CN105827338B - Indoor environment content identification method based on Wi-Fi signal and mobile phone - Google Patents
Indoor environment content identification method based on Wi-Fi signal and mobile phone Download PDFInfo
- Publication number
- CN105827338B CN105827338B CN201610143111.7A CN201610143111A CN105827338B CN 105827338 B CN105827338 B CN 105827338B CN 201610143111 A CN201610143111 A CN 201610143111A CN 105827338 B CN105827338 B CN 105827338B
- Authority
- CN
- China
- Prior art keywords
- value
- signal strength
- mobile phone
- indoor environment
- identification
- 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.)
- Active
Links
Classifications
-
- 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
- H04B17/318—Received signal strength
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The present invention provides a kind of indoor environment content identification method based on Wi Fi signals and mobile phone, at 2 seconds, 2 seconds, 6 seconds or 10 seconds it was the characteristic value under time window by extracting Wi Fi intensity curves, so that indoor environment type be identified, when being particularly suitable for that 1) interior space is empty；2) when target object stands or lies down beside the mobile phone；3) target object apart from described mobile phone 1m, 2m, 3m, 4m when stand or when walking about；4) target object move towards mobile phone or far from mobile phone when indoor environment content recognition, recognition efficiency is higher, and accuracy is preferable, and required training data point quantity is also less.Reduce the cost of indoor environment identification.
Description
Technical field
The present invention relates to indoor activity identification technology fields, are specifically related to a kind of room based on Wi-Fi signal and mobile phone
Internal environment holds recognition methods.
Background technology
Determine that the content of indoor environment is of great significance, researcher develops a variety of different algorithms.With based on
Universal, the indoor ring based on Wi-Fi of the development of the radio network technique of 802.11 specifications and wireless network and smart mobile phone
Border content recognition obtains extensive concern.
Ambient Property identification includes two steps：Training and identification.It is bent that training stage mainly creates Wi-Fi signal variation
Line database, indoor different ambient Property will produce different Wi-Fi signal strength variations, by ambient Property and corresponding letter
Number combined configuration information database of Strength Changes.In cognitive phase, the change in signal strength curve that mobile phone is received and letter
All data in breath database are compared, and find similar data information, then by existing algorithm, including KNN, random
Forest, naive Bayesian, support vector machines scheduling algorithm, environment-identification content.This method passes through existing Wi-Fi hotspot and common intelligence
Energy mobile phone, which is combined, can be achieved with indoor environment identification, and algorithm complexity is low, and recognition effect is good.
Indoor environment content recognition is strongly depend on the sampling signal frequency of signal receiver-mobile phone, but high sampling rate
Equipment restricted application, user needs the high sampling rate equipment for buying costliness that relevant range could be identified, and adopts
The more high then accuracy of identification of sample rate is higher, and the equipment of high sampling rate needs higher maintenance cost and sampling cost.
Secondly, high sampling rate equipment commonly uses the frequency domain character such as energy spectrum of wireless signal, and Fourier is carried out to spectrum
The huge algorithms of calculation amounts such as analysis support equipment electricity to propose high requirement, are not suitable for carrying out indoor environment for a long time
Content recognition.
It can be seen that although the technology of the indoor environment content recognition based on wireless signal has, algorithm is simple, accuracy of identification
Higher advantage, but purchase professional equipment cost is higher, restricted application directly faces in actual application
Problem, which is the shortage of professional equipment, to be caused using being obstructed.Therefore it is made for carrying out the popularization of ambient Property identification using wireless signal
At obstacle.
Invention content
The purpose of the present invention is to provide a kind of indoor environment content identification method based on Wi-Fi signal and mobile phone, should
Invention solves the technical issues of signal characteristic value selection of indoor environment content recognition under low sampling rate.
The present invention provides a kind of indoor environment content identification method based on Wi-Fi signal and mobile phone, includes the following steps：
Step S100：Training stage：Under variant indoor environmental condition, corresponding Wi-Fi signal strength curve is acquired,
Respectively with 2 seconds, 2 seconds, 6 seconds or 10 seconds for time window, the training for calculating separately each trained Wi-Fi signal strength curve of gained is special
Gained training characteristics value is carried out corresponding storage by value indicative with corresponding indoor environment state respectively；
Step S200：Cognitive phase：Certain point in indoor environment to be identified, with mobile phone acquisition mobile phone, institute can collected knowledge
Other Wi-Fi signal strength, and identification Wi-Fi signal strength characteristic value is calculated in time window, obtain identification feature value；
Step S300：Sorting phase：Identification feature value vector is formed with identification feature value, while being formed with training characteristics value
Training feature vector classifies to recognition feature vector by classification, be classified as in training feature vector it is a kind of compared with
It is similar；
Step S400：Decision stage：According to the training characteristics value obtained in step S100 and specific indoor environment state
Corresponding storage relationship judges the indoor environment situation residing for certain point in indoor environment to be identified；
Indoor environment includes：1) when the interior space is empty；2) when target object stands or lies down beside mobile phone；3) target
When standing or walk about when object distance mobile phone 1m, 2m, 3m, 4m；4) when target object moves towards mobile phone or separate mobile phone；
Training characteristics value includes the training average value of Wi-Fi signal strength in each time window, intermediate value, maximum value, most
Small value, summation, standard deviation, difference；
Identification feature value includes that the average value of Wi-Fi signal strength, intermediate value, maximum value, most are identified in each time window
Small value, summation, standard deviation, difference.
Further, the average value of training Wi-Fi signal strength or the average value of identification Wi-Fi signal strength, by formula
(1) it is calculated：
W in formula (1)tIndicate the window of t times, siIndicate i-th of signal strength values, sjIndicate j-th of signal strength
Value.
Further, the intermediate value of the intermediate value of training Wi-Fi signal strength or identification Wi-Fi signal strength, based on formula (2)
It calculates：
Wherein j indicates another Wi-Fi signal strength point different from i, sjIndicate j-th of signal strength values.
Further, the standard deviation of the standard deviation of training Wi-Fi signal strength or identification Wi-Fi signal strength, by formula
(6) it calculates：
Preferably, in step s 200, wavelet de-noising processing is carried out to identification Wi-Fi signal strength.
The technique effect of the present invention：
Indoor environment content identification method provided by the invention based on Wi-Fi signal and mobile phone, first with smart mobile phone
It collects Wi-Fi signal and creates an indoor signal delta data library, including the signal intensity held in a large amount of various indoor environments
Initial data；It then proposes that wavelet de-noising (Wavelet Denoise) filters out the noise in initial data, extracts its time domain spy
Sign；Finally, k-NN and classification tree algorithm are proposed, is identified based on temporal signatures.Method provided by the invention is adopted merely with low
The smart mobile phone of sample rate realizes content recognition.Be conducive to that popularization cost is greatly reduced.
Indoor environment content identification method provided by the invention based on Wi-Fi signal and mobile phone, uses standard intelligence hand
Machine carries out signal sampling, improves algorithm scalability；Classified using temporal signatures, reduce calculation amount and power consumption,
Therefore can effectively extend can orientation range by the present invention.
It specifically please refers to and according to the present invention is proposed based on the indoor environment content identification method of Wi-Fi signal and mobile phone
Various embodiments it is described below, will make apparent in terms of the above and other of the present invention.
Description of the drawings
Fig. 1 is that indoor environment content identification method flow of the preferred embodiment of the present invention based on Wi-Fi signal and mobile phone is shown
It is intended to.
Specific implementation mode
The attached drawing constituted part of this application is used to provide further understanding of the present invention, schematic reality of the invention
Example and its explanation are applied for explaining the present invention, is not constituted improper limitations of the present invention.
Indoor environment identification herein is specifically only limitted to：It is nobody or someone to room state；User's body state is
It lies down or standing state；User's body apart from mobile phone be 1m, 2m, 3m or 4m when；The direction of motion of user's body is proximate to mobile phone
Or far from mobile phone.
Referring to Fig. 1, the indoor environment content identification method provided by the invention based on Wi-Fi signal and mobile phone is main to wrap
Containing training and two stages of identification.Training stage includes signal collection, signal processing, classification storage three phases；Cognitive phase
Including signal collection, signal processing, content recognition three phases.
Include the following steps：
Step S100：Training stage：Under variant indoor environmental condition, corresponding Wi-Fi signal strength curve is acquired,
Respectively with 2 seconds, 2 seconds, 6 seconds or 10 seconds for time window, the training for calculating separately each trained Wi-Fi signal strength curve of gained is special
Gained training characteristics value is carried out corresponding storage by value indicative with corresponding indoor environment state respectively；
Step S200：Cognitive phase：Certain point in indoor environment to be identified, with mobile phone acquisition mobile phone, institute can collected knowledge
Other Wi-Fi signal strength, and identification Wi-Fi signal strength characteristic value is calculated in time window, obtain identification feature value；
Step S300：Sorting phase：Identification feature value vector is formed with identification feature value, while being formed with training characteristics value
Training feature vector classifies to recognition feature vector by classification, be classified as in training feature vector it is a kind of compared with
It is similar；
Step S400：Decision stage：According to the training characteristics value obtained in step S100 and specific indoor environment state
Corresponding storage relationship judges the indoor environment situation residing for certain point in indoor environment to be identified；
Indoor environment includes：1) when the interior space is empty；2) when target object stands or lies down beside mobile phone；3) target
When standing or walk about when object distance mobile phone 1m, 2m, 3m, 4m；4) when target object moves towards mobile phone or separate mobile phone；
Training characteristics value includes the training average value of Wi-Fi signal strength in each time window, intermediate value, maximum value, most
Small value, summation, standard deviation, difference；
Identification feature value includes that the average value of Wi-Fi signal strength, intermediate value, maximum value, most are identified in each time window
Small value, summation, standard deviation, difference.
Preferably, the average value of training Wi-Fi signal strength or the average value of identification Wi-Fi signal strength, by formula (1)
It is calculated：
W in formula (1)tIndicate the window of t times, siIndicate i-th of signal strength values.
Preferably, the intermediate value of the intermediate value of training Wi-Fi signal strength or identification Wi-Fi signal strength, based on formula (2)
It calculates：
Wherein, j indicates another Wi-Fi signal strength point different from i, sjIndicate j-th of signal strength values.
Method provided by the invention is used as characteristic value by extracting intermediate value, and intermediate value represents static nature change, than only carrying
It is averaged as characteristic value, acquired results have better robustness.
Preferably, the standard deviation of the standard deviation of training Wi-Fi signal strength or identification Wi-Fi signal strength, by formula (6)
It calculates：
The present invention has reacted the received letter of target object by extracting standard deviation of the signal strength under different time window
Number volatility and fluctuation, can estimate that training Wi-Fi signal near mobile phone changes, as whether target object has occurred movement
Deng.It is higher particularly with the recognition efficiency of this 4 class indoor environment.
Specifically include following steps：
Training stage
Step 1, signal collection.Under the conditions of different indoor environments identify, the Wi- that is acquired and recorded by mobile phone
Fi signal strengths change over time curve.Indoor environment classification specifically includes：1) when the interior space is empty；2) target object (
Can be described as user) it stands or when lying down beside mobile phone；3) target object apart from mobile phone 1m, 2m, 3m, 4m when stand or walk about
When；4) when target object moves towards mobile phone or separate mobile phone.Indoor environments hereinafter referred to as different for this 4 class indoor environment.
Step 2, signal processing.Under different indoor environments, respectively with 2 seconds, 2 seconds, 6 seconds or 10 seconds for time window
(time window) counts the totally 13 different Wi-Fi signal strength curves obtained in step 1 by formula (1) respectively
Calculate the average value of each training characteristics value：
To under different indoor environments, the corresponding trained Wi-Fi signal strength of each time window takes it flat by formula (1)
Mean value：
Wherein w in formula (1)tIndicate the window of t times, siIndicate i-th of letter on training Wi-Fi signal strength curve
Number intensity value；The static nature that the method provided by the present invention receives different indoor environments using average value representative signal strength changes
Become, is conducive to improve the differentiation accuracy to target object Position Approximate.
To under different indoor environments, the corresponding trained Wi-Fi signal strength of each time window is taken wherein by formula (2)
Value：
Method provided by the invention is used as characteristic value by extracting intermediate value, and intermediate value represents static nature change, than only carrying
It is averaged as characteristic value, acquired results have better robustness.J herein is a series of hot spots for opposite i
In another hot spot, asBottom right mark refer to bottom right mark according to arranging from small to large, S values are also according to from small
To longer spread, that is,
To under different indoor environments, the corresponding trained Wi-Fi signal strength of each time window takes it most by formula (3)
Big value：
To under different indoor environments, the corresponding trained Wi-Fi signal strength of each time window takes it most by formula (4)
Small value：
By formula (5) under different indoor environments, calculating the total of all trained Wi-Fi signal strengths in each time window
With：
Maximum value represents the maximum value of signal strength in a window, and minimum value represents in a window signal strength most
Small value, summation represent the characteristic value summation in a window, and summation, maximum value, minimum value three combination can be to indoor environments
Preferable differentiation effect is played in the variation of state；It is particularly suitable for this 4 kinds of situations proposed in the processing present invention.
Train the standard of Wi-Fi signal strength in each time window under different indoor environments, calculating by formula (6)
Difference：
According to the standard deviation that formula (6) is calculated, the volatility and fluctuation of the received signal of target object, energy are represented
Estimate that the training Wi-Fi signal near mobile phone changes, such as whether target object is moved.Particularly in this 4 class room
The recognition efficiency of environment is higher.
Train the difference of Wi-Fi signal strength in each time window under different indoor environments, calculating by formula (7)：
Diff=Max (wt)-Min(wt) (7)
The difference of training Wi-Fi signal strength represents in a time window, maximum signal and minimum signal strength
Difference, can be used as an index of trained Wi-Fi signal variation degree.
Step 3, classification storage.Step 2 has carried out the extraction of training characteristics value, step to the data being collected into step 1
These training characteristics are worth corresponding indoor environment state and carry out corresponding storage by three, and subsequent cognitive phase is facilitated to carry out
Classification and Identification.
Cognitive phase：
Step 1, signal collection.Certain point in indoor environment to be identified, frequency of the mobile phone to give tacit consent to, acquisition one per second
Secondary mobile phone can be obtained identification Wi-Fi signal strength；
Step 2, signal processing.Wavelet de-noising processing is carried out to collected identification Wi-Fi signal strength, is filtered out
Then abnormal point therein calculates the identification Wi-Fi signal strength identification in each time window in 4 kinds of time windows above-mentioned
Characteristic value；Abnormal point herein may be wifi jitters under the conditions of, caused by data exception point, such as a certain second do not have
The wifi signals of specific SSID are scanned, then the intensity value of software records is the 0 of acquiescence at this time, this exception information can cause
Variance feature value becomes larger suddenly, and then influences normal characteristic value identification process.Characteristic value herein refer to it is each when
Between the average value of identification Wi-Fi signal strength, intermediate value, maximum value, minimum value, summation, standard deviation, difference in window.
Step 3, Classification and Identification.Identification feature value vector is formed with the identification feature value of each time window of gained, according to instruction
It is foundation to practice the indoor environment situation of stage storage and the corresponding storage result of training characteristics value, is carried out to gained identification feature value
Classification, the classification may be used the machine learning methods such as K arest neighbors and classify.It is corresponding according to the characteristic value of classification results
Indoor environment situation judges the ambient condition of certain point in indoor environment to be identified, to realize the identification of ambient Property.
Below with as follows in the specific example stood or the indoor environment of state of couching is identified to target object：
Step S100：When record target object stands and lies down beside mobile phone, it is strong that each second acquires a Wi-Fi signal
Angle value, n seconds time obtain n data point.Set of data points when standing is recorded as R1(r1,r2,…,rn), data when lying down
Point set is recorded as R2(l1,l2,…,lm)
Step S200：Calculate R1Seven characteristic values, obtain the average value of first time windowIntermediate valueMaximum value max1=max (r1,r2), minimum value min1=min (r1,r2), summation s1=r1+r2, standard deviationDifference d1=| max1-min1|.With vector (m1,a1,max1,min1,s1,v1,d1) represent
One time window, therefore R1It is decomposed into n/2 vector, similarly R2Also it is decomposed into m/2 vector.Include the set of n data
R1.After the characteristic value of step S200 calculates, n/2 7 degree of freedom vector, i.e. { (m have been turned to1,a1,max1,min1,s1,v1,
d1),…,(mn/2,an/2,maxn/2,minn/2,sn/2,vn/2,dn/2), similarly R2Also m/2 7 degree of freedom vector has been turned to.
Step S300：When target object stands or lies down beside mobile phone, mobile phone record Wi-Fi signal strength at this time with
A data point on time changing curve is (s1,s2), calculate separately 7 characteristic values of gained intensity curves at this time：It is average
ValueIntermediate valueMaximum value max=max (s1,s2), minimum value min=min (s1,s2), summation s
=s1+s2, standard deviationDifference d=| max-min |.Gained vector (m, a, max, min, s, v,
D) state vector point at this time is represented.
Step S400：Using K nearest neighbor classifications, choose K=20, judge in the septuple space vector point (m, a, max,
Min, s, v, d) distance R1Representative vectorial point set or R2Representative vectorial point set who closer to, to judge, mesh at this time
Mark the indoor environment classification residing for object.
Thus it may recognize that the state of target object is in standing or to lie down.
Those skilled in the art will be clear that the scope of the present invention is not limited to example discussed above, it is possible to be carried out to it
Several changes and modification, the scope of the present invention limited without departing from the appended claims.Although oneself is through in attached drawing and explanation
The present invention is illustrated and described in book in detail, but such illustrate and describe only is explanation or schematical, and not restrictive.
The present invention is not limited to the disclosed embodiments.
By to attached drawing, the research of specification and claims, those skilled in the art can be in carrying out the present invention
Understand and realize the deformation of the disclosed embodiments.In detail in the claims, term " comprising " is not excluded for other steps or element,
And indefinite article "one" or "an" be not excluded for it is multiple.The certain measures quoted in mutually different dependent claims
The fact does not mean that the combination of these measures cannot be advantageously used.Any reference marker in claims is not constituted pair
The limitation of the scope of the present invention.
Claims (1)
1. a kind of indoor environment content identification method based on Wi-Fi signal and mobile phone, which is characterized in that include the following steps：
Step S100：Training stage：Under variant indoor environmental condition, corresponding Wi-Fi signal strength curve is acquired, respectively
With 2 seconds, 2 seconds, 6 seconds or 10 seconds for time window, the training characteristics of each trained Wi-Fi signal strength curve of gained are calculated separately
Value, corresponding storage is carried out by the gained training characteristics value with corresponding indoor environment state respectively；
Step S200：Cognitive phase：Certain point in indoor environment to be identified, acquiring the mobile phone institute with mobile phone can collected knowledge
Other Wi-Fi signal strength, and identification Wi-Fi signal strength characteristic value is calculated in the time window, obtain identification feature value；
Step S300：Sorting phase：Identification feature value vector is formed with the identification feature value, while with the training characteristics value
Composition training feature vector classified to the recognition feature vector by classification, be classified as with the training characteristics to
One kind in amount is relatively similar；
Step S400：Decision stage：According to the training characteristics value obtained in the step S100 and specific indoor environment shape
The correspondence storage relationship of state judges the indoor environment situation residing for certain point in the indoor environment to be identified；
The indoor environment includes：1) when the interior space is empty；2) when target object stands or lies down beside the mobile phone；3)
When standing or walk about when target object is apart from described mobile phone 1m, 2m, 3m, 4m；4) when target object moves towards mobile phone or separate mobile phone；
The training characteristics value includes the average value that Wi-Fi signal strength is trained described in each time window, intermediate value, maximum
Value, minimum value, summation, standard deviation, difference；
The identification feature value includes the average value that Wi-Fi signal strength is identified described in each time window, intermediate value, maximum
Value, minimum value, summation, standard deviation, difference；
The average value of the average value of the trained Wi-Fi signal strength or the identification Wi-Fi signal strength, based on formula (1)
It obtains：
W in formula (1)tIndicate the window of t times, siIndicate i-th of signal strength values；
The intermediate value of the intermediate value of the trained Wi-Fi signal strength or the identification Wi-Fi signal strength, is calculated by formula (2)：
Wherein j indicates another Wi-Fi signal strength point different from i, sjIndicate j-th of signal strength values；
The standard deviation of the standard deviation of the trained Wi-Fi signal strength or the identification Wi-Fi signal strength, based on formula (6)
It calculates：
In the step S200, wavelet de-noising processing is carried out to the identification Wi-Fi signal strength.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610143111.7A CN105827338B (en) | 2016-03-14 | 2016-03-14 | Indoor environment content identification method based on Wi-Fi signal and mobile phone |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610143111.7A CN105827338B (en) | 2016-03-14 | 2016-03-14 | Indoor environment content identification method based on Wi-Fi signal and mobile phone |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105827338A CN105827338A (en) | 2016-08-03 |
| CN105827338B true CN105827338B (en) | 2018-08-07 |
Family
ID=56987864
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610143111.7A Active CN105827338B (en) | 2016-03-14 | 2016-03-14 | Indoor environment content identification method based on Wi-Fi signal and mobile phone |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105827338B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108008350A (en) * | 2017-11-03 | 2018-05-08 | 平安科技（深圳）有限公司 | Localization method, device and storage medium based on Random Forest model |
| CN110493581A (en) * | 2019-08-13 | 2019-11-22 | 苏州兆见智能科技有限公司 | A kind of control method for supporting more people's immersion experience |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102724751A (en) * | 2012-06-04 | 2012-10-10 | 清华大学 | Wireless indoor positioning method based on off-site survey |
| CN102821194A (en) * | 2012-07-17 | 2012-12-12 | 西安电子科技大学 | Cellphone indoor positioning device and cellphone indoor positioning method on basis of various sensors |
| CN104798420A (en) * | 2012-12-24 | 2015-07-22 | 英特尔公司 | System and method for pilot sequence design in a communications system |
-
2016
- 2016-03-14 CN CN201610143111.7A patent/CN105827338B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102724751A (en) * | 2012-06-04 | 2012-10-10 | 清华大学 | Wireless indoor positioning method based on off-site survey |
| CN102821194A (en) * | 2012-07-17 | 2012-12-12 | 西安电子科技大学 | Cellphone indoor positioning device and cellphone indoor positioning method on basis of various sensors |
| CN104798420A (en) * | 2012-12-24 | 2015-07-22 | 英特尔公司 | System and method for pilot sequence design in a communications system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105827338A (en) | 2016-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Song et al. | A novel convolutional neural network based indoor localization framework with WiFi fingerprinting | |
| CN106407905B (en) | Wireless aware action identification method based on machine learning | |
| Carpineti et al. | Custom dual transportation mode detection by smartphone devices exploiting sensor diversity | |
| CN105678222B (en) | A kind of mobile device-based Human bodys' response method | |
| Guo et al. | HuAc: Human activity recognition using crowdsourced WiFi signals and skeleton data | |
| Bhattacharya et al. | Using unlabeled data in a sparse-coding framework for human activity recognition | |
| CN101695445B (en) | Acceleration transducer-based gait identification method | |
| Kjærgaard et al. | Mobile sensing of pedestrian flocks in indoor environments using wifi signals | |
| CN112036433B (en) | CNN-based Wi-Move behavior sensing method | |
| CN106685590B (en) | Indoor human body orientation identification method based on channel state information and KNN | |
| CN107844990A (en) | A kind of approaches to IM and its system, terminal device for intelligent shops | |
| CN109215010B (en) | Image quality judgment method and robot face recognition system | |
| Sun et al. | Multifloor Wi-Fi localization system with floor identification | |
| CN105827338B (en) | Indoor environment content identification method based on Wi-Fi signal and mobile phone | |
| Ding et al. | Device-free location-independent human activity recognition using transfer learning based on CNN | |
| CN109241870B (en) | Coal mine underground personnel identity identification method based on gait identification | |
| CN111901028B (en) | Human body behavior identification method based on CSI (channel State information) on multiple antennas | |
| CN117437599B (en) | Pedestrian abnormal event detection method and system for monitoring scene | |
| CN107230219A (en) | A kind of target person in monocular robot is found and follower method | |
| CN110730473A (en) | Signal feature extraction method for WiFi activity recognition | |
| Balazia et al. | Learning robust features for gait recognition by maximum margin criterion | |
| Liu et al. | A sensor-free crowd counting framework for indoor environments based on channel state information | |
| CN106971203B (en) | Identity recognition method based on walking characteristic data | |
| CN104679967A (en) | Method for judging reliability of psychological test | |
| He et al. | Crowd-flow graph construction and identification with spatio-temporal signal feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |