CN110456320A - A kind of ULTRA-WIDEBAND RADAR personal identification method based on free space gait temporal aspect - Google Patents
A kind of ULTRA-WIDEBAND RADAR personal identification method based on free space gait temporal aspect Download PDFInfo
- Publication number
- CN110456320A CN110456320A CN201910688715.3A CN201910688715A CN110456320A CN 110456320 A CN110456320 A CN 110456320A CN 201910688715 A CN201910688715 A CN 201910688715A CN 110456320 A CN110456320 A CN 110456320A
- Authority
- CN
- China
- Prior art keywords
- gait
- ultra
- doppler
- wideband radar
- distance
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/415—Identification of targets based on measurements of movement associated with the target
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of ULTRA-WIDEBAND RADAR personal identification methods based on free space gait temporal aspect, comprising the following steps: the gait signal of fixed duration is collected using ultra-wideband radar sensors；Gait signal is split to obtain a series of subsignals along the slow time, Fast Fourier Transform (FFT) is carried out to each subsignal and obtains distance-Doppler figure, a gait signal can obtain one group of distance-Doppler figure；Utilization orientation histogram of gradients carries out HOG feature extraction to each distance-Doppler figure；One group of HOG feature for belonging to the same gait signal is input in identification model according to slow time sequencing, Gait Recognition is carried out, obtains the identification result of target object.The ULTRA-WIDEBAND RADAR personal identification method can both rely on the identification that gait feature realizes target object not by the display of the direction of travel of target and angle.
Description
Technical field
The invention belongs to identify field, and in particular to a kind of ULTRA-WIDEBAND RADAR body based on free space gait temporal aspect
Part recognition methods.
Background technique
Environment sensing is the core function in the present or even following monitoring system, and scheme universal at present is by camera
To collect information from ambient enviroment.However, camera is to illumination condition, there are higher requirements, while being easy leakage privacy etc.
Disadvantage causes it not use sensitizing range within the family.Although other such as wearable device robustness are higher, by
It can be repelled by most people in invasive reason, therefore equally exist biggish limitation.With semiconductor technology
The volume of fast-developing chip significantly reduces, and present radar starts to play increasing effect in civil field.Radar
Sensor is typically mounted in ambient enviroment, by microwave as information carrier, therefore invasive and direct privacy is not present
Leakage problem.In addition, it is not influenced by factors such as weather, illumination, it is suitble to for a long time perceive indoor environment.
By long-term development, the effect of radar has been not limited only to simple ranging, angle measurement etc. at present, can also be right
Objective attribute target attribute in movement is characterized accordingly.Specifically, the target monitored during exercise can be to the frequency of radar signal
Rate and phase are modulated, to generate additional frequency component around the basic frequency of echo-signal, here it is micro-dopplers
Feature.In practical applications, can be used to classify to gait by micro-Doppler feature, to detect in intrusion family
The disease of stranger or even some similar parkinson's syndromes of auxiliary diagnosis (walking posture changes after illness).Forefathers grind
The target studied carefully possesses biggish biology discrimination mostly, such as distinguishes young man and old man, or identification double-arm sway, freedom
Arm swing and without three kinds of walking motions of arm swing.Importantly, existing method requires target in radar line of sight direction row mostly
It walks, this will bring inconvenience in practical application.
Summary of the invention
The present invention provides a kind of ULTRA-WIDEBAND RADAR personal identification methods based on free space gait temporal aspect, this is super
Wideband radar personal identification method can both rely on gait feature and realize target not by the display of the direction of travel of target and angle
The identification of object.
The technical solution of the present invention is as follows:
A kind of ULTRA-WIDEBAND RADAR personal identification method based on free space gait temporal aspect, comprising the following steps:
The gait signal of fixed duration is collected using ultra-wideband radar sensors；
Gait signal is split to obtain a series of subsignals along the slow time, each subsignal is carried out in quick Fu
Leaf transformation obtains distance-Doppler figure, and a gait signal can obtain one group of distance-Doppler figure；
Utilization orientation histogram of gradients carries out HOG feature extraction to each distance-Doppler figure；
One group of HOG feature for belonging to the same gait signal is input in identification model according to slow time sequencing,
Gait Recognition is carried out, the identification result of target object is obtained.
Preferably, the utilization orientation histogram of gradients includes: to the progress HOG feature extraction of each distance-Doppler figure
Adjust the distance-Dopplergram carry out gray processing processing；
Using the contrast of Gamma correction method adjustment gray processing treated distance-Doppler figure；
The gradient information of each pixel on distance-Doppler figure after calculating setting contrast, using horizontal gradient and
Vertical gradient just calculates amplitude and the direction of gradient, i.e. acquisition HOG feature.
Preferably, the ULTRA-WIDEBAND RADAR personal identification method further include:
Entire distance-Doppler figure is divided into the cell factory of the sizes such as several, counts the gradient in cell factory
Histogram；Multiple cell factories are combined into module unit, the histogram of gradients in module unit is connected and normalized, shape
At the HOG descriptor of current block；All descriptors of entire distance-Doppler figure mutually splice the HOG that just can be finally extracted
Feature.
In the present invention, the identification model is the network formed with length time memory network and classifier, through instructing
It gets, wherein length time memory network is used to carry out feature integration to the HOG feature of input, and classifier is used for integration
Feature carry out classification judgement, to export recognition result.
Wherein, the classifier includes softmax classifier.
Subsignal indicates the characteristic in the gait short time, therefore the length of subsignal is critically important, through Experimental Research, subsignal
When a length of 150ms~250ms, the subsignal of such length just can accurately indicate the characteristic of gait.
Compared with prior art, the device have the advantages that are as follows:
The ULTRA-WIDEBAND RADAR personal identification method, both can be quick and precisely not by the display of the direction of travel of target and angle
The identification of ground realization target object.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to do simply to introduce, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art, can be with root under the premise of not making the creative labor
Other accompanying drawings are obtained according to these attached drawings.
Fig. 1 is the stream of the ULTRA-WIDEBAND RADAR personal identification method proposed by the present invention based on free space gait temporal aspect
Cheng Tu；
Fig. 2 is the Gait Recognition system structure diagram for realizing ULTRA-WIDEBAND RADAR personal identification method；
Fig. 3 is that the segmentation of slow time of gait signal generates schematic diagram with RDM；
Fig. 4 is typical one group of RDM；
Fig. 5 is the accuracy rate and loss figure of identification model training process.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention more comprehensible, with reference to the accompanying drawings and embodiments to this
Invention is described in further detail.It should be appreciated that the specific embodiments described herein are only used to explain the present invention,
And the scope of protection of the present invention is not limited.
In order to identify the identity information of people, the present invention provides a kind of ultra wide bands based on free space gait temporal aspect
Radar personal identification method and the Gait Recognition system for realizing the ULTRA-WIDEBAND RADAR personal identification method.Above-mentioned Gait Recognition system
The application scenarios of the ULTRA-WIDEBAND RADAR personal identification method for sum of uniting may include at family's door pass --- it is identified using gait feature
Stranger's swarms into, at indoor spaciousness --- judge that the elderly's gait changes with corresponding disease of early warning etc..
As shown in Fig. 2, the ULTRA-WIDEBAND RADAR personal identification method the following steps are included:
S101 collects the gait signal of fixed duration using ultra-wideband radar sensors；
Gait signal was split to obtain a series of subsignals along the slow time by S102, carried out to each subsignal fast
Fast Fourier transformation obtains distance-Doppler figure, and a gait signal can obtain one group of distance-Doppler figure；
S103, utilization orientation histogram of gradients carry out feature extraction to each distance-Doppler figure；
One group of feature for belonging to the same gait signal is input to identification model according to slow time sequencing by S104
In, Gait Recognition is carried out, the identification result of target object is obtained.
The ULTRA-WIDEBAND RADAR personal identification method does not have to limit the direction of travel and angle of target, and can be to its row
It walks the ULTRA-WIDEBAND RADAR echo-signal that movement generates and carries out feature extraction and time series modeling, finally distinguish mesh by gait feature
Target identity.It is described in detail below for every part.
The present invention collects radar echo signal, as gait signal, traditional time-frequency figure using ultra-wideband radar sensors
(time-Doppler maps, TDM) is lost range information, in order to completely utilize the information contained in radar echo signal
And conclusion and separating capacity of the representative domain to feature are improved, the present invention constructs new data coding method, process such as Fig. 3 institute
Show.
Gait signal is carried out cutting along slow time-domain and obtains subsignal, these subsignal durations are in 200ms or so, table
The characteristic in the gait short time is shown.Fourier transformation is applied in the distance unit of each subsignal and RDM is calculated, and belongs to
Time, Doppler frequency and space length information in the characterized radar signal of multiple RDM of the same gait signal.
One group of RDM that typical gait signal generates is as shown in figure 4, every figure illustrates frequency in the period of current subsignal record
Distribution situation in each distance unit.
The walking motion of human body is a dynamic process, and one group of RDM shown in Fig. 4 has abandoned traditional " static state " expression side
Formula (relies on single TDM), the situation then feature for describing gait of " dynamic " changes with time, and this data representative domain is
It introduces Time-Series analysis and provides theories integration.
Compared to traditional image, these RDM are more succinct intuitive in shape and texture, can preferably by gradient or
Described by person's edge direction density.The present invention using a kind of this method in computer vision field extensive utilization of HOG as
Character description method carries out feature extraction to RDM.HOG is by the pixel situation of change in statistical picture regional area as table
The foundation of shape and texture is levied, and lower compared to other methods computation complexity.
After carrying out gray processing processing to RDM, HOG method adjusts the contrast of RDM by Gamma correction method, to subtract
The influence that remote reflection signal energy reduces on RDM caused by light propagation path is decayed:
P (a, b)=P (a, b)gamma
In formula, P (a, b) is the pixel value at coordinate (a, b)；
Then, the gradient information of each pixel on the RDM after calculating setting contrast:
Gx(a, b)=P (a+1, b)-P (a-1, b)
Gy(a, b)=P (a, b+1)-P (a, b-1)
Amplitude and the direction of gradient can be calculated using this horizontal gradient and vertical gradient；
Finally, entire RDM to be divided into the cell factory (Cell) of the sizes such as several, the gradient in cell factory is counted
Histogram；Multiple cell factories are further combined into bigger module unit (Block), to the histogram of gradients in module unit
It is connected and is normalized, form the HOG descriptor of current block；All descriptors of entire RDM, which mutually splice, just can obtain finally
The HOG feature of extraction.
There is sequential relationship, Ke Yiyong between each other for the feature that one group of RDM that one gait signal generates is extracted
Some time series modeling techniques handle feature.Therefore, the net formed with length time memory network and classifier is selected
Network is trained the identification network as identification network, obtains identification model.
A kind of special case of the LSTM as Recognition with Recurrent Neural Network (Recruited Neural Networks, RNN), using thin
Born of the same parents' state (Cell State) and hidden state (Hidden State) transmit between unit, stores information come to long interval information into
Row memory.Information flow situation in unit between unit is as shown in the following formula:
ft=σ (Wxfxf+Whfht-1+bf)
it=σ (Wxixt+Whiht-1+bi)
rt=tanh (Wxrxt+Whrht-1+br)
Ct=ft*Ct-1+it*rt
ot=σ (Wxoxt+Whoht-1+bo)
ht=ot*tanh(Ct)
The one group of feature obtained by histograms of oriented gradients is inputted in LSTM, LSTM is final by the update of time step
Probability of all categories is calculated by Softmax layers of result feeding to obtain the identity information of target: Softmax function such as following formula
It is shown:
Embodiment
For sharpening method flow of the invention, more intuitively show coming as a result, enclosing embodiment herein to it for method
Further remarked additionally.Altogether there are four experimental subjects in the present embodiment, their corresponding body physical feature such as 1 institutes of table
Show:
1 experimental subjects physical features of table
| Identity | Height | Weight |
| Target 1 | 170cm | 55kg |
| Target 2 | 179cm | 80kg |
| Target 3 | 181cm | 75kg |
| Target 4 | 173cm | 54kg |
Each experimental subjects is walked (do not limit direction and angle) in free space, is acquired using ULTRA-WIDEBAND RADAR
534 gait sample of signal: each sample record time is 2 seconds, includes 500 radar frames, then along slow time shaft to step
State signal is split, and segmentation length is 100 points, is forward lapped 50 points, therefore each sample generates 9 subsignals and is used to generate
RDM is simultaneously adjusted to fixed size (being 64*32 in the present embodiment).Four experimental subjects totally 2136 samples according to 7:3 ratio
It is divided into training set and test set, wherein training set is used for cross validation acquisition optimized parameter model, and test set is used to test
Practical manifestation of the model under truthful data.Under 5 folding cross validations, initial data is divided into 5 groups, wherein 4 when each iteration
Group is taken as training set, and remaining one group is taken as verifying to collect, the evaluation criteria by the average result of 5 iteration as model.
The cross validation assessment accuracy rate of model is as shown in table 2:
2 model parameter assessment result of table
The perception domain of cell size and image is closely bound up, due to smaller, the excessive cell of the size of RDM in this embodiment
Size can reduce the capturing ability to image detail, lead to the decline of recognition performance；Meanwhile the excessive increasing of LSTM network complexity
Add and also will affect final accuracy rate.It is 79.67% that maximum assessment accuracy rate, which can be obtained, by table 2.By the 5 of cross validation Shi Suofen
A part is merged into final training set, and is re-started and trained and to the step of test set on training set with optimized parameter model
State data are tested.The accuracy rate of training process and loss are as shown in Figure 5:
Reach 79.10% test accuracy rate when exercise wheel number is 100, however as can be seen from Fig. 5 when exercise wheel number
When more than 60, the generalization ability of model tends to decline, therefore the number of iterations of suitable control training is needed in practical application, sacrificial
Domestic animal a part of training set ability to express retains enough generalization abilities as far as possible.Table 3 is given a in the form of confusion matrix
Accurate rate (Precision), false dismissed rate (FNR) and the false alarm rate (FPR) of target identities identification.It can be seen that this method is at target 2
Performance is good, has shown ideal effect；The accurate rate of visual target 3 is smaller simultaneously, false dismissed rate is larger, and reason is
Its feature walked does not protrude, it is easy to be classified as in other classifications；Since the walk of target 1 and target 4 changes greatly, hold
Easily cover the gait characteristics of other targets, therefore be easy to mislead the identification of other targets, leads to biggish false alarm rate.Total comes
It says, the ULTRA-WIDEBAND RADAR personal identification method performance provided by the invention based on free space gait temporal aspect is good, there is one
Fixed application potential.
3 test set confusion matrix of table
| Really/prediction | Target 1 | Target 2 | Target 3 | Target 4 | FNR |
| Target 1 | 87.1% | 2.9% | 4.7% | 5.3% | 12.9% |
| Target 2 | 5.1% | 87.1% | 5.2% | 2.6% | 12.9% |
| Target 3 | 7.6% | 6.9% | 66.5% | 19.0% | 33.5% |
| Target 4 | 14.0% | 2.5% | 8.3% | 75.2% | 24.8% |
| FPR | 26.7% | 12.3% | 18.2% | 26.9% | - |
The present embodiment additionally provides Gait Recognition system, as shown in Figure 1, including ultra-wideband radar sensors, embedded place
Manage device and cloud database.Ultra-wideband radar sensors find out extraneous gait signal, are transmitted to embeded processor, embedded processing
Device carries out signal segmentation and feature extraction to the gait signal of acquisition, is divided using feature of the identification model to extraction
Class, export recognition result, the recognition result of acquisition be transmitted to cloud database carry out save preservation recognition result be sent to use
The terminal at family is informed.
The above Gait Recognition system and ULTRA-WIDEBAND RADAR personal identification method completely remain time in radar signal,
The information of space length and Doppler frequency, and do not need the walking angle and route of limited target.It realizes to personal part
Identification.
Technical solution of the present invention and beneficial effect is described in detail in above-described specific embodiment, Ying Li
Solution is not intended to restrict the invention the foregoing is merely presently most preferred embodiment of the invention, all in principle model of the invention
Interior done any modification, supplementary, and equivalent replacement etc. are enclosed, should all be included in the protection scope of the present invention.
Claims (6)
1. a kind of ULTRA-WIDEBAND RADAR personal identification method based on free space gait temporal aspect, comprising the following steps:
The gait signal of fixed duration is collected using ultra-wideband radar sensors；
Gait signal is split to obtain a series of subsignals along the slow time, fast Fourier change is carried out to each subsignal
Get distance-Doppler figure in return, a gait signal can obtain one group of distance-Doppler figure；
Utilization orientation histogram of gradients carries out HOG feature extraction to each distance-Doppler figure；
One group of HOG feature for belonging to the same gait signal is input in identification model according to slow time sequencing, is carried out
Gait Recognition obtains the identification result of target object.
2. the ULTRA-WIDEBAND RADAR personal identification method as described in claim 1 based on free space gait temporal aspect, special
Sign is that the utilization orientation histogram of gradients carries out HOG feature extraction to each distance-Doppler figure and includes:
Adjust the distance-Dopplergram carry out gray processing processing；
Using the contrast of Gamma correction method adjustment gray processing treated distance-Doppler figure；
The gradient information of each pixel on distance-Doppler figure after calculating setting contrast, using horizontal gradient and vertically
Gradient just calculates amplitude and the direction of gradient, i.e. acquisition HOG feature.
3. the ULTRA-WIDEBAND RADAR personal identification method as claimed in claim 2 based on free space gait temporal aspect, special
Sign is, the ULTRA-WIDEBAND RADAR personal identification method further include:
Entire distance-Doppler figure is divided into the cell factory of the sizes such as several, counts the gradient histogram in cell factory
Figure；Multiple cell factories are combined into module unit, the histogram of gradients in module unit is connected and normalized, formation is worked as
Preceding piece of HOG descriptor；All descriptors of entire distance-Doppler figure mutually splice the HOG feature that just can be finally extracted.
4. the ULTRA-WIDEBAND RADAR personal identification method as described in claim 1 based on free space gait temporal aspect, special
Sign is that the identification model is the network formed with length time memory network and classifier, trained to obtain,
In, length time memory network is used to carry out the HOG feature of input feature integration, and classifier is used to carry out the feature of integration
Classification judgement, to export recognition result.
5. the ULTRA-WIDEBAND RADAR personal identification method as claimed in claim 4 based on free space gait temporal aspect, special
Sign is that the classifier includes softmax classifier.
6. the ULTRA-WIDEBAND RADAR identification as claimed in any one of claims 1 to 5 based on free space gait temporal aspect
Method, which is characterized in that when a length of 150ms~250ms of subsignal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910688715.3A CN110456320B (en) | 2019-07-29 | 2019-07-29 | Ultra-wideband radar identity recognition method based on free space gait time sequence characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910688715.3A CN110456320B (en) | 2019-07-29 | 2019-07-29 | Ultra-wideband radar identity recognition method based on free space gait time sequence characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110456320A true CN110456320A (en) | 2019-11-15 |
| CN110456320B CN110456320B (en) | 2021-08-03 |
Family
ID=68483760
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910688715.3A Active CN110456320B (en) | 2019-07-29 | 2019-07-29 | Ultra-wideband radar identity recognition method based on free space gait time sequence characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110456320B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021068470A1 (en) * | 2020-04-09 | 2021-04-15 | 浙江大学 | Radar signal-based identity and gesture recognition method |
| CN112885431A (en) * | 2021-01-13 | 2021-06-01 | 佛山市顺德区美的洗涤电器制造有限公司 | Diet recommendation method and device, range hood, processor and storage medium |
| CN113096791A (en) * | 2020-01-08 | 2021-07-09 | 富士通株式会社 | Personnel identification device, method and system based on wireless signals |
| CN113189555A (en) * | 2021-04-12 | 2021-07-30 | 浙江大学 | Multi-target segmentation method based on time-distance bounding box target interception, gait recognition method, device, system and storage medium |
| CN113435283A (en) * | 2021-06-18 | 2021-09-24 | 浙江大学 | Ultra-wideband radar identity recognition method based on breath sample space |
| WO2021203364A1 (en) * | 2020-04-09 | 2021-10-14 | 浙江大学 | Radar signal processing method for identity recognition |
| WO2022079170A3 (en) * | 2020-10-15 | 2022-06-16 | Security Alarms & Co. S.A. | Devices and system |
| CN115015867A (en) * | 2022-05-31 | 2022-09-06 | 中国人民解放军总医院 | Ultra-wideband radar-based identity recognition and fall detection method |
Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101630364A (en) * | 2009-08-20 | 2010-01-20 | 天津大学 | Method for gait information processing and identity identification based on fusion feature |
| CN102236093A (en) * | 2010-05-01 | 2011-11-09 | 天津工程师范学院 | Device for measuring walking gait characteristic of human body |
| CN102298145A (en) * | 2011-08-15 | 2011-12-28 | 天津职业技术师范大学 | Pseudo-random code measuring device with capacity of simultaneously extracting walking features of multiple pedestrians |
| CN102765365A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision |
| CN102928835A (en) * | 2012-10-09 | 2013-02-13 | 北京航空航天大学 | Human body target motion state identifying method based on improved generalized S conversion |
| EP2589979A1 (en) * | 2011-11-03 | 2013-05-08 | Thales Nederland B.V. | System for characterizing motion of an individual, notably a human individual |
| CN103473539A (en) * | 2013-09-23 | 2013-12-25 | 智慧城市***服务（中国）有限公司 | Gait recognition method and device |
| CN104237865A (en) * | 2014-10-07 | 2014-12-24 | 电子科技大学 | Method for analyzing time and frequencies of micro-movement signals of human objects on basis of terahertz radar echoes |
| CN104794449A (en) * | 2015-04-27 | 2015-07-22 | 青岛科技大学 | Gait energy image acquisition method based on human body HOG (histogram of oriented gradient) features and identity identification method |
| CN106443626A (en) * | 2016-09-23 | 2017-02-22 | 西安思丹德信息技术有限公司 | Unmanned area target detection method |
| CN107358250A (en) * | 2017-06-07 | 2017-11-17 | 清华大学 | Body gait recognition methods and system based on the fusion of two waveband radar micro-doppler |
| CN107423729A (en) * | 2017-09-20 | 2017-12-01 | 湖南师范大学 | A kind of remote class brain three-dimensional gait identifying system and implementation method towards under complicated visual scene |
| KR101872017B1 (en) * | 2017-11-07 | 2018-06-27 | 엘아이지넥스원 주식회사 | Target Identifying Method Based on Extracted Scattering Point in Millimeter Wave Seeker and Recording Medium Storing Computer Program thereof |
| CN108256488A (en) * | 2018-01-19 | 2018-07-06 | 中国人民解放军陆军装甲兵学院 | A kind of radar target identification method based on micro-Doppler feature extraction and deep learning |
| CN108828548A (en) * | 2018-06-26 | 2018-11-16 | 重庆邮电大学 | A kind of three Parameter fusion data set construction methods based on fmcw radar |
| CN108919247A (en) * | 2018-03-30 | 2018-11-30 | 中国科学院电子学研究所 | A kind of multiple target physical examination survey and localization method based on constant false alarm rate detection |
| US20180348343A1 (en) * | 2017-06-05 | 2018-12-06 | Metawave Corporation | Intelligent metamaterial radar for target identification |
| CN109343046A (en) * | 2018-09-19 | 2019-02-15 | 成都理工大学 | Radar gait recognition method based on multifrequency multiple domain deep learning |
| CN109765539A (en) * | 2019-01-28 | 2019-05-17 | 珠海格力电器股份有限公司 | Indoor user behavior monitoring method and device, electrical equipment and home monitoring system |
| KR101988182B1 (en) * | 2018-11-02 | 2019-06-11 | 세종대학교산학협력단 | Apparatus for hand motion recognition using continuous wave radar and method thereof |
-
2019
- 2019-07-29 CN CN201910688715.3A patent/CN110456320B/en active Active
Patent Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101630364A (en) * | 2009-08-20 | 2010-01-20 | 天津大学 | Method for gait information processing and identity identification based on fusion feature |
| CN102236093A (en) * | 2010-05-01 | 2011-11-09 | 天津工程师范学院 | Device for measuring walking gait characteristic of human body |
| CN102765365A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision |
| CN102298145A (en) * | 2011-08-15 | 2011-12-28 | 天津职业技术师范大学 | Pseudo-random code measuring device with capacity of simultaneously extracting walking features of multiple pedestrians |
| EP2589979A1 (en) * | 2011-11-03 | 2013-05-08 | Thales Nederland B.V. | System for characterizing motion of an individual, notably a human individual |
| CN102928835A (en) * | 2012-10-09 | 2013-02-13 | 北京航空航天大学 | Human body target motion state identifying method based on improved generalized S conversion |
| CN103473539A (en) * | 2013-09-23 | 2013-12-25 | 智慧城市***服务（中国）有限公司 | Gait recognition method and device |
| CN104237865A (en) * | 2014-10-07 | 2014-12-24 | 电子科技大学 | Method for analyzing time and frequencies of micro-movement signals of human objects on basis of terahertz radar echoes |
| CN104794449A (en) * | 2015-04-27 | 2015-07-22 | 青岛科技大学 | Gait energy image acquisition method based on human body HOG (histogram of oriented gradient) features and identity identification method |
| CN106443626A (en) * | 2016-09-23 | 2017-02-22 | 西安思丹德信息技术有限公司 | Unmanned area target detection method |
| US20180348343A1 (en) * | 2017-06-05 | 2018-12-06 | Metawave Corporation | Intelligent metamaterial radar for target identification |
| CN107358250A (en) * | 2017-06-07 | 2017-11-17 | 清华大学 | Body gait recognition methods and system based on the fusion of two waveband radar micro-doppler |
| CN107423729A (en) * | 2017-09-20 | 2017-12-01 | 湖南师范大学 | A kind of remote class brain three-dimensional gait identifying system and implementation method towards under complicated visual scene |
| KR101872017B1 (en) * | 2017-11-07 | 2018-06-27 | 엘아이지넥스원 주식회사 | Target Identifying Method Based on Extracted Scattering Point in Millimeter Wave Seeker and Recording Medium Storing Computer Program thereof |
| CN108256488A (en) * | 2018-01-19 | 2018-07-06 | 中国人民解放军陆军装甲兵学院 | A kind of radar target identification method based on micro-Doppler feature extraction and deep learning |
| CN108919247A (en) * | 2018-03-30 | 2018-11-30 | 中国科学院电子学研究所 | A kind of multiple target physical examination survey and localization method based on constant false alarm rate detection |
| CN108828548A (en) * | 2018-06-26 | 2018-11-16 | 重庆邮电大学 | A kind of three Parameter fusion data set construction methods based on fmcw radar |
| CN109343046A (en) * | 2018-09-19 | 2019-02-15 | 成都理工大学 | Radar gait recognition method based on multifrequency multiple domain deep learning |
| KR101988182B1 (en) * | 2018-11-02 | 2019-06-11 | 세종대학교산학협력단 | Apparatus for hand motion recognition using continuous wave radar and method thereof |
| CN109765539A (en) * | 2019-01-28 | 2019-05-17 | 珠海格力电器股份有限公司 | Indoor user behavior monitoring method and device, electrical equipment and home monitoring system |
Non-Patent Citations (7)
| Title |
|---|
| JAMES PARK; JOEL T. JOHNSON; GRAEME SMITH: "《Multi-frequency Range-Doppler signatures of human gaits》", 《2013 IEEE ANTENNAS AND PROPAGATION SOCIETY INTERNATIONAL SYMPOSIUM (APSURSI)》 * |
| O. RYAN FOGLE; BRIAN D. RIGLING: "《Micro-Range/Micro-Doppler Decomposition of Human Radar Signatures》", 《IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS》 * |
| 任磊,周金海,吴祥飞,金韬,金晓峰: "《 移动设备佩戴位置自适应识别的跌倒检测方法》", 《计算机工程与应用》 * |
| 刘骊、付晓东: "《计算机服装建模与仿真》", 31 December 2018, 云南大学出版社 * |
| 武勇: "《外辐射源雷达信号处理与目标跟踪方法研究》", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 裴明涛、赵猛: "《视频时间的分析与理解》", 31 March 2019, 北京理工大学出版社 * |
| 邵禹铭: "《基于深度学习方法的人体微动特征识别》", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113096791B (en) * | 2020-01-08 | 2024-06-04 | 富士通株式会社 | Personnel identification device, method and system based on wireless signals |
| CN113096791A (en) * | 2020-01-08 | 2021-07-09 | 富士通株式会社 | Personnel identification device, method and system based on wireless signals |
| WO2021068470A1 (en) * | 2020-04-09 | 2021-04-15 | 浙江大学 | Radar signal-based identity and gesture recognition method |
| US11947002B2 (en) | 2020-04-09 | 2024-04-02 | Zhejiang University | Method for recognizing identity and gesture based on radar signals |
| WO2021203364A1 (en) * | 2020-04-09 | 2021-10-14 | 浙江大学 | Radar signal processing method for identity recognition |
| EP4235610A3 (en) * | 2020-10-15 | 2024-01-24 | Algorized Sàrl | Devices and system |
| WO2022079170A3 (en) * | 2020-10-15 | 2022-06-16 | Security Alarms & Co. S.A. | Devices and system |
| CN112885431B (en) * | 2021-01-13 | 2023-09-05 | 佛山市顺德区美的洗涤电器制造有限公司 | Diet recommendation method and device, range hood, processor and storage medium |
| CN112885431A (en) * | 2021-01-13 | 2021-06-01 | 佛山市顺德区美的洗涤电器制造有限公司 | Diet recommendation method and device, range hood, processor and storage medium |
| CN113189555B (en) * | 2021-04-12 | 2023-02-14 | 浙江大学 | Multi-target segmentation method based on time-distance bounding box target interception, gait recognition method, device, system and storage medium |
| CN113189555A (en) * | 2021-04-12 | 2021-07-30 | 浙江大学 | Multi-target segmentation method based on time-distance bounding box target interception, gait recognition method, device, system and storage medium |
| CN113435283A (en) * | 2021-06-18 | 2021-09-24 | 浙江大学 | Ultra-wideband radar identity recognition method based on breath sample space |
| CN115015867A (en) * | 2022-05-31 | 2022-09-06 | 中国人民解放军总医院 | Ultra-wideband radar-based identity recognition and fall detection method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110456320B (en) | 2021-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110456320A (en) | A kind of ULTRA-WIDEBAND RADAR personal identification method based on free space gait temporal aspect | |
| Chouhan et al. | Soft computing approaches for image segmentation: a survey | |
| CN107169435B (en) | Convolutional neural network human body action classification method based on radar simulation image | |
| Luo et al. | Temporal convolutional networks for multiperson activity recognition using a 2-d lidar | |
| Planinc et al. | Introducing the use of depth data for fall detection | |
| CN109272017B (en) | Vibration signal mode identification method and system of distributed optical fiber sensor | |
| CN104586398A (en) | Old man falling detecting method and system based on multi-sensor fusion | |
| CN108932479A (en) | A kind of human body anomaly detection method | |
| CN114942434B (en) | Fall gesture recognition method and system based on millimeter wave Lei Dadian cloud | |
| Luo et al. | Binarized neural network for edge intelligence of sensor-based human activity recognition | |
| CN107679516B (en) | Lower extremity movement recognition methods based on multiple dimensioned Gauss-Markov random field models | |
| CN108492298A (en) | Based on the multispectral image change detecting method for generating confrontation network | |
| CN104637242A (en) | Elder falling detection method and system based on multiple classifier integration | |
| US20220026557A1 (en) | Spatial sensor system with background scene subtraction | |
| CN112711979A (en) | Non-contact vital sign monitoring under slow random motion based on biological radar | |
| CN115205891A (en) | Personnel behavior recognition model training method, behavior recognition method and device | |
| CN115422962A (en) | Gesture and gesture recognition method and device based on millimeter wave radar and deep learning algorithm | |
| CN116148801A (en) | Millimeter wave radar-based target detection method and system | |
| CN113341392B (en) | Human behavior classification method based on multi-station radar micro-Doppler motion direction finding | |
| CN111403021B (en) | Monitoring method and device | |
| Lin et al. | Wi-Fi based non-invasive detection of indoor wandering using LSTM model | |
| Li et al. | Gait Recognition Using Spatio‐Temporal Information of 3D Point Cloud via Millimeter Wave Radar | |
| CN116092193A (en) | Pedestrian track reckoning method based on human motion state identification | |
| CN114511873B (en) | Static gesture recognition method and device based on millimeter wave radar imaging | |
| CN113221709B (en) | Method and device for identifying user motion and water heater |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |