CN108966131A - Fusion floor estimating method based on indoor positioning - Google Patents
Fusion floor estimating method based on indoor positioning Download PDFInfo
- Publication number
- CN108966131A CN108966131A CN201810812602.5A CN201810812602A CN108966131A CN 108966131 A CN108966131 A CN 108966131A CN 201810812602 A CN201810812602 A CN 201810812602A CN 108966131 A CN108966131 A CN 108966131A
- Authority
- CN
- China
- Prior art keywords
- floor
- cluster
- signal strength
- initial
- generated
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/021—Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The present invention provides the fusion floor estimating method based on indoor positioning, comprising steps of acquiring the signal strength of each floor signal access point in building；If collected signal strength is divided into Ganlei, if generating cluster set by Ganlei, the initial cluster center of cluster set is extracted, clustering processing is carried out to collected signal strength and initial cluster center using clustering algorithm, generates cluster result；Cluster result is mapped as corresponding true floor by set floor mapping table.The present invention is not necessarily to add additional hardware indoors, signal strength is acquired by mobile terminal, use hierarchical clustering algorithm for K-means algorithm picks initial cluster center, reduce the influence of noise spot and isolated point to cluster, revaluation using Baum-Welch algorithm to Hidden Markov Model parameter, the mapping relations between true floor and cluster set are established, realize the deduction of building floor.
Description
Technical field
The present invention relates to indoor positioning technologies fields, more particularly to the fusion floor estimating method based on indoor positioning.
Background technique
With the continuous propulsion that social cityization is built, ultra-large type store, airport, high-speed rail station, subway hinge, meeting venue
Sharply increase with the room area of the densely populated places such as conference and exhibition center, determines that the location information of itself is in complicated strange environment
One cumbersome process often will cause the dual waste of time and energy.However, room area is the blind of satellite navigation and positioning
Area, global position system GPS (Global Position System) and Beidou BDS (BeiDou Navigation
Satellite System) satellite navigation and location system can not work normally, for this problem, lot of domestic and foreign scientific research machine
Structure also expansion research in succession, proposes the indoor orientation method based on technologies such as WiFi, bluetooth, RFID, UWB, infrared rays, wherein
WiFi is wireless signal most commonly seen in city, and inexpensive popularization and application are a big advantages of WiFi indoor positioning.
Currently, numerous indoor positioning technologies based on WiFi are studied from the angle of two-dimensional surface, and uses and be based on
The matched method of location fingerprint can obtain higher positioning accuracy, but multilayered structure is the mainstream construction of current heavy construction,
The accurate deduction of floor is the emphasis that indoor positioning technologies further develop, therefore needs a kind of fusion building based on indoor positioning
Layer estimating method.
Summary of the invention
For overcome the deficiencies in the prior art, the purpose of the present invention is to provide the fusion floor deductions based on indoor positioning
Method maps corresponding floor by indoor radio signal positioning, the deduction suitable for three-dimensional floor.
The present invention provides the fusion floor estimating method based on indoor positioning, comprising the following steps:
Signal strength acquisition, acquires the signal strength of each floor signal access point in building；
Signal clustering processing generates cluster set by several classes if collected signal strength is divided into Ganlei,
The initial cluster center for extracting the cluster set, using clustering algorithm in collected signal strength and the initial clustering
The heart carries out clustering processing, generates cluster result；
Floor is inferred, the cluster result is mapped as corresponding true floor by set floor mapping table.
Further, in signal strength acquisition, by the reception of wireless signals module built in mobile terminal according to setting
The signal strength of all access points of each floor in fixed sampling interval acquisition building.
Further, in the signal clustering processing, the initial cluster center is extracted using hierarchical clustering algorithm, including
Following steps:
Sample data input, inputs using collected signal strength as sample data, and by each institute of input
It states sample data and is divided into a class；
Between class distance is calculated, sample set is generated by the sample data, initial clustering set is generated by the class,
It calculates separately the distance between all kinds of in the initial clustering set；
Initial cluster center is generated, different classes is merged, and calculates the class center after merging, is generated described initial
Cluster centre.
Further, in the signal clustering processing, using K-means clustering algorithm to collected signal strength and institute
It states initial cluster center and carries out clustering processing, comprising the following steps:
Cluster sample is chosen, several samples are randomly selected in the initial cluster center as cluster centre；
Euclidean distance is calculated, the Europe of remaining sample and each cluster centre in the initial cluster center is calculated separately
The Euclidean distance is divided into several distance sets by formula distance；
Cluster centre is calculated, the cluster centre of several distance sets is calculated；
Cluster result is generated, judges whether K-means clustering algorithm restrains, is, go to step and calculate Euclidean distance,
Otherwise cluster result is exported.
Further, the set floor mapping table is generated by Hidden Markov Model, comprising the following steps:
Model parameter is generated, by the corresponding signal strength of each floor, several cluster set, in the movement of different floor gaps
Transition probability, model under time of day to any observation state shift probability, original state when model be in observation state
Probability separately constitute true floor state set, observation state set, state-transition matrix, emission probability matrix, model at the beginning of
Beginning probability distribution over states；
Hidden Markov Model is generated, square is shifted by the true floor state set, observation state set, state
Battle array, emission probability matrix, model primitive probability distribution generate Hidden Markov Model；
Mapping table is generated, the mapping table between true floor and cluster set is generated by the Hidden Markov Model.
Further, the generation mapping table further includes estimating the model parameter using Baum-Welch algorithm, is generated
Parameter Estimation structure obtains the mapping table between true floor and cluster set by the parameter Estimation structure.
Compared with prior art, the beneficial effects of the present invention are:
The present invention provides the fusion floor estimating method based on indoor positioning, comprising the following steps: signal strength acquisition is adopted
Collect the signal strength of each floor signal access point in building；Collected signal strength is divided into several by signal clustering processing
Class extracts the initial cluster center of cluster set, using clustering algorithm to collected letter if generating cluster set by Ganlei
Number intensity and initial cluster center carry out clustering processing, generate cluster result；Floor is inferred, will be gathered by gathering floor mapping table
Class result is mapped as corresponding true floor.The present invention is not necessarily to add additional hardware indoors, without carrying out net to region to be measured
Lattice divide, and without calculating the height of each floor in advance, acquire signal strength by indoor moving terminal, acquisition path is free, only
Wireless signal is wanted to cover floor area, while without being acquired in the fixed point of each floor, without record collection point
Location information, realize that three-dimensional floor is inferred, since the selection of initial cluster center can seriously affect the calculating of K-means algorithm
Amount and operation result use hierarchical clustering algorithm in advance for K-means calculation to reduce the influence of noise spot and isolated point to cluster
Method choose initial cluster center, it is advanced optimized, improve noise it is larger when classification and regression problem there are the feelings intended
Condition realizes the revaluation to model parameter using Baum-Welch algorithm, simplifies set sequence by constructing Hidden Markov Model
Column establish the mapping relations between true floor and cluster set, realize the deduction of building floor, be not related to building for coordinate system
It is vertical, the location information of record collection point and reference mode AP is not needed yet, operand is greatly reduced, reduces human and material resources
It expends.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And can be implemented in accordance with the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention and the accompanying drawings.
A specific embodiment of the invention is shown in detail by following embodiment and its attached drawing.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the fusion floor estimating method flow chart of the invention based on indoor positioning.
Specific embodiment
In the following, being described further in conjunction with attached drawing and specific embodiment to the present invention, it should be noted that not
Under the premise of conflicting, new implementation can be formed between various embodiments described below or between each technical characteristic in any combination
Example.
Fusion floor estimating method based on indoor positioning, as shown in Figure 1, comprising the following steps:
Signal strength acquisition, acquires the signal strength of each floor signal access point in building；Preferably, signal strength is adopted
It concentrates, each floor in building is acquired according to the sampling interval of setting by the reception of wireless signals module built in mobile terminal
The signal strength of all access points.In the present embodiment, the reception of wireless signals module built in mobile terminal is specially WiFi signal
Receiving module.It is acquired different from general location fingerprint, building is acquired by the WiFi signal receiving module built in mobile terminal
The signal strength of all access points of each floor in object, acquisition path is free, as long as wireless signal covers floor area, nothing
It need to be acquired in the fixed point of each floor, also not need the location information of record collection point.
Collected signal strength is divided into several by signal clustering processing according to " aggregation-discrete " characteristic of WiFi signal
Class, so that the sample in a class both is from same floor, and the number of floor levels of building is it is known that therefore final clusters number
It can be confirmed, if generating cluster set by Ganlei, the initial cluster center of cluster set be extracted, using clustering algorithm to adopting
The signal strength and initial cluster center collected carries out clustering processing, generates cluster result；
In one embodiment, the selection of initial cluster center can seriously affect the calculation amount and operation knot of K-means algorithm
Fruit, it is preferred that in signal clustering processing, use hierarchical clustering algorithm for K-means algorithm picks initial cluster center, reduction is made an uproar
The influence of sound point and isolated point to cluster, comprising the following steps:
Sample data input, inputs using collected signal strength as sample data, and by each sample of input
Notebook data is divided into a class Si=ri；Wherein, riFor sample data, SiFor class.
Between class distance is calculated, sample set R is generated by sample datas=(r1, r2..., ri), it is generated by class initial
Cluster set S=(S1, S2..., Si), calculate separately the distance between all kinds of d in initial clustering setmin=(Si, Sj)；Its
In, RsFor sample set, S is initial clustering set, dminIt is the distance between all kinds of.
Initial cluster center is generated, by different class SiAnd SjIt merges, and calculates the class center C after mergingi, generate
Initial cluster center.It is subsequent K- by realizing the extraction of initial cluster center apart from the selection of percentage to clustering object
Means algorithm provides basis.
In one embodiment, it is preferred that in signal clustering processing, using K-means clustering algorithm to collected signal
Intensity and initial cluster center carry out clustering processing, comprising the following steps:
Cluster sample is chosen, several samples are randomly selected in initial cluster center as cluster centre；The present embodiment
In, K sample is randomly selected in initial cluster center as cluster centre, the cluster centre C=(c of K sample composition1,
c2..., ck) indicate, ci∈Rs, use Ci, i=1,2 ..., K indicate corresponding class.
Euclidean distance is calculated, the Euclidean distance d of remaining sample and K cluster centre in initial cluster center is calculated separately
(ri, cj), Euclidean distance is divided into K distance set；d(ri, cj)=min { d (ri, cj), j=1,2 ..., K }, ri∈cj。
Cluster centre is calculated, the cluster centre of K distance set is calculated,NjTable
Show the sample number in such.
Cluster result is generated, judges whether K-means clustering algorithm restrains, is, go to step and calculate Euclidean distance,
Until cluster centre no longer changes, cluster result is otherwise exported.
In one embodiment, it is the mapping relations for finding true floor and cluster set, passes through the observation sequence after clustered
Column construct Hidden Markov Model with the mobility in personal walking process.Collected signal strength passes through clustering processing quilt
It is referred in corresponding set, obtains true floor information by mapping.For a random process { ξn: n >=0 }, ξnIt indicates
State of the random process in moment n, following state ξn+1Only depend on current state ξn, with past state ξn-1...
ξ0Independently of each other, then Markov random process is formed:The value of expression state, PijTurn for state
Change probability.
When carrying out parameter Estimation by Hidden Markov Model, the initial value of parameter can seriously affect finally obtained parameter
As a result.Therefore, each parameter expression in Hidden Markov Model five-tuple (S, O, A, B, ∏) need to be determined, it is preferred that set floor
Mapping table is generated by Hidden Markov Model, comprising the following steps:
Model parameter is generated, by the corresponding signal strength of each floor, several cluster set, in the movement of different floor gaps
Transition probability, model under time of day to any observation state shift probability, original state when model be in observation state
Probability separately constitute true floor state set, observation state set, state-transition matrix, emission probability matrix, model at the beginning of
Beginning probability distribution over states；In the present embodiment, true floor state set S={ S1, S2, S3..., SK, respectively indicate K building
Layer, floor1, floor2, floor3..., floorK；Observation state set O={ O1, O2, O3..., OK, respectively indicate K
Cluster set, is indicated by the observation sequence that observation state generates with sequence of sets: Seqi={ O1, O2, O3, O4... }；State turns
Move the transition probability that matrix A indicates mobile in different floor gaps, A={ aij, aijIt indicates from t moment to t+1 moment model state
By SiIt is transferred to SjProbability, A={ aij}=P (qt+1=Sj|qt=Si), 1≤i, j≤K, wherein aij> 0,Collection
The simplified processing of sequence is closed, floor is 0 from transition probability, it is as follows to obtain state-transition matrix:
Emission probability matrix B is indicated in t moment model in time of day SjThe lower probability shifted to any observation state, B
={ bj(Ot)=P (Ot|qt=Si), really 1≤j≤K belongs to the probability of set j in the collected signal strength of floor i, i.e.,
Mapping relations between floor and cluster set.Wherein, the line number of matrix indicates that number of floor levels, matrix column number indicate cluster set
It closes, matrix is expressed as follows:
The distribution of ∏ representative model initial state probabilities, ∏={ π1, π2, π3..., πk, it indicates when original state at model
In the probability of certain state.
Hidden Markov Model is generated, true floor state set, observation state set, state-transition matrix, hair are passed through
Penetrate probability matrix, model primitive probability distribution generates Hidden Markov Model.
Mapping table is generated, the mapping table between true floor and cluster set is generated by Hidden Markov Model.
In one embodiment, it is preferred that generating mapping table further includes that model parameter is estimated using Baum-Welch algorithm, raw
At parameter Estimation structure, for observation sequence O={ O1, O2, O3..., OK, suitable model parameter is chosen, utilization is recursive
Method seeks the local maximum of probability P (O | λ), then acquires required model parameter.The maximum likelihood estimation of λ indicates
Are as follows:
Log-likelihood function is expressed as follows:
L (λ)=ln P (O | λ)
In conjunction with forward variable αtAnd backward variable β (i),t(i), it can obtain:
Introducing auxiliary function Q indicates likelihood functionExpectation, expression formula is as follows:
Wherein, Ωt(j, i) is that t moment state is j, and t+1 moment state is the probability of i, expression formula are as follows:
The revaluation formula of Baum-Welch algorithm is obtained, when moment t is 1, system is in the probability of i:
Wherein,Indicate the expectation of model from the number that state i transfer is other states,Indicate mould
The expectation that type is state j from state i transfer.
Obtain new model parameterCalculate probabilityIfThen
Illustrate that the parameter that revaluation obtains is more excellent than original parameter, repeats above-mentioned calculating, the parameter of object module, parameter can be obtained
The structure of estimation is the mapping relations of true floor and cluster set.
Floor is inferred, cluster result is mapped as corresponding true floor by set floor mapping table.
The present invention provides the fusion floor estimating method based on indoor positioning, comprising the following steps: signal strength acquisition is adopted
Collect the signal strength of each floor signal access point in building；Collected signal strength is divided into several by signal clustering processing
Class extracts the initial cluster center of cluster set, using clustering algorithm to collected letter if generating cluster set by Ganlei
Number intensity and initial cluster center carry out clustering processing, generate cluster result；Floor is inferred, will be gathered by gathering floor mapping table
Class result is mapped as corresponding true floor.The present invention is not necessarily to add additional hardware indoors, without carrying out net to region to be measured
Lattice divide, and without calculating the height of each floor in advance, acquire signal strength by indoor moving terminal, acquisition path is free, only
Wireless signal is wanted to cover floor area, while without being acquired in the fixed point of each floor, without record collection point
Location information, realize that three-dimensional floor is inferred, since the selection of initial cluster center can seriously affect the calculating of K-means algorithm
Amount and operation result use hierarchical clustering algorithm in advance for K-means calculation to reduce the influence of noise spot and isolated point to cluster
Method choose initial cluster center, it is advanced optimized, improve noise it is larger when classification and regression problem there are the feelings intended
Condition realizes the revaluation to model parameter using Baum-Welch algorithm, simplifies set sequence by constructing Hidden Markov Model
Column establish the mapping relations between true floor and cluster set, realize the deduction of building floor, be not related to building for coordinate system
It is vertical, the location information of record collection point and reference mode AP is not needed yet, operand is greatly reduced, reduces human and material resources
It expends.
More than, only presently preferred embodiments of the present invention is not intended to limit the present invention in any form；All current rows
The those of ordinary skill of industry can be shown in by specification attached drawing and above and swimmingly implement the present invention；But all to be familiar with sheet special
The technical staff of industry without departing from the scope of the present invention, is made a little using disclosed above technology contents
The equivalent variations of variation, modification and evolution is equivalent embodiment of the invention；Meanwhile all substantial technologicals according to the present invention
The variation, modification and evolution etc. of any equivalent variations to the above embodiments, still fall within technical solution of the present invention
Within protection scope.
Claims (6)
1. the fusion floor estimating method based on indoor positioning, it is characterised in that the following steps are included:
Signal strength acquisition, acquires the signal strength of each floor signal access point in building；
Signal clustering processing generates cluster set by several classes, extracts if collected signal strength is divided into Ganlei
The initial cluster center of the cluster set, using clustering algorithm to collected signal strength and the initial cluster center into
Row clustering processing generates cluster result；
Floor is inferred, the cluster result is mapped as corresponding true floor by set floor mapping table.
2. the fusion floor estimating method based on indoor positioning as described in claim 1, it is characterised in that: the signal strength
In acquisition, each floor in building is acquired according to the sampling interval of setting by the reception of wireless signals module built in mobile terminal
All access points signal strength.
3. the fusion floor estimating method based on indoor positioning as claimed in claim 2, which is characterized in that the signal cluster
In processing, the initial cluster center is extracted using hierarchical clustering algorithm, comprising the following steps:
Sample data input, inputs using collected signal strength as sample data, and by each of input sample
Notebook data is divided into a class；
Between class distance is calculated, sample set is generated by the sample data, initial clustering set is generated by the class, respectively
It calculates the distance between all kinds of in the initial clustering set；
Initial cluster center is generated, different classes is merged, and calculates the class center after merging, generates the initial clustering
Center.
4. the fusion floor estimating method based on indoor positioning as claimed in claim 3, which is characterized in that the signal cluster
In processing, clustering processing, packet are carried out to collected signal strength and the initial cluster center using K-means clustering algorithm
Include following steps:
Cluster sample is chosen, several samples are randomly selected in the initial cluster center as cluster centre；
Calculate Euclidean distance, calculate separately in the initial cluster center remaining sample and each cluster centre it is European away from
From the Euclidean distance is divided into several distance sets；
Cluster centre is calculated, the cluster centre of several distance sets is calculated；
Cluster result is generated, judges whether K-means clustering algorithm restrains, is, go to step and calculate Euclidean distance, otherwise
Export cluster result.
5. the fusion floor estimating method based on indoor positioning as claimed in claim 4, which is characterized in that the set floor
Mapping table is generated by Hidden Markov Model, comprising the following steps:
Model parameter is generated, by the corresponding signal strength of each floor, several cluster set, in the mobile transfer of different floor gaps
Probability, model under time of day to any observation state shift probability, original state when model be in the general of observation state
Rate separately constitutes true floor state set, observation state set, state-transition matrix, emission probability matrix, model initial shape
State probability distribution；
Hidden Markov Model is generated, the true floor state set, observation state set, state-transition matrix, hair are passed through
Penetrate probability matrix, model primitive probability distribution generates Hidden Markov Model；
Mapping table is generated, the mapping table between true floor and cluster set is generated by the Hidden Markov Model.
6. the fusion floor estimating method based on indoor positioning as claimed in claim 5, it is characterised in that: the generation mapping
Table further includes estimating the model parameter using Baum-Welch algorithm, generates parameter Estimation structure, passes through the parameter Estimation
Structure obtains the mapping table between true floor and cluster set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810812602.5A CN108966131A (en) | 2018-07-23 | 2018-07-23 | Fusion floor estimating method based on indoor positioning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810812602.5A CN108966131A (en) | 2018-07-23 | 2018-07-23 | Fusion floor estimating method based on indoor positioning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108966131A true CN108966131A (en) | 2018-12-07 |
Family
ID=64463438
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810812602.5A Pending CN108966131A (en) | 2018-07-23 | 2018-07-23 | Fusion floor estimating method based on indoor positioning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108966131A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110413655A (en) * | 2019-06-21 | 2019-11-05 | 中国矿业大学 | A kind of floor recognition methods based on improvement Hidden Markov Model |
| CN110824421A (en) * | 2019-11-15 | 2020-02-21 | 广东博智林机器人有限公司 | Position information processing method and device, storage medium and electronic equipment |
| CN110824461A (en) * | 2019-11-18 | 2020-02-21 | 广东博智林机器人有限公司 | Positioning method |
| CN112229411A (en) * | 2020-10-15 | 2021-01-15 | 广州小鹏自动驾驶科技有限公司 | Data processing method and device |
| CN113099386A (en) * | 2021-03-25 | 2021-07-09 | 湘潭大学 | Multi-floor indoor position identification method and application thereof in museum navigation |
| CN115022961A (en) * | 2021-12-31 | 2022-09-06 | 荣耀终端有限公司 | Positioning method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104699606A (en) * | 2015-03-06 | 2015-06-10 | 国网四川省电力公司电力科学研究院 | Method for predicting state of software system based on hidden Markov model |
| CN105554704A (en) * | 2015-12-21 | 2016-05-04 | 西安电子科技大学 | Fake-locus-based location privacy protection method for use in recommendation system |
| CN106888426A (en) * | 2017-03-03 | 2017-06-23 | 深圳墨兔科技有限公司 | The indoor orientation method and system of a kind of bluetooth, WiFi and inertial navigation technique fusion |
| CN107896362A (en) * | 2017-10-25 | 2018-04-10 | 电子科技大学 | A kind of WIFI location fingerprints localization method and system based on deep learning |
-
2018
- 2018-07-23 CN CN201810812602.5A patent/CN108966131A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104699606A (en) * | 2015-03-06 | 2015-06-10 | 国网四川省电力公司电力科学研究院 | Method for predicting state of software system based on hidden Markov model |
| CN105554704A (en) * | 2015-12-21 | 2016-05-04 | 西安电子科技大学 | Fake-locus-based location privacy protection method for use in recommendation system |
| CN106888426A (en) * | 2017-03-03 | 2017-06-23 | 深圳墨兔科技有限公司 | The indoor orientation method and system of a kind of bluetooth, WiFi and inertial navigation technique fusion |
| CN107896362A (en) * | 2017-10-25 | 2018-04-10 | 电子科技大学 | A kind of WIFI location fingerprints localization method and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 高洪晔: "基于HMM的室内楼层定位研究", 《中国优秀硕士学位论文全文数据库（信息科技辑）》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110413655A (en) * | 2019-06-21 | 2019-11-05 | 中国矿业大学 | A kind of floor recognition methods based on improvement Hidden Markov Model |
| CN110413655B (en) * | 2019-06-21 | 2023-05-26 | 中国矿业大学 | Floor identification method based on improved hidden Markov model |
| CN110824421A (en) * | 2019-11-15 | 2020-02-21 | 广东博智林机器人有限公司 | Position information processing method and device, storage medium and electronic equipment |
| CN110824461A (en) * | 2019-11-18 | 2020-02-21 | 广东博智林机器人有限公司 | Positioning method |
| CN112229411A (en) * | 2020-10-15 | 2021-01-15 | 广州小鹏自动驾驶科技有限公司 | Data processing method and device |
| CN112229411B (en) * | 2020-10-15 | 2021-12-07 | 广州小鹏自动驾驶科技有限公司 | Data processing method and device |
| CN113099386A (en) * | 2021-03-25 | 2021-07-09 | 湘潭大学 | Multi-floor indoor position identification method and application thereof in museum navigation |
| CN115022961A (en) * | 2021-12-31 | 2022-09-06 | 荣耀终端有限公司 | Positioning method and device |
| CN115022961B (en) * | 2021-12-31 | 2023-06-02 | 荣耀终端有限公司 | Positioning method and equipment |
| CN116867061A (en) * | 2021-12-31 | 2023-10-10 | 荣耀终端有限公司 | Positioning method and equipment |
| CN116867061B (en) * | 2021-12-31 | 2024-04-16 | 荣耀终端有限公司 | Positioning method and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108966131A (en) | Fusion floor estimating method based on indoor positioning | |
| Tao et al. | An object detection system based on YOLO in traffic scene | |
| CN103796305B (en) | Indoor positioning method based on Wi-Fi position fingerprint | |
| Bae et al. | Large-scale indoor positioning using geomagnetic field with deep neural networks | |
| CN105933975A (en) | WiFi fingerprint-based accuracy improved indoor positioning method | |
| Liu et al. | End-to-end trajectory transportation mode classification using Bi-LSTM recurrent neural network | |
| CN107182036A (en) | The adaptive location fingerprint positioning method merged based on multidimensional characteristic | |
| CN105636201B (en) | Indoor orientation method based on sparse signal fingerprint database | |
| CN103905992B (en) | Indoor positioning method based on wireless sensor networks of fingerprint data | |
| CN107121690B (en) | A kind of dwell point recognition methods and device based on parameter of doing more physical exercises | |
| CN103634901B (en) | Novel location fingerprint based on Density Estimator gathers extracting method | |
| CN105869173A (en) | Stereoscopic vision saliency detection method | |
| CN106568445B (en) | Indoor trajectory predictions method based on bidirectional circulating neural network | |
| CN109671102A (en) | A kind of composite type method for tracking target based on depth characteristic fusion convolutional neural networks | |
| CN108521627A (en) | The indoor locating system and method for wifi and earth magnetism fusion based on HMM | |
| CN109798896A (en) | A kind of positioning of Indoor Robot with build drawing method and device | |
| Zhu et al. | Indoor/outdoor switching detection using multisensor DenseNet and LSTM | |
| CN110536245A (en) | A kind of indoor wireless positioning method and system based on deep learning | |
| CN108985195A (en) | Activity recognition method, apparatus, computer equipment and storage medium | |
| CN110213710A (en) | A kind of high-performance indoor orientation method, indoor locating system based on random forest | |
| CN101908213A (en) | SAR image change detection method based on quantum-inspired immune clone | |
| Zhu et al. | Accurate WiFi-based indoor localization by using fuzzy classifier and mlps ensemble in complex environment | |
| CN107194949B (en) | A kind of interactive video dividing method and system matched based on block and enhance Onecut | |
| CN106951735B (en) | A kind of prediction technique of signal peptide and its cleavage site based on layering mixed model | |
| CN105974360A (en) | Monitoring analysis method based on ADL and apparatus thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181207 |