2, collecting data by a system, wherein the embodiment adopts two modes to collect data by sampling, namely, the time of 6s is specified, and the data packet is written into a file with a dat format; the other method is that the specified 600 packet numbers are adopted to write the data packets into the file with the dat format; wherein, both modes correspond to a sampling rate of 100HZ, that is, the time interval between packets is 0.01 s;
a3, acquiring actions, arranging experimenters to go to a data acquisition point 3 to perform specified actions, and acquiring corresponding data; the present embodiment is planned as follows within 6 seconds: the first 2s, standing still on the boundary of the experimental region range, as shown in fig. 6, after two seconds, going to the data acquisition point 3, then executing corresponding action, and after executing the action, standing still;
a4, signal processing, comprising the following steps:
a4.1, preprocessing signals, namely reading a dat file, and then extracting csi subcarriers from the dat file by utilizing an extraction function contained in a disclosed csi tool kit, wherein csi is extracted in a frequency response mode, so that the csi data format is a complex number, namely a + bi mode, and | a + bi | is taken as a basic signal of the system through modulus (namely amplitude) extraction;
a4.2, signal denoising and filtering, wherein the data packet is denoised through wavelet transformation, and is filtered through a Butterworth filter to obtain an initial available signal;
a4.3, feature extraction, namely extracting feature values of the initial available signals to obtain signal features; the step is mainly to improve the calculation power through the optimization of the algorithm, and the formula is obtained by integrating a large number of papers;
a4.4, feature normalization, wherein two possible schemes are provided at the beginning of the step, one is normalization, the other is normalization, and the practice proves that the variance of some feature values is extremely small, so that the data is amplified, and therefore the normalization is adopted in the embodiment;
a4.5, SVM classification, namely determination of hyper-parameters c and g of a vector machine in an SVM classification model by adopting a code set related to libsvm, and finding the optimal hyper-parameters c and g of each training set by adopting functions of the optimal hyper-parameters c and g;
a5, connecting a computer and a router through a TCP protocol to realize real-time data transmission, locally setting the computer as a server, then connecting the router, and setting the router as a client to realize TCP transmission;
a6, forming a function _ model function by a signal processing model, setting an amplitude disturbance condition, and ensuring the execution of the system through double while loops, wherein the first layer of while loops are disconnected after the connection time without data is over long and can wait for the connection again, and the second layer of while loops are used for ensuring the real-time collection, updating the data and executing the model; and finally, outputting a result of judging each action.
The sampling rate used in this embodiment is 100HZ and the data packets are 6s intervals, then theoretically, a data packet should contain 600 packets. However, since the quality of the channel link cannot be guaranteed to meet the optimal condition, and a delay of a packet may be slightly long, so that 600 packets cannot be received when the packet is received within a predetermined time, in this embodiment, the packet number of the data packet is counted each time the packet is received, and if the packet number cannot receive 600 packets after 6s at a sampling rate of 100HZ in the current link environment, the embodiment will adopt that 600 packets are received even if the time exceeds 6 s.
In this embodiment, the number of the sending antennas 1 of the router is 1, and the number of the receiving antennas of this embodiment is 3, so after the received dat files are extracted into csi, one csi corresponds to one matrix, and one subcarrier matrix is 1 × 3 × 30. The system model of this embodiment does not require all subcarriers received by 3 antennas, so we compare the sensitivity of each antenna to the motion by observation, we select the antenna most sensitive to the motion as the antenna we need, and we select the third antenna after many tests; and the purpose of this embodiment is to perform data analysis with time as axis, so that only one of the 30 sub-carriers is needed, and therefore, the sub-carrier most sensitive to action is taken here, i.e. the first sub-carrier is used. The step is the key for improving the operation efficiency, most people use the distribution variance of 30 subcarriers to replace the distribution variance, so the operation cost is greatly improved, and the efficiency is reduced.
In the embodiment, 9 data acquisition points 3 are provided, each point acquires 30 dat files, and there are two movements in total, namely a bending movement and a falling movement (because the bending movement is approximately similar from the view point of a data graph), and 5 experimenters, and there are 2700 data sets in total, so that data errors caused by the geographic position are well reduced.
The wavelet transformation is used for denoising, after a large amount of data research and analysis and an actual test experiment, wavelet transformation denoising is found to be the most suitable mode for csi signal denoising, and then the fact that the action frequencies of different parts of a human body are different and the hand action is short in period and high in frequency is known; falls are long, infrequent; therefore, the rough frequency range is calculated to be 0-80 HZ by calculating the action period, the frequency range is expanded in a small range, action details are prevented from being lost, and irrelevant signals outside the frequency range are filtered out by using a Butterworth filter so as to improve the accuracy.
The embodiment has two versions of the optimizing function, one is SVMcg, and the other is SVMcgPP; the SVMcgPP is an enhanced version of SVMcg, and both the SVMcgP and the SVMcgPP use a grid searching method, namely a stepping method (enumeration method) to find out the optimal hyperparameter c, g. In this embodiment, an enumeration function is used to obtain an optimal SVM classification model, which is obtained by: and setting a certain parameter range, setting a gradient, and continuously training and comparing results to obtain the SVM classification model with the maximum accuracy and the optimal corresponding parameter value. Although the method is slow in generating the model and needs to occupy certain calculation resources, the final SVM classification model is generated, and after the model is configured in an actual application scene, the model is extremely fast in application speed, occupies few calculation resources, and can be applied to most scenes.
As another embodiment of the present invention, the method comprises the following steps:
1, performing real-time data collection, opening a local 8090 port through a tcpip function by using a TCP (transmission control protocol) protocol to establish the port as a server, and setting the input buffer size and the waiting time of the server; inputting a command line at a computer terminal, connecting an 8090 port, and waiting for system collection; through a packet sending command, the CSI Tool performs bidirectional CSI acquisition in a monitoring mode, namely simultaneously sends and receives data;
b2, judging whether an action is generated, firstly extracting a csi matrix, judging whether the action is generated according to an amplitude disturbance judgment condition, and intercepting effective action fragments through a sliding window according to the amplitude disturbance judgment condition; if the amplitude disturbance judgment condition is met, performing the next step, and if the amplitude disturbance judgment condition is not met, continuously collecting the data;
b3, carrying out data processing, calling a function _ model function, carrying out data preprocessing, carrying out noise reduction and filtering by using a Butterworth filter, and then carrying out characteristic value extraction on the processed data, wherein the characteristic values comprise 26 time domain characteristic values and frequency domain characteristic values; after obtaining the 26 eigenvalue matrixes, normalizing the 26 eigenvalue matrixes;
and B4, classifying the action, namely putting the normalized characteristic value matrix into an SVM classification model for classification, and rapidly obtaining the action type.
This embodiment mainly collects data for training the model in the following cases:
the first condition is as follows: the router and the microcomputer are started, so that an experimental scene is exposed to a WIFI environment, the WIFI signal can be stably received by the data receiving module, the matlab on the microcomputer is opened, and a real-time detection program is executed to prepare for receiving data.
The specific operation comprises the following steps:
firstly, opening matlab through a command on a Linux system;
secondly, using a self-writing read _ quiethcess.m file on port connection and data transmission;
thirdly, in the TCP protocol, a local IP port 8090 is used as a server, and port parameters are set;
in the program, the CSI data stream acquired in real time is placed into an ap1_ mp1 array, the last bit of the ap1_ mp1 array is written into each CSI data stream, and each bit of the ap1_ mp1 array is pushed forward by one bit through for circulation so that each received datum is continuously transmitted in the ap1_ mp1 array;
fifthly, establishing a pandean array, and putting every 250 th data value of ap1_ mp1 into the last bit of the pandean, so that the data of the pandean array and the ap1_ mp1 array are continuously updated and transmitted, and the pandean array is a variable applied to the condition of the amplitude disturbance judgment condition;
setting amplitude disturbance judgment conditions, if the conditions are met, putting the effective action fragments under the current window into an SVM classification model for classification judgment, and if the conditions are not met, continuously receiving the effective action fragments, greatly reducing the amount of calculation and improving the calculation efficiency;
and finally, if the effective fragments are subjected to classification judgment, resetting the current effective fragments, namely the panduan array and the ap1_ mp1, and avoiding interference caused by repeated judgment.
And opening a new terminal in the microcomputer, performing log _ to _ server operation, inputting a local IP address to realize the connection with an 8090 port, and performing a ping command to enable a local server end of the computer to receive channel information from the router and transmit the channel information to a matlab program.
The specific operation is as follows:
firstly, entering a netlink file, performing log _ to _ server, and inputting a local IP address to realize the communication with a local 8090 port;
secondly, inputting a ping command to enable the local server to receive router channel information and transmit the router channel information to the matlab program;
thirdly, according to the method in the second step, the signals received by the matlab are processed, amplitude disturbance conditions are judged, and finally classification is carried out;
and fourthly, if the classification model judges that the person falls down, displaying that the person falls down, and if the person does not judge that the person falls down, displaying that the person does not fall down.
Because the ideal environment of no person is adopted in the situation, the amplitude disturbance judgment condition can not be triggered.
And a second condition: in the real-time detection system based on WIFI in the target stooping environment, the operation steps of the second case are basically consistent with those of the first case, and the first case is a static unmanned environment, while the second case is a target stooping environment.
In this case, the target makes a bowing motion between the receiving-end antenna and the transmitting-end antenna to cause disturbance, the bowing causes a large fluctuation in the image, and then the curve returns to the original position, as shown in fig. 3.
Since the target bending environment is adopted in the present case, the amplitude disturbance judgment condition is triggered, but the classification result belongs to non-falling.
Case three: in the present case, the real-time detection system based on WIFI is used in an environment where the target falls, operation steps of the case three and the case two and the case one are basically the same, and the case three is an environment where the target falls.
In this case, the disturbance is made by making a fall between the receiving end antenna and the transmitting end antenna, the fall causes the image to fluctuate greatly, but the image of the fall action and the image of the bow action have different trends, and the difference is that the bow is the original position of the return curve after the disturbance, and the fall is to keep the position of the return curve after the disturbance, as shown in fig. 4.
In this case, the target stooping environment is adopted, so that an amplitude disturbance judgment condition is triggered, and the classification result is that the object falls down, and at the moment, alarm information is sent out.
The invention also discloses a real-time falling detection device based on the channel state information, which comprises: the device comprises a data acquisition module, a carrier acquisition module, an action detection module, a first data module, a second data module, a third data module, a classification module and an alarm module;
the data acquisition module is used for connecting wifi signal sending equipment by using a TCP protocol and acquiring CSI data stream in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressive data packets;
the carrier acquisition module is used for analyzing the data stream and acquiring an effective carrier from the data stream;
the action detection module is used for presetting amplitude disturbance judgment conditions; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
the first data module is used for denoising the effective carrier in the current window by using wavelet transformation and filtering by using a Butterworth filter when an effective action fragment is intercepted, so as to obtain an initial available signal;
the second data module is used for extracting characteristic values of the initial available signals to obtain signal characteristics;
the third data module is used for normalizing the signal characteristics to obtain a characteristic value matrix;
the classification module is used for importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
the alarm module is used for judging whether the action classification result is falling down, and if yes, sending alarm information.
The invention also discloses a terminal device, which comprises a processor and a storage device, wherein the storage device is used for storing one or more programs; when the one or more programs are executed by the processor, the processor implements the above-described real-time fall detection method based on channel state information. The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, which is the control center for the test equipment and connects the various parts of the overall test equipment using various interfaces and lines.
The storage device may be used to store computer programs and/or modules, and the processor may implement various functions of the terminal device by running or executing the computer programs and/or modules stored in the storage device and calling data stored in the storage device. The storage device may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the terminal device, and the like. In addition, the storage device may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
Wherein, the integrated module/unit of the real-time fall detection device based on the channel state information can be stored in a computer readable storage medium if it is implemented in the form of a software functional unit and sold or used as a stand-alone product. Based on such understanding, all or part of the flow in the method according to the embodiments of the present invention may also be implemented by a computer program, which may be stored in at least one computer-readable storage medium and used for instructing related hardware to implement the steps of the above-described embodiments of the method when executed by a processor. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying computer program code, recording medium, U.S. disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution media, and the like.
It should be noted that the above-described embodiments of the apparatus and device are merely schematic, where units illustrated as separate components may or may not be physically separate, and components illustrated as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
The above-mentioned embodiments are provided to further explain the objects, technical solutions and advantages of the present invention in detail, and it should be understood that the above-mentioned embodiments are only examples of the present invention and are not intended to limit the scope of the present invention. It should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (9)
1. The real-time fall detection method based on the channel state information is characterized by comprising the following steps:
connecting wifi signal sending equipment by using a TCP protocol, and acquiring a data stream of CSI in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressive data packets;
analyzing the data stream to obtain effective carriers;
presetting an amplitude disturbance judgment condition; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
when an effective action fragment is intercepted, carrying out noise reduction on an effective carrier in a current window by using wavelet transformation, and carrying out filtering by using a Butterworth filter to obtain an initial available signal;
extracting characteristic values of the initial available signals to obtain signal characteristics;
normalizing the signal characteristics to obtain a characteristic value matrix;
importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
and judging whether the action classification result is a fall or not, and if so, sending alarm information.
2. The real-time fall detection method based on channel state information according to claim 1, wherein: the signal characteristics include: the method comprises the steps of time domain mean, time domain standard deviation, time domain maximum, time domain minimum, time domain range, number of over-mean points, time domain 1/4 quantiles, time domain 3/4 quantile and time domain quartile range, first large FFT, frequency corresponding to the first large FFT, third large FFT, frequency corresponding to the third large FFT, fifth large FFT, frequency corresponding to the fifth large FFT, frequency domain average, frequency domain standard deviation, frequency domain 1/4 quantile, frequency domain 3/4 quantile, frequency domain quartile range, amplitude statistical skewness, amplitude statistical kurtosis, shape statistical mean, shape statistical standard deviation, shape statistical skewness and shape statistical kurtosis.
3. The real-time fall detection method based on channel state information according to claim 1, wherein the feature value extraction is performed on the initial available signal to obtain signal features, specifically:
performing wavelet transformation on the initial available signals again to obtain two wavelet signals, and performing characteristic value extraction on the two wavelet signals to obtain signal characteristics;
the signal characteristics comprise a time domain second mean value, a time domain second standard deviation, a time domain second range, a time domain second mean value point number and a time domain second quartile range, which are extracted from one wavelet signal; and extracting a time domain third mean value, a time domain third standard deviation, a time domain third range, a time domain third mean value point number and a time domain third quartile range from another wavelet signal.
4. The real-time fall detection method based on the channel state information is characterized in that the method uses a TCP protocol to connect with a wifi signal sending device and obtains a CSI data stream in real time through a wifi signal receiving device; wherein the data stream comprises a number of consecutively progressive data packets; the method specifically comprises the following steps:
setting at least one wifi signal sending device and m (since n is a variable hereinafter, m is repeatedly changed for avoiding the phenomenon) wifi signal receiving devices in a space to be detected, and simultaneously obtaining data sent by at least one wifi signal sending device through the m wifi signal receiving devices to obtain 1 dat file; wherein m is a positive integer;
extracting the data stream of the csi from the acquired dat file; the data stream comprises a plurality of data packets which are continuously progressive, and each data packet corresponds to a1 x m x 30 subcarrier matrix.
5. The method for fall detection in real time based on channel state information as claimed in claim 4, wherein the analyzing the data stream to obtain the effective carrier is specifically:
at preset time intervals, the sensitivities of the 1 × m × 30 subcarriers to the motion are compared, and the subcarrier most sensitive to the motion is selected as the effective carrier.
6. The real-time fall detection method based on the channel state information according to claim 1, wherein the preset amplitude disturbance judgment condition is set; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not; the method specifically comprises the following steps:
presetting an amplitude disturbance judgment condition, wherein the amplitude disturbance judgment condition comprises a segment starting condition and an amplitude disturbance threshold;
sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
according to the segment starting condition, a first action segment in the data stream is intercepted through a sliding window;
and detecting whether the first action segment is larger than the amplitude disturbance threshold value in real time, and if so, outputting an effective action segment.
7. The real-time fall detection method based on the channel state information according to claim 1, wherein the preset amplitude disturbance judgment condition is set; sequentially carrying out matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and carrying out sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; wherein, the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window, and the method further comprises the following steps:
acquiring data samples of different movements of a human body at different spatial positions, wherein the data samples comprise at least one fallen data sample and at least one non-fallen data sample;
and marking the data samples, and distinguishing amplitude disturbance differences of fallen data samples and non-fallen data samples to obtain amplitude disturbance judgment conditions.
8. Real-time fall detection equipment based on channel state information is characterized by comprising: the device comprises a data acquisition module, a carrier acquisition module, an action detection module, a first data module, a second data module, a third data module, a classification module and an alarm module;
the data acquisition module is used for connecting wifi signal sending equipment by using a TCP protocol and acquiring CSI data stream in real time through wifi signal receiving equipment; wherein the data stream comprises a number of consecutively progressing data packets;
the carrier acquisition module is used for analyzing the data stream and acquiring an effective carrier from the data stream;
the action detection module is used for presetting amplitude disturbance judgment conditions; sequentially performing matrix extraction on the effective carrier wave corresponding to each data packet to obtain a matrix parameter corresponding to each data packet, and performing sliding window interception on the effective carrier wave according to the amplitude disturbance judgment condition to obtain an effective action fragment; the amplitude disturbance judgment condition is used for judging whether a plurality of matrix parameters containing corresponding human body actions exist in the sliding window or not;
the first data module is used for denoising the effective carrier wave in the current window by utilizing wavelet transformation and filtering by utilizing a Butterworth filter to obtain an initial available signal when an effective action segment is intercepted;
the second data module is used for extracting characteristic values of the initial available signals to obtain signal characteristics;
the third data module is used for normalizing the signal characteristics to obtain a characteristic value matrix;
the classification module is used for importing the characteristic value matrix into an SVM classification model to obtain an action classification result;
the alarm module is used for judging whether the action classification result is falling down, and if yes, sending alarm information.
9. A computer-readable storage medium, comprising a stored computer program, wherein when the computer program runs, the computer-readable storage medium controls an apparatus to execute the method for real-time fall detection based on channel state information according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210475689.8A CN115035686A (en) | 2022-04-29 | 2022-04-29 | Real-time falling detection method, system and medium based on channel state information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210475689.8A CN115035686A (en) | 2022-04-29 | 2022-04-29 | Real-time falling detection method, system and medium based on channel state information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115035686A true CN115035686A (en) | 2022-09-09 |
Family
ID=83118949
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210475689.8A Pending CN115035686A (en) | 2022-04-29 | 2022-04-29 | Real-time falling detection method, system and medium based on channel state information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115035686A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116058298A (en) * | 2023-03-06 | 2023-05-05 | 北京市农林科学院信息技术研究中心 | Livestock behavior monitoring method and device |
| CN116304844A (en) * | 2023-05-23 | 2023-06-23 | 山东科技大学 | Personnel entry and exit counting and counting system and method |
| CN116528358A (en) * | 2023-06-30 | 2023-08-01 | 华南师范大学 | Wi-Fi-based fall positioning joint detection method, device and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107749143A (en) * | 2017-10-30 | 2018-03-02 | 安徽工业大学 | A kind of indoor occupant fall detection system and method through walls based on WiFi signal |
| US20180198548A1 (en) * | 2017-01-08 | 2018-07-12 | At&T Intellectual Property I, L.P. | Interference cancelation for 5g or other next generation network |
| CN109394229A (en) * | 2018-11-22 | 2019-03-01 | 九牧厨卫股份有限公司 | A kind of fall detection method, apparatus and system |
| CN109635837A (en) * | 2018-11-10 | 2019-04-16 | 天津大学 | A kind of carefree fall detection system of scene based on commercial wireless Wi-Fi |
| CN111225354A (en) * | 2020-02-14 | 2020-06-02 | 重庆邮电大学 | CSI human body falling identification method under WiFi interference environment |
| CN112330924A (en) * | 2020-11-05 | 2021-02-05 | 北京理工大学 | Falling event detection method based on channel state information in indoor environment |
| CN112364770A (en) * | 2020-11-11 | 2021-02-12 | 天津大学 | Commercial Wi-Fi-based human activity recognition and action quality evaluation method |
| CN114358162A (en) * | 2021-12-27 | 2022-04-15 | 中国人民解放军国防科技大学 | Falling detection method and device based on continuous wavelet transform and electronic equipment |
-
2022
- 2022-04-29 CN CN202210475689.8A patent/CN115035686A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180198548A1 (en) * | 2017-01-08 | 2018-07-12 | At&T Intellectual Property I, L.P. | Interference cancelation for 5g or other next generation network |
| CN107749143A (en) * | 2017-10-30 | 2018-03-02 | 安徽工业大学 | A kind of indoor occupant fall detection system and method through walls based on WiFi signal |
| CN109635837A (en) * | 2018-11-10 | 2019-04-16 | 天津大学 | A kind of carefree fall detection system of scene based on commercial wireless Wi-Fi |
| CN109394229A (en) * | 2018-11-22 | 2019-03-01 | 九牧厨卫股份有限公司 | A kind of fall detection method, apparatus and system |
| US20200163590A1 (en) * | 2018-11-22 | 2020-05-28 | Jomoo Kitchen & Bath Co., Ltd | Fall detection method, device, and system |
| CN111225354A (en) * | 2020-02-14 | 2020-06-02 | 重庆邮电大学 | CSI human body falling identification method under WiFi interference environment |
| CN112330924A (en) * | 2020-11-05 | 2021-02-05 | 北京理工大学 | Falling event detection method based on channel state information in indoor environment |
| CN112364770A (en) * | 2020-11-11 | 2021-02-12 | 天津大学 | Commercial Wi-Fi-based human activity recognition and action quality evaluation method |
| CN114358162A (en) * | 2021-12-27 | 2022-04-15 | 中国人民解放军国防科技大学 | Falling detection method and device based on continuous wavelet transform and electronic equipment |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116058298A (en) * | 2023-03-06 | 2023-05-05 | 北京市农林科学院信息技术研究中心 | Livestock behavior monitoring method and device |
| CN116058298B (en) * | 2023-03-06 | 2023-09-12 | 北京市农林科学院信息技术研究中心 | Livestock behavior monitoring method and device |
| CN116304844A (en) * | 2023-05-23 | 2023-06-23 | 山东科技大学 | Personnel entry and exit counting and counting system and method |
| CN116304844B (en) * | 2023-05-23 | 2023-09-01 | 山东科技大学 | Personnel entry and exit counting and counting system and method |
| CN116528358A (en) * | 2023-06-30 | 2023-08-01 | 华南师范大学 | Wi-Fi-based fall positioning joint detection method, device and system |
| CN116528358B (en) * | 2023-06-30 | 2023-09-12 | 华南师范大学 | Wi-Fi-based fall positioning joint detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115035686A (en) | Real-time falling detection method, system and medium based on channel state information | |
| CN109243470B (en) | Broiler cough monitoring method based on audio technology | |
| CN109817339B (en) | Patient grouping method and device based on big data | |
| CN109620244B (en) | Infant abnormal behavior detection method based on condition generation countermeasure network and SVM | |
| CN109864736A (en) | Processing method, device, terminal device and the medium of electrocardiosignal | |
| CN109512390B (en) | Sleep staging method and wearable device based on EEG time domain multi-dimensional features and M-WSVM | |
| CN110737201B (en) | Monitoring method and device, storage medium and air conditioner | |
| CN110414607A (en) | Classification method, device, equipment and the medium of capsule endoscope image | |
| CN113723157B (en) | Crop disease identification method and device, electronic equipment and storage medium | |
| CN114781463A (en) | Cross-scene robust indoor tumble wireless detection method and related equipment | |
| CN114580572B (en) | Abnormal value identification method and device, electronic equipment and storage medium | |
| CN117787510A (en) | Optimization method of pesticide residue monitoring process based on time sequence predictive analysis | |
| CN110163103B (en) | Live pig behavior identification method and device based on video image | |
| CN116596879A (en) | Grape downy mildew self-adaptive identification method based on quantiles of boundary samples | |
| CN114465678A (en) | Complex activity WIFI perception method based on deep learning | |
| CN114488841A (en) | Data collection processing method of intelligent wearable device | |
| CN117235650B (en) | Method, device, equipment and medium for detecting high-altitude operation state | |
| CN110309767A (en) | In vivo detection equipment, recognition methods, device and storage medium | |
| CN110414562A (en) | Classification method, device, terminal and the storage medium of X-ray | |
| CN113837122A (en) | Wi-Fi channel state information-based non-contact human body behavior identification method and system | |
| US20230409422A1 (en) | Systems and Methods for Anomaly Detection in Multi-Modal Data Streams | |
| CN114692683B (en) | Fall detection method and device based on CSI and storage medium | |
| CN106326672A (en) | Falling into sleep detecting method and system | |
| CN109086207B (en) | Page response fault analysis method, computer readable storage medium and terminal device | |
| CN118298467B (en) | Intelligent recognition method and system for individual anomalies of breeding hens based on machine vision |
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 |
Application publication date: 20220909 |
|
| RJ01 | Rejection of invention patent application after publication |