CN105989694A - Human body falling-down detection method based on three-axis acceleration sensor - Google Patents
Human body falling-down detection method based on three-axis acceleration sensor Download PDFInfo
- Publication number
- CN105989694A CN105989694A CN201510074961.1A CN201510074961A CN105989694A CN 105989694 A CN105989694 A CN 105989694A CN 201510074961 A CN201510074961 A CN 201510074961A CN 105989694 A CN105989694 A CN 105989694A
- Authority
- CN
- China
- Prior art keywords
- axis acceleration
- human body
- detection method
- falling
- behavior
- 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
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a human body falling-down detection method based on a three-axis acceleration sensor. The human body falling-down detection method comprises the steps of acquiring X-axis acceleration data, Y-axis acceleration data and Z-axis acceleration data by an acceleration sensor when a user moves, and processing the acquired data; judging whether a Y-axis acceleration peak value reaches a threshold TH1 or not; when the Y-axis acceleration peak value exceeds TH1, intercepting a time period within 2s after the Y-axis acceleration peak value appears, extracting features in the time period, and carrying out matching on difference time sequences and feature sequences of a falling-down behavior in sequence by using DTW (Dynamic Time Warping) in the time period so as to recognize whether a user falls down or not. The human body falling-down detection method is convenient to apply in falling-down detection, and can effectively differentiate falling-down and daily behaviors.
Description
Technical field:
The present invention relates to pattern recognition and sensor technical field, be specifically related to a kind of tumble detection method for human body based on 3-axis acceleration sensor.
Background technology:
In current contingency, there are about 29% is directly or indirectly to be caused by falling.The lighter that falls can cause the wounded to abrade, fracture, and severe one is even
Cause death, especially to disadvantaged group such as old peoples from the point of view of, falling is grave danger of its healthy living.Medical research shows: as long as human body is falling
Fall and given treatment in time after occurring, it is possible to avoid the generation of major disease.Therefore, disadvantaged group are carried out fall detection, improves to greatest extent
Their health level, saves medical treatment expense, is then a highly important medical care problem and social problem.
At present, the method fallen of detection has two big classes: fall detection based on environmental remote sensing technology and fall detection based on Wearable sensor.(1)
Vibrating sensor or photographic head that fall detection system based on environmental remote sensing technology is fixed by placement location in specific environment obtain human body and transport
Dynamic information is entered, and obtains a range of ground vibration signal characteristic around from vibrating sensor, obtains human motion figure from one or more photographic head
As feature, and then the movable information of extraction carried out point by technology such as data mining, pattern recognition, data fusion, wavelet analysis, artificial intelligences
Analysis processes, it is judged that whether human body falls phenomenon.This type of method is only effective in specific environment, easily by environmental disturbances, and bad adaptability, equipment valency
Lattice are the highest, the most susceptible to user acceptance.(2) fall detection system based on Wearable sensor, it is common that embedding in daily living article
Enter microsensor (acceleration transducer, gyroscope, pressure transducer etc.) be placed on the head of human body, chest, waist, arm, thigh,
The signal characteristics such as the accekeration of station acquisition human body, angle value, force value such as vola, monitor physical activity, in real time at the movement parameter of human body
Judge whether there occurs by certain algorithm when having bigger change and fall, once there occurs and fall, by wireless sending module, this situation is carried out
Location and warning.Fall detection process based on Wearable sensor is not limited by environment, overcomes environmental remote sensing technology can only be limited to certain limit
Problem, motility is strong, has vast potential for future development.
Acceleration transducer, compared to other Wearable sensors such as gyroscope, pressure transducer, has following superiority: 1. gyroscope can only be measured
The directional information of each axle.And acceleration transducer can not only gather the directional information of acceleration, moreover it is possible to gather the size information of acceleration.2. pressure
Sensor is normally placed at vola, lacks motility, to be typically monitored physical activity with the work of other sensor synergisms.And acceleration transducer
Each position of the trunk that can place, discrimination is high, can be used alone fall detection, has been the main of current fall detection research
Trend.
Summary of the invention:
Technical problem: present invention solves the technical problem that to be to overcome and using limitation environmentally, suitable based on environmental remote sensing technology fall detection
The drawback that answering property is poor, be easily disturbed, overcomes gyroscope and obtains the incomplete drawback of body motion information, overcome pressure transducer very flexible
Defect, it is provided that a kind of tumble detection method for human body based on 3-axis acceleration sensor, detection method is simple, can preferable each from human body
The behavior of kind detects and falls.
Technical scheme: the invention provides a kind of tumble detection method for human body based on 3-axis acceleration sensor, solves by the following technical solutions
State technical problem:
A. template generation:
1. utilize 3-axis acceleration sensor to gather human body X, Y, Z 3-axis acceleration data sequence during falling.
2. pair above-mentioned X, Y, Z 3-axis acceleration data sequence collected carries out pretreatment, is defined on the t moment point gained original X, Y, Z
3-axis acceleration data are respectively ax(t)、ay(t)、azT (), carries out smoothing denoising pretreatment with the moving average filter that width m is 3,
If a 'x(t)、a′y(t)、a′zT () is the 3-axis acceleration after pretreatment, then shown in method such as formula (1), (2) (3):
3.Y axle acceleration peak computational formula is as shown in (4):
Ymax=max ay′(t) (4)
Human body is during falling, and the acceleration change on vertical direction is the most obvious and different from the change of activities of daily living, the most right
Y-axis acceleration is analyzed.Judge Y-axis acceleration peak value YmaxWhether more than TH1, if Ymax> TH1, then perform next step, otherwise again adopt
Collection data.
4. from Y-axis acceleration peak value YmaxOccur starting to YmaxAfter appearance, 2s is as the time period of feature extraction, has 100 numbers in this time period
The characteristic sequence at strong point.In order to correctly be distinguished with behavior of not falling by daily behavior, the present invention is extracted two features and carries out data after pretreatment
Identify, be resultant acceleration and inclination angle respectively.
1. resultant acceleration
Resultant acceleration shows the severe degree of human motion, and its value is the biggest, moves the most violent.Within 2s, body action amplitude is not after falling for human body
Can be too big, resultant acceleration is less, can fluctuate up and down at 1g.As shown in (5) in the resultant acceleration computing formula of t sequence of points:
2. inclination angle
The inclined degree of incidence meter person of good sense's body, its value is the biggest, and inclined degree is the biggest.When human body is in erectility, inclination angle is 0 °, lying status
Time inclination angle be 90 °.Human body cannot oneself be stood up within 2s after falling, and is in lying status after therefore falling within 2s, and inclination angle is at 90 °
Neighbouring fluctuation.The Dip countion formula of t sequence of points is as shown in (6):
5. finally giving the characteristic sequence of four behaviors of falling as sample form, if i & lt is fallen, the characteristic sequence of behavior is Fi, it calculates
Formula is as shown in (7):
Wherein 1≤i≤4,1≤t≤n, n=100, i, t ∈ Z.Two templates of i, j are asked for corresponding to resultant acceleration and inclination angle with DTW algorithm
Minimum Cumulative Distance between characteristic sequence is respectively D (FAi, FAi)、D(FQi, FQj), then the distance such as formula (8) between two templates of i, j
Shown in:
D (i, j)=D (FAi, FAj)+D(FQi, FQj) (8)
Ask for the distance of other two form assemblies by formula (8), have 6 kinds of combinations, according to formula (9) ask for two fall between behavior average
Distance:
DM and F the most at lastiIt is stored in template base.
B. fall detection:
1. the behavior being likely to occur during user completes daily life, gathers human body X, Y, Z during falling with 3-axis acceleration sensor
3-axis acceleration data ax(t)、ay(t)、az(t);
2. by formula (1) (2) (3), raw acceleration data is carried out pretreatment, obtain pretreated acceleration information a 'x(t)、a′y(t)、
a′z(t);
3. ask for Y-axis acceleration peak value Y in motor process according to formula (4)max, and judge formula YmaxWhether exceed threshold value TH2, according to formula (10)
Acquisition judged result:
If judged result f1It is 1, then enters next step and judge, otherwise, Resurvey data.When human body contacts ground during falling, Y-axis adds
Speed can sharply increase and reach maximum, by Y-axis acceleration peak value YmaxThe detection of size, it is known that the motion of human body in the vertical direction
The most violent.
4. extract Y-axis acceleration peak value Y according to formula (5) (6)maxOccur starting to YmaxThe characteristic sequence F of 2s time period after appearancep, i.e.
Resultant acceleration characteristic sequence and inclination angle characteristic sequence, as shown in formula (11):
5. by the characteristic sequence F of the behaviorpMate with the characteristic sequence of 4 groups of templates, fallen with i-th by formula (8) the calculating behavior
Distance D between behaviour template (p, i), calculates the behavior and the average distance fallen between behavior according to such as following formula (12):
6. fall the Status Flag f of Activity recognition2As shown in following formula (13):
Wherein DM/DC shows this group behavior and the similarity degree fallen, and its value is the least, illustrates that this group behavior is closer to the behavior of falling, right
DM/DC sets threshold value, by judging whether that reaching threshold value can identify the behavior of falling.If Status Flag f2Be 1, then the behavior is to fall
Behavior, otherwise, the behavior is non-behavior of falling.
The present invention compared with prior art, has the advantage that
1. overcome vibrating sensor, photographic head equipment cost high, range is little, by drawbacks such as environmental disturbances are big, use acceleration sensing
Device can obtain body motion information whenever and wherever possible, will not impact the daily life of user, and low cost, it is easier to is easily accepted by a user.
2. pair Y-axis acceleration peak value sets threshold value, by judging whether the Y-axis acceleration peak value of certain behavior motor process exceedes threshold value, to behavior
Carry out pre-judgement, excluded and the bigger motion of gap of falling.
3. in terms of human motion amplitude and human motion attitude two, it is extracted the feature higher to human motion discrimination, uses DTW side
Human motion state is judged by method, can reach higher discrimination, and the present invention can effectively identify the behavior of falling.
Accompanying drawing illustrates:
Fig. 1 is template generation flow chart
Fig. 2 is fall detection flow chart
Fig. 3 is human action model 3-axis acceleration directional diagram
Fig. 4 chooses figure for extracting characteristic time section
Detailed description of the invention:
The daily life motion of human body mainly includes jump, running, walking, sits down, lies down, bends over.The present invention is according to the daily life of human body
Motion of living is different from the motion feature of the behavior of falling, and detects, based on 3-axis acceleration sensor, the behavior of falling.
In order to make the purpose of the present invention, technical scheme and advantage clearer, a kind of based on 3-axis acceleration to the present invention below in conjunction with accompanying drawing
The tumble detection method for human body of sensor further describes.Should be appreciated that embodiments of the present invention are not limited to this.
As it is shown in figure 1, the matching template flow chart for setting up human body fall detection provided for the present invention, concrete steps include:
Step 101: user completes the behavior of falling, acceleration transducer collection the 3-axis acceleration data of process of falling, with level to the left as X
Axle, is Y-axis vertically upward, and horizontal forward for Z axis, X, Y, Z tri-axles are orthogonal, and three direction of principal axis are shown in Fig. 3.
Step 102: the original 3-axis acceleration data obtained carry out pretreatment, i.e. uses moving average filter to carry out smoothing denoising process,
Processing procedure is shown in technical scheme a. template generation 2.
Step 103: extract Y-axis acceleration peak value Ymax, it is judged that YmaxWhether more than TH1, the most then carry out step 104, otherwise carry out
Step 101.
Step 104: intercept Y-axis acceleration peak value and go out to start now to Y-axis acceleration peak value 2s this time period T occurs, as shown in Figure 4.
Step 105: according to the resultant acceleration characteristic sequence in a. template generation 4 extraction time section T in technical scheme and inclination angle characteristic sequence.
Step 106: judge whether behavior of the falling number of times meeting above-mentioned steps is 4, the most then carry out step 107, otherwise carry out step 101.
Step 107: set up template according to a. template generation 5 in technical scheme, this template includes the characteristic sequence of 4 behaviors of falling and twice
Fall the average distance DM between behaviour template.
As in figure 2 it is shown, for the present invention provide for user movement is carried out fall detection flow chart, concrete steps include:
Step 201: user complete daily in the crawler behavior that is likely to occur, and gather the 3-axis acceleration data of motor process.
Step 202: 3-axis acceleration data are carried out pretreatment.
Step mule 203: extract behavior Y-axis acceleration peak value Ymax, judge according to b. fall detection 3 in technical scheme, if Status Flag
Position f1It is 1, shows that the motion of human body in the vertical direction is more violent, perform step 204, otherwise perform step 201.
Step 204: from Y-axis acceleration peak value go out now T start timing 2s, obtain the pretreated 3-axis acceleration in T time section
Data.
Step 205: extract the resultant acceleration characteristic sequence of 3-axis acceleration data and inclination angle characteristic sequence in T time section.
Step 206: the characteristic sequence of the behavior is used successively with 4 groups of behaviors of falling in template base according to b. fall detection 5 in technical scheme
DTW method is mated, and tries to achieve the behavior and the average distance DC of behaviour template of falling.Obtain according to according to b. fall detection 6 in technical scheme
Take state flag bit f2If, state flag bit f2Be 1, then the match is successful, otherwise, it fails to match.
Step 207: it fails to match, the behavior is non-behavior of falling.
Step 208: the match is successful, the behavior is the behavior of falling.
Step 209: behavior of falling, reports to the police.
Claims (7)
1. a tumble detection method for human body based on 3-axis acceleration sensor, it is characterised in that the method comprises template generation and fall detection two
Part, specifically comprises the following steps that
A. template generation:
Step (1): user completes the behavior of falling, by X, Y, Z 3-axis acceleration data during acceleration transducer acquisition human motion;
Step (2): the original 3-axis acceleration data collected are carried out pretreatment;
Step (3): extract through pretreated Y-axis acceleration peak value Ymax, it is judged that YmaxWhether exceed threshold value TH1, if Ymax> TH1,
Then perform step (4);Otherwise, perform step (1);
Step (4): extract the feature of special time period, and generate matching template;
B. fall detection:
Step (5): user completes the behavior being likely to occur in daily life, obtains 3-axis acceleration data;
Step (6): the original 3-axis acceleration data obtained are carried out pretreatment;
Step (7): extract through pretreated Y-axis acceleration peak value Ymax, it is judged that YmaxWhether exceed threshold value TH1, if Ymax> TH1,
Then perform step (8), otherwise, perform step (5);
Step (8): extract the feature of special time period;
Step (9): successively different time sequence in special time period is mated with the template of generation with DTW, it is judged that whether user falls.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 1, it is characterised in that step (2)
It is that smoothing denoising processes that original 3-axis acceleration to gather described with step (6) carries out pretreatment.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 2, it is characterised in that described smoothing is gone
Make an uproar and be processed as moving average filter.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 1, it is characterised in that step (4)
Human motion range parameter, human body attitude parameter is included with the feature of step (8) described extraction.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 1, it is characterised in that generate coupling mould
When plate and fall detection, gained 3-axis acceleration data after pretreatment extract characteristic sequence.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 1, it is characterised in that step (4),
Step (8) and step (9) described special time period be after Y-axis acceleration peak value occurs starting to occur to Y-axis acceleration peak value in 2s this time
Between section.
A kind of tumble detection method for human body based on 3-axis acceleration sensor the most according to claim 1, it is characterised in that step (4),
Matching template described in step (9) includes that 4 groups of human bodies are fallen the characteristic sequence of behavior, and fall behavior and the average distance of falling between behavior.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510074961.1A CN105989694A (en) | 2015-02-05 | 2015-02-05 | Human body falling-down detection method based on three-axis acceleration sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510074961.1A CN105989694A (en) | 2015-02-05 | 2015-02-05 | Human body falling-down detection method based on three-axis acceleration sensor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105989694A true CN105989694A (en) | 2016-10-05 |
Family
ID=57041876
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510074961.1A Pending CN105989694A (en) | 2015-02-05 | 2015-02-05 | Human body falling-down detection method based on three-axis acceleration sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105989694A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106910314A (en) * | 2017-02-03 | 2017-06-30 | 同济大学 | A kind of personalized fall detection method based on the bodily form |
CN107105092A (en) * | 2017-04-18 | 2017-08-29 | 华东师范大学 | A kind of human body tumble recognition methods based on dynamic time warping |
CN107126217A (en) * | 2017-06-05 | 2017-09-05 | 河池学院 | Wearable falls down detection device |
CN107909771A (en) * | 2017-10-08 | 2018-04-13 | 南京邮电大学 | A kind of personnel's tumble alarm system and its implementation based on wireless sensor network |
CN108960016A (en) * | 2017-05-25 | 2018-12-07 | 富士通株式会社 | The method and apparatus for detecting specific action |
CN109171738A (en) * | 2018-07-13 | 2019-01-11 | 杭州电子科技大学 | Fall detection method based on human body acceleration multiple features fusion and KNN |
CN111192434A (en) * | 2020-01-19 | 2020-05-22 | 中国建筑第四工程局有限公司 | Safety protective clothing recognition system and method based on multi-mode perception |
CN111210595A (en) * | 2020-01-15 | 2020-05-29 | 广东工业大学 | Human body falling detection and warning method, device and computer readable storage medium |
CN112466090A (en) * | 2020-11-26 | 2021-03-09 | 汤泽金 | Intelligent rescue bracelet |
CN113499066A (en) * | 2021-07-13 | 2021-10-15 | 西安邮电大学 | Multi-node falling early warning method and system based on DTW gait difference |
CN113822182A (en) * | 2021-09-08 | 2021-12-21 | 河南理工大学 | Motion detection method and system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102509117A (en) * | 2011-09-30 | 2012-06-20 | 成都君晟科技有限公司 | DTW (dynamic time warping) based method for automatically distinguishing abnormal measurements from plantar pressure measurement |
CN104082905A (en) * | 2014-06-18 | 2014-10-08 | 南京纳联信息科技有限公司 | Multifunctional intelligent insole and gait similarity detection method |
US20140309964A1 (en) * | 2013-04-11 | 2014-10-16 | Microsoft Corporation | Internal Sensor Based Personalized Pedestrian Location |
CN106066995A (en) * | 2016-05-25 | 2016-11-02 | 西安交通大学 | A kind of wireless unbundling human body behavioral value algorithm |
-
2015
- 2015-02-05 CN CN201510074961.1A patent/CN105989694A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102509117A (en) * | 2011-09-30 | 2012-06-20 | 成都君晟科技有限公司 | DTW (dynamic time warping) based method for automatically distinguishing abnormal measurements from plantar pressure measurement |
US20140309964A1 (en) * | 2013-04-11 | 2014-10-16 | Microsoft Corporation | Internal Sensor Based Personalized Pedestrian Location |
CN104082905A (en) * | 2014-06-18 | 2014-10-08 | 南京纳联信息科技有限公司 | Multifunctional intelligent insole and gait similarity detection method |
CN106066995A (en) * | 2016-05-25 | 2016-11-02 | 西安交通大学 | A kind of wireless unbundling human body behavioral value algorithm |
Non-Patent Citations (2)
Title |
---|
SASIWAN PAIYAROM等: "Fall detection and activity monitoring system using dynamic time warping for elderly and disabled people", 《I-CREATE "09 PROCEEDINGS OF THE 3RD INTERNATIONAL CONVENTION ON REHABILITATION ENGINEERING & ASSISTIVE TECHNOLOGY》 * |
黄帅: "老年人跌倒检测系统中相关算法的研究及应用", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106910314A (en) * | 2017-02-03 | 2017-06-30 | 同济大学 | A kind of personalized fall detection method based on the bodily form |
CN107105092A (en) * | 2017-04-18 | 2017-08-29 | 华东师范大学 | A kind of human body tumble recognition methods based on dynamic time warping |
CN108960016A (en) * | 2017-05-25 | 2018-12-07 | 富士通株式会社 | The method and apparatus for detecting specific action |
CN107126217A (en) * | 2017-06-05 | 2017-09-05 | 河池学院 | Wearable falls down detection device |
CN107909771B (en) * | 2017-10-08 | 2020-04-10 | 南京邮电大学 | Personnel falling alarm system based on wireless sensor network and implementation method thereof |
CN107909771A (en) * | 2017-10-08 | 2018-04-13 | 南京邮电大学 | A kind of personnel's tumble alarm system and its implementation based on wireless sensor network |
CN109171738A (en) * | 2018-07-13 | 2019-01-11 | 杭州电子科技大学 | Fall detection method based on human body acceleration multiple features fusion and KNN |
CN111210595A (en) * | 2020-01-15 | 2020-05-29 | 广东工业大学 | Human body falling detection and warning method, device and computer readable storage medium |
CN111192434A (en) * | 2020-01-19 | 2020-05-22 | 中国建筑第四工程局有限公司 | Safety protective clothing recognition system and method based on multi-mode perception |
CN111192434B (en) * | 2020-01-19 | 2024-02-09 | 中国建筑第四工程局有限公司 | Multi-mode perception based safety protection suit identification system and method |
CN112466090A (en) * | 2020-11-26 | 2021-03-09 | 汤泽金 | Intelligent rescue bracelet |
CN113499066A (en) * | 2021-07-13 | 2021-10-15 | 西安邮电大学 | Multi-node falling early warning method and system based on DTW gait difference |
CN113499066B (en) * | 2021-07-13 | 2024-05-03 | 西安邮电大学 | Multi-node fall early warning method and system based on DTW gait difference |
CN113822182A (en) * | 2021-09-08 | 2021-12-21 | 河南理工大学 | Motion detection method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105989694A (en) | Human body falling-down detection method based on three-axis acceleration sensor | |
CN101558996B (en) | Gait recognition method based on orthogonal projection three-dimensional reconstruction of human motion structure | |
Tong et al. | HMM-based human fall detection and prediction method using tri-axial accelerometer | |
Dikovski et al. | Evaluation of different feature sets for gait recognition using skeletal data from Kinect | |
CN103337132B (en) | Tumble detection method for human body based on 3-axis acceleration sensor | |
Hamäläinen et al. | Jerk-based feature extraction for robust activity recognition from acceleration data | |
Wang et al. | Real time accelerometer-based gait recognition using adaptive windowed wavelet transforms | |
CN108958482B (en) | Similarity action recognition device and method based on convolutional neural network | |
CN105303183B (en) | A kind of child posture discriminance analysis system and method based on wearable device | |
KR20130073361A (en) | Apparatus and method for classifing pattern of electromyogram signals | |
Cao et al. | A fall detection method based on acceleration data and hidden Markov model | |
CN105138967A (en) | Living body detection method and apparatus based on active state of human eye region | |
Martínez-Villaseñor et al. | Deep learning for multimodal fall detection | |
CN107169334B (en) | The user authen method based on straight punch motion detection for hand wearable device | |
CN114818952A (en) | Intelligent falling posture classification and identification method based on mobile phone sensor | |
CN106971203A (en) | Personal identification method based on characteristic on foot | |
CN109634419A (en) | The recognition methods of healing robot motion intention and its computer readable storage medium | |
CN109584521A (en) | A kind of incorrect sitting-pose monitoring method based on Notch sensor | |
CN111626273B (en) | Fall behavior recognition system and method based on atomic action time sequence characteristics | |
CN103251411A (en) | Complexity based pressure center nonlinear feature extraction method | |
Zhang et al. | An efficient feature selection method for activity classification | |
Liao et al. | The application of EMD in activity recognition based on a single triaxial accelerometer | |
Hai et al. | PCA-SVM algorithm for classification of skeletal data-based eigen postures | |
Huang et al. | Comparison of feature-level and kernel-level data fusion methods in multi-sensory fall detection | |
Del Pozo et al. | Speed-independent gait identification for mobile devices |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161005 |