WO2021054502A1

WO2021054502A1 - Floor identification apparatus and method

Info

Publication number: WO2021054502A1
Application number: PCT/KR2019/012418
Authority: WO
Inventors: 박용완; 허수정; 임란아시라프
Original assignee: 영남대학교 산학협력단
Priority date: 2019-09-17
Filing date: 2019-09-24
Publication date: 2021-03-25
Also published as: KR102080532B1

Abstract

A floor identification apparatus and method are disclosed. A floor identification apparatus according to one disclosed embodiment comprises: a fingerprint storage unit for storing a magnetic fingerprint database for each floor of a building of which a floor is to be identified; a reception unit for receiving acceleration data, gyroscope data, and magnetic data from a user terminal; a direction determination unit for determining the direction of the user terminal on the basis of the acceleration data and the gyroscope data; a transformation unit for transforming the magnetic data by using the direction of the user terminal as a reference coordinate system; a distance calculation unit for calculating the Euclidean distance between the transformed magnetic data and the magnetic fingerprints; and a floor identification unit for identifying, as the floor on which a user is located, the floor from which a magnetic fingerprint allowing a minimum Euclidean distance is collected.

Description

Stair identification device and method

The disclosed embodiments relate to a stair identification technique using magnetic data.

With the development of GPS (Global Positioning System) technology, it is possible to accurately and quickly determine the location of a user who has a terminal capable of receiving satellite signals. Through this, rapid progress has been made in the fields of transportation, transportation, and military, but at the same time, the need for a technology for tracking the user's location in a building environment where it is difficult to receive GPS satellite signals has emerged.

In order to accurately track the user's location inside a multi-story building, the process of identifying the stair where the user is located must precede. To this end, a technology for identifying the stair where the user is located by receiving a signal from a service access point (SAP) using a Wi-Fi network has been proposed, but there is a problem that the identification result is sensitive to building materials, wall separation, etc. do.

The disclosed embodiments are to identify which floor the user is on by using acceleration data, gyroscope data, or magnetic data.

A stair identification device according to an embodiment includes a fingerprint storage unit that stores a magnetic fingerprint database for each floor of a stair identification target building, acceleration data from a user terminal, gyroscope data, and magnetic ) A receiving unit for receiving data, a direction determining unit for determining the direction of the user terminal based on the acceleration data and the gyroscope data, a conversion unit for converting the magnetic data by using the direction of the user terminal as a reference coordinate system, transformation A distance calculation unit that calculates an Euclidean distance between the magnetic data and the magnetic fingerprint and the stairs on which the magnetic fingerprints having the minimum Euclidean distance are collected are identified as the stairs where the user is located. It includes a stair identification unit.

The fingerprint storage unit may store the magnetic fingerprint based on global coordinates regardless of the direction of the user terminal.

The direction determination unit may further include an activity classification unit for classifying an activity of a user with the user terminal into daily walking, a call, and a terminal swing based on the acceleration data and the gyroscope data. .

The direction determination unit may determine the direction of the user terminal based on the classified user's activity.

The activity classifier may be a machine learning model using one of K-Nearest Neighbors (K-NN), Decision Trees (DT), and Naive Bayes (NB) algorithms.

The conversion unit, Equation 1 below

[Equation 1]

(At this time,

Is the magnetic data based on the global coordinate system,

Is the magnetic data with the direction of the user terminal as a reference coordinate system,

Is the Roll rotation matrix,

Is the pitch rotation matrix,

Is the yaw rotation matrix)

In this way, the magnetic data can be converted.

The receiving unit receives the acceleration data, the gyroscope data, and the magnetic data from the user terminal by repeating a preset number of times at a preset time interval, and the direction determining unit repeats a preset number of times at a preset time interval To determine the direction of the user terminal, the conversion unit converts the magnetic data by repeating a preset number of times at a preset time interval, and the distance calculating unit repeats a preset number of times at a preset time interval The Clidian distance is calculated, and the stair identification unit generates a set of stair candidates including a plurality of stair candidates by repeatedly identifying the stair where the user is located by repeating a preset number of times at a preset time interval, and a mode value among the stair candidates Can be identified as the final stair where the user is located.

The stair identification unit, Equation 2 below

[Equation 2]

(At this time,

Is the stairs where the user is located,

Is the set of stair candidates,

Is the set of Euclidean distances)

By this, the stairs on which the user is located can be identified.

The stair identification device calculates a maximum value of the acceleration data by repeating a preset number of times at a preset time interval, and determines whether the user is moving on the stairs of the building based on the maximum value. It may further include a determination unit.

The stair identification device calculates a maximum change amount per time interval of the acceleration data by repeating a preset number of times at a preset time interval, and determines whether the user is moving on the stairs of the building based on the maximum change amount. It may further include a stair change determination unit to determine.

The stair identification method according to an embodiment includes storing a magnetic fingerprint database for each floor of a stair identification target building, acceleration data, gyroscope data, and magnetic data from a user terminal. Receiving, determining a direction of the user terminal based on the acceleration data and the gyroscope data, converting the magnetic data using the direction of the user terminal as a reference coordinate system, the converted magnetic data and the And calculating a Euclidean distance between magnetic fingerprints, and identifying a layer in which the magnetic fingerprints having the minimum Euclidean distance are collected as a layer on which the user is located.

The storing may include storing the magnetic fingerprint based on global coordinates irrespective of the direction of the user terminal.

The determining may further include classifying an activity of a user with the user terminal into a daily walking, a call, or a terminal swing based on the acceleration data and the gyroscope data. .

In the determining step, the direction of the user terminal may be determined based on the classified activity of the user.

In the classifying step, one of K-Nearest Neighbors (K-NN), Decision Trees (DT), and Naive Bayes (NB) algorithms may be used.

The converting step is Equation 1 below

[Equation 1]

(At this time,

Is the magnetic data based on the global coordinate system,

Is the Roll rotation matrix,

Is the pitch rotation matrix,

Is the yaw rotation matrix)

In this way, the magnetic data can be converted.

The receiving step includes receiving the acceleration data, the gyroscope data, and the magnetic data from the user terminal by repeating a preset number of times at a preset time interval, and the determining step includes a preset time interval. The direction of the user terminal is determined by repeating a number of times, and the converting step is to convert the magnetic data by repeating a preset number of times at a preset time interval, and the calculating step is a preset time interval. The Euclidean distance is calculated by repeating a number of times, and the identifying step is repeated for a preset number of times at a preset time interval to identify the stairs on which the user is located, thereby generating a set of stairs candidates including a plurality of stairs. And, the mode among the stair candidates may be identified as the final stair in which the user is located.

The identifying step is Equation 2 below

[Equation 2]

(At this time,

Is the stairs where the user is located,

Is the set of stair candidates,

Is the set of Euclidean distances)

By this, the stairs on which the user is located can be identified.

The stair identification method includes calculating a maximum value of the acceleration data by repeating a preset number of times at a preset time interval, and determining whether the user is moving on the stairs of the building based on the maximum value. It may contain more.

The stair identification method includes calculating a maximum change amount per time interval of the acceleration data by repeating a preset number of times at a preset time interval, and determining whether the user is moving on the stairs of the building based on the maximum change amount. It may further include the step of determining.

According to the disclosed embodiments, by collecting a magnetic fingerprint for each floor of a target building to identify a floor, a magnetic database that is stable against changes other than changes in magnetic materials inside the building may be obtained.

In addition, according to the disclosed embodiments, by performing stair identification for a predetermined number of times at a preset time interval, and determining the mode of the identified results as the finally identified stair, relatively compared to the one-time stair identification method. It can provide accurate stair identification results.

In addition, according to the disclosed embodiments, by calculating a physical quantity derived from acceleration data, it may be determined whether a user is moving on a staircase in a building.

1 is a block diagram illustrating an apparatus for identifying a stair according to an exemplary embodiment;

2 is a block diagram illustrating an apparatus for identifying a stair according to another exemplary embodiment;

3 is a flowchart illustrating a method for identifying a stair according to an exemplary embodiment

4 is a flowchart illustrating a method for determining a stair change according to an exemplary embodiment

5 is a flowchart illustrating a method of determining a stair change according to another embodiment

Hereinafter, a specific embodiment will be described with reference to the drawings. The following detailed description is provided to aid in a comprehensive understanding of the methods, devices, and/or systems described herein. However, this is only an example, and the disclosed embodiments are not limited thereto.

In describing the embodiments, if it is determined that a detailed description of related known technologies may unnecessarily obscure the subject matter of the disclosed embodiments, a detailed description thereof will be omitted. In addition, terms to be described later are terms defined in consideration of functions in the disclosed embodiments and may vary according to the intention or custom of users or operators. Therefore, the definition should be made based on the contents throughout the present specification. The terms used in the detailed description are only for describing the embodiments, and should not be limiting. Unless explicitly used otherwise, expressions in the singular form include the meaning of the plural form. In the present description, expressions such as "comprising" or "feature" are intended to refer to certain features, numbers, steps, actions, elements, some or combination thereof, and one or more It should not be construed to exclude the presence or possibility of other features, numbers, steps, actions, elements, any part or combination thereof.

1 is a block diagram of an apparatus 100 for identifying stairs according to an exemplary embodiment.

Referring to FIG. 1, as shown, the stair identification apparatus 100 according to an embodiment includes a fingerprint storage unit 110, a reception unit 120, a direction determination unit 130, a conversion unit 140, and a distance calculation. It includes a unit 150 and a stair identification unit 160.

The fingerprint storage unit 110 stores a magnetic fingerprint database for each floor of a building to be identified.

Hereinafter, the fingerprint refers to a form of magnetic data of a target building, which is measured in advance for stair identification. Specifically, a plurality of measurement points for each layer may be set at regular intervals and magnetic data may be collected at the corresponding points at a preset sampling frequency. In this case, magnetic data of an empty space between one measurement point and another measurement point may be generated through spline interpolation based on an average of the collected magnetic data of the two points.

According to an embodiment, the fingerprint storage unit 110 may receive normalized magnetic data after being measured in an offline state from an external device and store it as a magnetic fingerprint.

According to an embodiment, the fingerprint storage unit 110 may store a magnetic fingerprint based on global coordinates irrespective of the direction of the user terminal.

Specifically, hereinafter, the global coordinate system refers to a fixed earth coordinate system.

The receiver 120 receives acceleration data, gyroscope data, and magnetic data from a user terminal.

Hereinafter, the gyroscope data refers to data measured through a gyro sensor in the user terminal, and is composed of a yaw, a pitch, and a roll. At this time, yaw refers to the angle at which the x-axis or y-axis rotates when the z-axis is rotated, and the pitch refers to the angle at which the x-axis or z-axis rotates when the y-axis is rotated, and roll Means the angle at which the y-axis or z-axis rotates when the x-axis is rotated.

At this time, each data can be divided into values for each direction of the x-axis, y-axis, and z-axis, and in particular, acceleration data is divided into values for each direction and then the acceleration obtained by correcting the bias in each direction due to gravity. It can be data.

According to an embodiment, the receiver 120 may receive acceleration data, gyroscope data, and magnetic data from a user terminal by repeating a preset number of times at preset time intervals.

For example, the receiving unit 120 may receive data from the user terminal over a total of 5 times, once per second, or receive data from the user terminal over a total of 5 times, once per frame. I can. However, the data reception time interval and the number of receptions are not limited thereto. Hereinafter, all the functions'repetitively performed by a preset number of times at each preset time interval' may be exemplarily performed at the time interval and the number of times of reception, and a more detailed description thereof will be omitted.

The direction determination unit 130 determines the direction of the user terminal based on acceleration data and gyroscope data.

According to an embodiment, the direction determination unit 130 may determine the direction of the user terminal by repeating a preset number of times at each preset time interval.

In addition, according to an embodiment, the direction determination unit 130 is based on the acceleration data and the gyroscope data, based on the activity of the user holding the user terminal, a normal walking activity, a call listening, and a terminal. It may further include an activity classification unit (not shown) for classifying as a swing.

In addition, according to an embodiment, the direction determination unit 130 may determine the direction of the user terminal based on the classified user's activity.

Hereinafter, the user holding the terminal and moving while holding the terminal in a certain space is a daily walking, the user holding the terminal close to the ear and communicating with the other party is a call, and the user holding the terminal in hand and moving his arm back and forth. Moving while moving is referred to as a terminal swing. However, the activity classifier (not shown) may classify the user's activities according to any criteria other than the above-described criteria, provided that it is a classification criterion for determining the direction of the user terminal.

Specifically, each magnitude of the values in the x-axis, y-axis, and z-axis directions of acceleration data and gyroscope data is

And

On the unit, it can be distinguished according to the above three activity states of the user.

Accordingly, the direction determination unit 130 refers to information on the x-axis, y-axis, and z-axis directions of the user terminal, which are mainly measured during a specific user's activity, and refers to the gyroscope data repeatedly received by the receiver 120. It can be used to determine the direction of the user terminal. For example, as the direction determination unit 130 repeatedly determines the direction of the user terminal, information on the x-axis, y-axis, and z-axis directions of the user terminal according to the accumulated user activity increases, so the direction determination unit ( 130) can be improved.

In this case, the activity classification unit (not shown) may be a machine learning model using one of K-Nearest Neighbors (K-NN), Decision Trees (DT), and Naive Bayes (NB) algorithms, but is not limited thereto, If the user's activity can be classified into a preset limited type of activity, it may be a regression analysis model or a deep learning model.

The conversion unit 140 converts magnetic data using the determined direction of the user terminal as a reference coordinate system.

According to an embodiment, the conversion unit 140 may convert magnetic data according to Equation 1 below.

[Equation 1]

At this time,

Is the magnetic data based on the global coordinate system,

Is magnetic data with the direction of the user terminal as the reference coordinate system,

Is the Roll rotation matrix,

Is the pitch rotation matrix,

Represents the yaw rotation matrix.

Specifically, the roll rotation matrix

Can be expressed by Equation 2 below.

[Equation 2]

At this time,

Represents the rotation angle around the x-axis.

Also specifically, the pitch rotation matrix

Can be expressed by Equation 3 below.

[Equation 3]

At this time,

Represents the rotation angle around the y-axis.

Also specifically, the yaw rotation matrix

Can be expressed by Equation 4 below.

[Equation 4]

At this time,

Represents the rotation angle around the z-axis.

In addition, according to an embodiment, the conversion unit 140 may convert the magnetic data by repeating a preset number of times at each preset time interval.

The distance calculator 150 calculates an Euclidean distance between the converted magnetic data and the magnetic fingerprint.

According to an embodiment, the distance calculator 150 may calculate the Euclidean distance by repeating a preset number of times at each preset time interval.

In addition, according to an embodiment, the distance calculator 150 may calculate a Euclidean distance between the magnetic data converted by Equation 5 below and the magnetic fingerprint.

[Equation 5]

At this time,

Is a set of Euclidean streets,

Silver converted magnetic data

i sets,

Denotes a set of j magnetic fingerprints for each stair for building b.

In the above formula, the distance calculation unit 150 is the converted magnetic data

For i, the Euclidean distance can be calculated by sequentially substituting j fingerprints corresponding to j floors of the building in Equation 5 above.

The stair identification unit 160 identifies a stair in which a magnetic fingerprint whose calculated Euclidean distance is the minimum is collected as a stair in which the user is located.

Specifically, the stair identification unit 160 may identify the stair where the user is located according to Equation 6 below.

[Equation 6]

At this time,

Is the stair where the user is located,

Is the set of stair candidates,

Denotes a set of Euclidean distances.

In the above formula, the stair identification unit 160 is

Corresponding to the minimum Euclidean distance among the elements of

The elements of

It can be set as.

According to an embodiment, the stair identification unit 160 generates a set of stair candidates including a plurality of stair candidates by repeatedly identifying a stair where the user is located by repeating a preset number of times at a preset time interval, and calculating a mode value among the stair candidates. It can be identified as the last stair where the user is located.

At this time, the stair identification unit 160 repeats the calculation through Equation 6 described above as many times as a preset number for each preset time interval,

Corresponding to the minimum Euclidean distance among the elements of

Multiple elements of can be derived. Subsequently, the stair identification unit 160 is

The mode of the elements of

It can be set as.

2 is a block diagram of an apparatus 200 for identifying stairs according to another exemplary embodiment.

Components corresponding to components in the embodiment described with reference to FIG. 1 perform the same or similar functions as described in the embodiment, and thus a more detailed description thereof will be omitted.

Referring to FIG. 2, the stair identification apparatus 200 according to an embodiment further includes a stair change determination unit 210.

According to an embodiment, the stair change determination unit 210 calculates a maximum value of acceleration data by repeating a preset number of times at a preset time interval, and determines whether the user is moving on the stairs of the building based on the maximum value. I can judge.

Specifically, the stair change determination unit 210 sets a plurality of measurement points in the building to be identified at a predetermined interval, and the maximum value of the acceleration data when the user is moving on the stairs at the predetermined sampling frequency at the corresponding point. It may include a machine learning model that collects a maximum value of acceleration data when moving through a flat corridor.

Subsequently, the stair change determination unit 210 inputs the maximum value of the calculated acceleration data to the machine learning model learned based on the maximum values of the collected acceleration data, thereby determining whether the user is moving on the stairs of the building. I can judge.

In addition, according to an embodiment, the stair change determination unit 210 calculates the maximum amount of change per time interval of acceleration data by repeating a preset number of times at each preset time interval, and based on the maximum amount of change, the user moves on the stairs of the building. You can determine whether you are moving.

Specifically, the stair change determination unit 210 sets a plurality of measurement points in the building to be identified at a predetermined interval, and at the predetermined sampling frequency at the corresponding point, the acceleration data per time interval of the acceleration data when the user is moving on the stairs. A machine learning model that collects the maximum amount of change and the maximum amount of change per time interval of acceleration data while moving through a flat corridor may be included.

Subsequently, the stair change determination unit 210 inputs the maximum amount of change per time interval of the calculated acceleration data to the machine learning model learned based on the maximum amount of change per time interval of the collected acceleration data, so that the user It can be determined whether or not the stomach is moving.

3 is a flowchart illustrating a method for identifying stairs according to an exemplary embodiment.

The method shown in FIG. 3 may be performed, for example, by the above-described

stair identification devices

100 and 200. In the illustrated flowchart, the stair identification method is divided into a plurality of steps, but at least some of the steps are performed in a different order, combined with other steps, performed together, omitted, divided into detailed steps, or not shown. One or more steps may be added and performed.

Referring to FIG. 3, first, the

stair identification devices

100 and 200 store a magnetic fingerprint database for each floor of a stair identification target building (310).

Thereafter, the

stairs identification apparatuses

100 and 200 receive acceleration data, gyroscope data, and magnetic data from the user terminal (320).

According to an embodiment, the

stairs identification apparatuses

100 and 200 may receive acceleration data, gyroscope data, and magnetic data from a user terminal by repeating a preset number of times at preset time intervals.

For example, the

stair identification devices

100 and 200 may receive data from the user terminal once per second for a total of 5 times, or once per frame for a total of 5 times. You can receive data. However, the data reception time interval and the number of receptions are not limited thereto. Hereinafter, all functions performed by'repeating a preset number of times at each preset time interval' may be exemplarily performed in the above-described time interval and the number of times of reception, and a more detailed description thereof will be omitted.

Thereafter, the

stairs identification devices

100 and 200 determine the direction of the user terminal based on the received acceleration data and gyroscope data (330).

According to an embodiment, the

stairs identification apparatuses

100 and 200 may determine the direction of the user terminal by repeating a preset number of times at each preset time interval.

In addition, according to an embodiment, in determining the direction of the user terminal (330), the stairs identification device (100, 200) routinely monitors the activity of the user holding the user terminal based on acceleration data and gyroscope data. It can be classified into walking, talking, and terminal swing.

In this case, the

stair identification devices

100 and 200 use one of the K-Nearest Neighbors (K-NN), Decision Trees (DT) and Naive Bayes (NB) algorithms in classifying the activities of the user with the user terminal. It may be a machine learning model, but is not limited thereto, and may be a regression analysis model or a deep learning model if the user's activity can be classified into a preset limited type of activity.

Accordingly, the

stairs identification apparatuses

100 and 200 may determine the direction of the user terminal based on the classified user's activity.

Thereafter, the

stairs identification apparatuses

100 and 200 convert the received magnetic data using the determined direction of the user terminal as a reference coordinate system (340).

According to an embodiment, the

stair identification devices

100 and 200 may convert magnetic data by repeating a preset number of times at a preset time interval.

In addition, according to an embodiment, the

stairs identification devices

100 and 200 may convert magnetic data according to Equation 1 described above.

Thereafter, the

stairs identification apparatuses

100 and 200 calculate a Euclidean distance between the converted magnetic data and the magnetic fingerprint (350).

According to an embodiment, the

stairs identification apparatuses

100 and 200 may calculate the Euclidean distance by repeating a preset number of times at each preset time interval.

Further, according to an embodiment, the

stairs identification apparatuses

100 and 200 may calculate a Euclidean distance between the magnetic data converted by Equation (5) and the magnetic fingerprint.

Thereafter, the

stairs identification apparatuses

100 and 200 identify the stairs in which the magnetic fingerprints of which the calculated Euclidean distance is the minimum are collected as the stairs in which the user is located (360).

According to an embodiment, the

stair identification apparatuses

100 and 200 generate a set of stair candidates including a plurality of stair candidates by repeatedly identifying the stair where the user is located by repeating a predetermined number of times at a preset time interval, and among the stair candidates. The mode can be identified as the final stair where the user is located.

In addition, according to an embodiment, the stair identification device 200 may identify the stair where the user is located according to Equation 6 described above.

4 is a flowchart illustrating a method of determining a stair change according to an exemplary embodiment.

According to an embodiment, the stair identification device 200 calculates a maximum value of acceleration data by repeating a preset number of times at a preset time interval, and determines whether the user is moving on the stairs of the building based on the maximum value. can do.

5 is a flowchart illustrating a method of determining a stair change according to another exemplary embodiment.

According to another embodiment, the stair identification device 200 calculates a maximum change amount per time interval of acceleration data by repeating a preset number of times at a preset time interval, and whether the user is moving on the stairs of the building based on the maximum change amount. You can judge whether or not.

In the flowcharts shown in FIGS. 4 and 5, the method for identifying the stairs is divided into a plurality of steps, but at least some of the steps are performed in a different order, combined with other steps, performed together, omitted, or divided into detailed steps. It may be performed, or may be performed by adding one or more steps not shown.

Although the exemplary embodiments of the present invention have been described in detail above, those of ordinary skill in the art to which the present invention pertains will understand that various modifications can be made to the above-described embodiments without departing from the scope of the present invention. . Therefore, the scope of the present invention is limited to the described embodiments and should not be determined, and should not be determined by the claims to be described later, but also by the claims and equivalents.

Claims

A fingerprint storage unit that stores a magnetic fingerprint database for each floor of a building to be identified;

A receiver for receiving acceleration data, gyroscope data, and magnetic data from a user terminal;

A direction determination unit determining a direction of the user terminal based on the acceleration data and the gyroscope data;

A conversion unit for converting the magnetic data using the direction of the user terminal as a reference coordinate system;

A distance calculator configured to calculate an Euclidean distance between the converted magnetic data and the magnetic fingerprint; And

And a stair identification unit for identifying a stair in which the magnetic fingerprint having the minimum Euclidean distance is collected as a stair in which the user is located.
The method according to claim 1,

The fingerprint storage unit stores the magnetic fingerprint based on global coordinates irrespective of the direction of the user terminal.
The method according to claim 1,

The direction determination unit further comprises an activity classification unit for classifying an activity of a user holding the user terminal into daily walking, a call, and a terminal swing based on the acceleration data and the gyroscope data. Identification device.
The method of claim 3,

The direction determining unit determines a direction of the user terminal based on the classified user's activity.
The method of claim 4,

The activity classification unit is a machine learning model using one of K-Nearest Neighbors (K-NN), Decision Trees (DT), and Naive Bayes (NB) algorithms, a stair identification device.
The method according to claim 1,

The conversion unit, Equation 1 below

[Equation 1]

(At this time,
Is the magnetic data based on the global coordinate system,
Is the magnetic data with the direction of the user terminal as a reference coordinate system,
Is the Roll rotation matrix,
Is the pitch rotation matrix,
Is the yaw rotation matrix)

A stair identification device for converting the magnetic data by means of.
The method according to claim 1,

The receiving unit receives the acceleration data, the gyroscope data, and the magnetic data from the user terminal by repeating a preset number of times at each preset time interval,

The direction determining unit determines the direction of the user terminal by repeating a preset number of times at each preset time interval,

The conversion unit converts the magnetic data by repeating a preset number of times at each preset time interval,

The distance calculator calculates the Euclidean distance by repeating a preset number of times at each preset time interval,

The stair identification unit generates a set of stair candidates including a plurality of stair candidates by repeatedly identifying a stair in which the user is located by repeating a preset number of times at a preset time interval, and determining a mode among the stair candidates as a final stair in which the user is To identify, stair identification device.
The method of claim 7,

The stair identification unit, Equation 2 below

[Equation 2]

(At this time,
Is the stairs where the user is located,
Is the set of stair candidates,
Is the set of Euclidean distances)

By identifying the stairs on which the user is located, the stairs identification device.
The method according to claim 1,

Further comprising a stair change determination unit that repeatedly calculates a maximum value of the acceleration data by repeating a preset number of times at a preset time interval, and determines whether the user is moving on the stairs of the building based on the maximum value, Stair identification device.
The method according to claim 1,

A stair change determination unit that repeats a preset number of times at a preset time interval, calculates the maximum change amount per time interval of the acceleration data, and determines whether the user is moving on the stairs of the building based on the maximum change amount. Further comprising, stair identification device.
Storing a magnetic fingerprint database for each floor of the building to be identified;

Receiving acceleration data, gyroscope data, and magnetic data from a user terminal;

Determining a direction of the user terminal based on the acceleration data and the gyroscope data;

Converting the magnetic data using the direction of the user terminal as a reference coordinate system;

Calculating a Euclidean distance between the converted magnetic data and the magnetic fingerprint; And

And identifying a stair in which the magnetic fingerprint having the minimum Euclidean distance is collected as a stair in which the user is located.
The method of claim 11,

The storing step includes storing the magnetic fingerprint based on global coordinates regardless of the direction of the user terminal.
The method of claim 11,

The determining step further comprises classifying an activity of a user carrying the user terminal into daily walking, a call, and a terminal swing based on the acceleration data and the gyroscope data. Identification method.
The method of claim 13,

The determining step determines the direction of the user terminal based on the classified activity of the user.
The method of claim 14,

The classifying step uses one of K-Nearest Neighbors (K-NN), Decision Trees (DT) and Naive Bayes (NB) algorithms.
The method of claim 11,

The converting step is Equation 1 below

[Equation 1]

(At this time,
Is the magnetic data based on the global coordinate system,
Is the magnetic data with the direction of the user terminal as a reference coordinate system,
Is the Roll rotation matrix,
Is the pitch rotation matrix,
Is the yaw rotation matrix)

Transforming the magnetic data by means of a step identification method.
The method of claim 11,

The receiving step includes receiving the acceleration data, the gyroscope data, and the magnetic data from the user terminal by repeating a preset number of times at a preset time interval,

In the determining step, the direction of the user terminal is determined by repeating a preset number of times at each preset time interval,

In the converting step, the magnetic data is converted by repeating a preset number of times at a preset time interval,

In the calculating, the Euclidean distance is calculated by repeating a preset number of times at a preset time interval,

In the step of identifying, a set of stairs candidates including a plurality of stairs is generated by repeating a predetermined number of times at a preset time interval to identify the stairs on which the user is located, and a mode value among the stairs candidates is determined as a final How to identify stairs, identifying by stairs.
The method of claim 17,

The identifying step is Equation 2 below

[Equation 2]

(At this time,
Is the stairs where the user is located,
Is the set of stair candidates,
Is the set of Euclidean distances)

By identifying the stairs on which the user is located, the stairs identification device.
The method of claim 11,

Step identification further comprising calculating a maximum value of the acceleration data by repeating a preset number of times at a preset time interval, and determining whether the user is moving on the stairs of the building based on the maximum value Way.
The method of claim 11,

The step of calculating the maximum change amount per time interval of the acceleration data by repeating a preset number of times at a preset time interval, and determining whether the user is moving on the stairs of the building based on the maximum change amount. That, how to identify stairs.