JP6536162B2 - Movement information calculation method, movement information calculation apparatus, and movement information calculation program - Google Patents

Description

  The present invention relates to a movement information calculation method, a movement information calculation apparatus, and a movement information calculation program.

  BACKGROUND In recent years, an acceleration sensor has come to be mounted on a mobile terminal device such as a mobile phone or a smartphone, and a method of utilizing information of acceleration detected by the acceleration sensor has been studied.

  For example, a pedometer using acceleration information is also known (see, for example, Patent Document 1). The pedometer separates the acceleration vector of the walking body into a gravity acceleration vector and a walking acceleration vector generated by the walking of the walking body, and calculates an inner product value of the gravity acceleration vector and the walking acceleration vector. Then, the pedometer counts the number of steps on the basis of the time change of the calculated inner product value.

  There is also known a step counting device that easily and accurately distinguishes walking and non-walking (see, for example, Patent Document 2). The step counting device converts the acceleration into a digital sensor value, obtains a square value of the sensor value, and integrates the square value to calculate energy every time the square value becomes zero. Then, the step counting device counts the number of times the energy is equal to or more than a predetermined threshold, and divides the counter value by 2 to calculate the number of steps.

Unexamined-Japanese-Patent No. 2010-257395 JP, 2008-171347, A

  2. Description of the Related Art In recent years, a business that analyzes the behavior of a person based on the movement trajectory of the person and the like and provides advice or information specialized for each person has attracted attention. In order to analyze the movement trajectory of a person with a portable terminal device, it is desirable to implement a simple algorithm capable of calculating the movement trajectory even with a low central processing unit (CPU) mounted on the portable terminal device.

  The step counting method of Patent Document 1 is an algorithm capable of counting the number of steps even by a CPU having a relatively low processing capacity. However, this step counting method is not an algorithm for calculating a movement trajectory of a person. In addition, although a method using a particle filter is known as a method of calculating a movement trajectory, this method has a large amount of calculation and it is difficult to operate it with the CPU of a portable terminal device.

  Such a problem occurs not only in the case of calculating the movement locus in the portable terminal device but also in the case of calculating the movement locus in another information processing apparatus such as a server.

  In one aspect, the present invention aims to reduce the computational load of movement information.

  In one scheme, the computer converts a plurality of acceleration vectors representing accelerations of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system into a plurality of acceleration vectors in the second coordinate system. Convert. Then, the computer converts a plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations, obtains the moving distance and the moving direction of the moving object from the coefficients of terms of a predetermined order in the series, The movement information indicating the movement direction is output.

  According to one embodiment, the computational load of movement information can be reduced.

It is a functional block diagram of a movement information calculation apparatus. It is a flowchart of movement information calculation processing. It is a functional block diagram which shows the 1st specific example of a movement information calculation apparatus. It is a flowchart which shows the specific example of a movement information calculation process. It is a figure which shows a movement trace. It is a figure which shows a device coordinate system and a global coordinate system. It is a figure which shows the added movement vector. It is a figure which shows the movement trace of 2 steps. It is a figure which shows the time change of the component of a walking acceleration. It is a figure which shows the unit vector on xy plane. It is a functional block diagram which shows the 2nd specific example of a movement information calculation apparatus. It is a figure which shows geomagnetic mode and a gyro mode. It is a flowchart of coordinate system calculation processing. It is a flowchart of a 1st update process. It is a flowchart of a 2nd update process. It is a block diagram of an information processor.

Hereinafter, embodiments will be described in detail with reference to the drawings.
FIG. 1 shows an example of the functional configuration of the movement information calculation apparatus of the embodiment. The movement information calculation apparatus 101 of FIG. 1 includes a storage unit 111, a conversion unit 112, a calculation unit 113, and an output unit 114. The storage unit 111 stores the acceleration of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system. The conversion unit 112 and the calculation unit 113 obtain movement information using the acceleration at a plurality of times stored in the storage unit 111, and the output unit 114 outputs the movement information.

  FIG. 2 is a flowchart showing an example of movement information calculation processing performed by the movement information calculation device 101 of FIG. First, the conversion unit 112 converts a plurality of acceleration vectors representing accelerations at a plurality of times stored in the storage unit 111 into a plurality of acceleration vectors in the second coordinate system (step 201).

  Next, the calculation unit 113 converts the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations (step 202), and the moving distance and the moving direction of the moving object from the coefficients of the terms of the predetermined order in the series And (step 203). Then, the output unit 114 outputs movement information indicating the movement distance and the movement direction (step 404).

  According to the movement information calculation apparatus 101 of FIG. 1, the calculation load of movement information can be reduced.

  FIG. 3 shows a first example of the movement information calculation device 101 of FIG. The movement information calculation apparatus 101 of FIG. 3 is, for example, a portable terminal apparatus, and includes a storage unit 111, a conversion unit 112, a calculation unit 113, an output unit 114, an acceleration sensor 301, and a geomagnetic sensor 302. The conversion unit 112 includes a cycle calculation unit 311, a coordinate system calculation unit 312, and an acceleration conversion unit 313. The calculation unit 113 includes a coefficient calculation unit 321 and a movement vector calculation unit 322.

  The acceleration sensor 301 detects the acceleration of the moving object at each time, and the storage unit 111 stores information of the detected acceleration. The moving object corresponds to, for example, a user who carries the movement information calculation apparatus 101, and the detected acceleration represents an acceleration generated on the user's body. The geomagnetic sensor 302 detects the geomagnetic field at each time, and the storage unit 111 stores information of the detected magnetic field.

  The cycle calculation unit 311 calculates the number of time series data included in one cycle corresponding to a predetermined period. For example, the pedometer of Patent Document 1 counts the number of acceleration vectors detected while walking in P steps (P is an integer of 1 or more) in order to count the number of steps based on the acceleration vector of a person at each time. It is possible. When a period corresponding to P steps is used as one cycle, the number of acceleration vectors detected during that period can be used as the number of time-series data included in one cycle.

  In this case, the cycle calculation unit 311 separates an acceleration vector representing the acceleration of the moving object detected at each time into a gravity acceleration vector and a movement acceleration vector generated by the movement of the moving object. Next, the cycle calculation unit 311 calculates the inner product of the gravity acceleration vector and the movement acceleration vector, and counts the number of accelerations detected in a period from when the sign of the inner product is inverted to when it is inverted to the code before inversion. Do. This period corresponds to a half step.

  Then, the cycle calculation unit 311 determines the number of time-series data based on the counted number. For example, when one cycle corresponds to P steps, the cycle calculation unit 311 can repeat the counting operation 2P times to calculate the sum of the count values, and determine the sum as the number of time series data.

  The coordinate system calculation unit 312 uses the acceleration vector at the number of times calculated by the cycle calculation unit 311 and the plurality of magnetic field vectors representing the magnetic field at those times to calculate the global coordinate system from the coordinate system of the movement information calculation device 101. Find the transformation matrix to Usually, the mobile terminal device is not fixed, and when the user walks or runs, the position and posture change sequentially, so the coordinate system of the movement information calculation device 101 is in a state of rotating sequentially. Therefore, when calculating the movement distance and the movement direction, it is preferable to use a fixed global coordinate system. Below, the coordinate system of the movement information calculation apparatus 101 may be described as an apparatus coordinate system.

  The acceleration conversion unit 313 converts a plurality of acceleration vectors in the device coordinate system into a plurality of acceleration vectors in the global coordinate system using the conversion matrix obtained by the coordinate system calculation unit 312, and stores the converted acceleration vector in the storage unit 111. Store in

  The coefficient calculation unit 321 converts a plurality of acceleration vectors in the global coordinate system into a Fourier series by Fourier transform, and calculates cosine coefficients and sine coefficients of a predetermined order in the Fourier series. The movement vector calculation unit 322 obtains a movement distance and a movement direction from the cosine coefficient and the sine coefficient obtained by the coefficient calculation unit 321, calculates a movement vector indicated by the movement distance and the movement direction, and stores the movement vector in the storage unit 111.

  FIG. 4 is a flowchart showing a specific example of the movement information calculation process performed by the movement information calculation device 101 of FIG. First, the cycle calculation unit 311 calculates the number of time-series data included in one cycle using an acceleration vector representing the detected acceleration (step 401).

  Next, the coordinate system calculation unit 312 obtains a transformation matrix of coordinate conversion using acceleration vectors at the number of times calculated by the cycle calculation unit 311 and magnetic field vectors at those times (step 402).

  Next, the acceleration conversion unit 313 converts a plurality of acceleration vectors in the device coordinate system into a plurality of acceleration vectors in the global coordinate system using a conversion matrix of coordinate conversion (step 403).

  Next, the coefficient calculation unit 321 performs Fourier transform on a plurality of acceleration vectors in the global coordinate system to calculate cosine coefficients and sine coefficients in the Fourier series (step 404).

  Next, the movement vector calculation unit 322 calculates a movement vector from the cosine coefficient and the sine coefficient obtained by the coefficient calculation unit 321 (step 405), and the output unit 114 displays the movement vector on the screen (step 406) .

  Next, the movement information calculation device 101 determines whether to calculate the next cycle (step 407). The movement information calculation apparatus 101 may determine whether to calculate the next cycle based on an instruction input from the user via the input device. The movement information calculation apparatus 101 repeats the processing of step 401 and the subsequent steps when calculating the next cycle (step 407, NO), and ends the processing when the calculation of the next cycle is not performed (step 407, YES). .

  FIG. 5 shows an example of the movement trajectory displayed on the screen. In the xy plane of FIG. 5, the positive direction of the x axis represents east, and the positive direction of the y axis represents north. In this example, four cycles of processing results are displayed. First, the movement vector 501 is calculated in the first cycle, the movement vector 501 is displayed starting from the origin of the xy plane, and then the movement vector 502 is calculated in the second cycle, the end point of the movement vector 501 is the start A movement vector 502 is displayed.

  Similar calculations are repeated in the third and fourth cycles, and the motion vector 503 and the motion vector 504 are added. From the movement locus represented by movement vector 501 to movement vector 504, it is known that the user is at the end point of movement vector 504 starting from the position of the origin.

  Next, an example of calculating the movement vector in the fifth cycle will be described. The acceleration detected by the acceleration sensor 301 at time 1 to time m + 1 (m is an integer of 2 or more) is expressed as follows.

( _{Ax, 1} , _{ay, 1} , _{az, 1} ),. . . , (A _{x, m} , a _{y, m} , a _{z, m} ),
( _{Ax, m + 1} , _{ay, m + 1} , _{az, m + 1} )

a _{x, i} (i = 1 to _m + 1) represents the x component of the acceleration at time i in the device coordinate system, and a _{y, i} and a _{z, i} represent the y and z components of the acceleration at time i, respectively Represents ( _{Ax, i} , _{ay, i} , _{az, i} ) represents an acceleration vector at time i.

  A magnetic field detected by the geomagnetic sensor 302 at time 1 to time m + 1 is represented as follows.

( _{M x, 1} , m _{y, 1} , m _{z, 1} ),. . . , ( _{M x, m} , m _{y, m} , m _{z, m} ),
( _{M x, m + 1} , m _{y, m + 1} , m _{z, m + 1} )

m _{x, i} represents the x component of the magnetic field at time i in the device coordinate system, and m _{y, i} and m _{z, i} represent the y and z components of the magnetic field at time i, respectively. ( _{M x, i} , m _{y, i} , m _{z, i} ) represents the magnetic field vector at time i.

Among these, (ax _{, 1} , _{ay, 1} , _{az, 1} ) to (ax _{, m} , _{ay, m} , _{az, m} ) and ( _{mx, 1} , my _{, 1} , _{mz) , 1} ) to (m _{x, m} , m _{y, m} , m _{z, m} ) are used for calculation of movement vectors 501 to 504. In this case, using the acceleration vector and the magnetic field vector after time m + 1, the movement vector of the fifth cycle is calculated.

  For example, when a period corresponding to P steps is used as one cycle, the cycle calculation unit 311 obtains the number n of acceleration vectors detected during that period. Then, the coordinate system calculation unit 312 obtains a transformation matrix using n acceleration vectors and n magnetic field vectors after time m + 1. Here, in order to simplify the description, the representations of these acceleration vectors and magnetic field vectors are changed as follows.

( _{Ax, m + 1} , _{ay, m + 1} , _{az, m + 1} ), (ax _{, m + 2} , _{ay, m + 2} , _{az, m + 2} ),. . . ,
( _{Ax, m + n} , _{ay, m + n} , _{az, m + n} )
→ (ax _{, 1} , _{ay, 1} , _{az, 1} ), (ax _{, 2} , _{ay, 2} , _{az, 2} ),. . . ,
( _{Ax, n} , _{ay, n} , _{az, n} )

( _{M x, m + 1} , m _{y, m + 1} , m _{z, m + 1} ), (m _{x, m + 2} , m _{y, m + 2} , m _{z, m + 2} ),. . . ,
( _{M x, m + n} , m _{y, m + n} , m _{z, m + n} )
→ (m _{x, 1} , m _{y, 1} , m _{z, 1} ), (m _{x, 2} , m _{y, 2} , m _{z, 2} ),. . . ,
( _{M x, n} , m _{y, n} , m _{z, n} )

  FIG. 6 shows an example of a device coordinate system and a global coordinate system. The acceleration vector in the device coordinate system 601 is converted into an acceleration vector in the global coordinate system 602 by the conversion matrix obtained by the coordinate system calculation unit 312. The positive orientations of the x-axis, y-axis, and z-axis of global coordinate system 602 represent east, north, and height, respectively.

First, the coordinate system calculation unit 312 calculates an average of n acceleration vectors according to the following equation to obtain a gravity direction vector (g _x , g _y , g _z ).

Gravitational acceleration is a fixed value of acceleration and has a larger value than the moving acceleration caused by the movement of a moving object. Therefore, averaging n acceleration vectors cancels out the moving acceleration component and leaves only the gravity acceleration component. . In this example, the gravity direction vector (g _x , g _y , g _z ) represents the positive direction of the z axis in the global coordinate system 602.

Next, the coordinate system calculation unit 312 calculates an average of n magnetic field vectors according to the following equation to obtain an average magnetic field vector (m _x , m _y , m _z ).

The mean magnetic field vector (m _x , m _y , m _z ) includes a component in the north direction, but also includes a component in the gravity direction, and therefore is not orthogonal to the gravity direction vector. Therefore, the coordinate system calculation unit 312 calculates the outer product of the average magnetic field vector and the gravity direction vector by the following equation to obtain an east direction vector (e _x , e _y , e _z ).

(E _x , e _y , e _z )
= ( _{M y} g _z- m _z g _y , m _z g _x- m _x g _z , m _x g _y- m _y g _z ) (3)

East direction vector _{(e x, e y, e} z) , both the direction of gravity and north are orthogonal to a vector directed to the east.

Next, the coordinate system calculation unit 312 calculates the outer product of the gravity direction vector and the east direction vector by the following equation, and obtains the north direction vector (n _x , n _y , n _z ).

(N _x , n _y , n _z )
= (G _y e _z- g _z e _y , g _z e _x- g _x e _z , g _x e _y- g _y e _z ) (4)

The north direction vector (n _x , n _y , n _z ) is a vector that is orthogonal to both the gravity direction and the east direction and faces north. A global coordinate system is determined by three mutually orthogonal vectors of gravity direction vector, east direction vector, and north direction vector.

Depending on the specification of the acceleration sensor 301, the gravity direction vector (g _x , g _y , g _z ) of the equation (1) may represent the negative direction of the z axis in the global coordinate system 602. In this case, the order of the average magnetic field vector and the gravity direction vector may be switched in Equation (3) to calculate the outer product, and the order of the gravity direction vector and the east direction vector may be switched in Equation (4) to calculate the outer product.

  The coordinate system calculation unit 312 normalizes the three obtained vectors, and uses a component of each normalized vector to obtain a transformation matrix H of the following expression.

  The transformation matrix H is a rotation matrix that transforms a vector in the device coordinate system 601 into a vector in the global coordinate system 602.

The acceleration conversion unit 313 performs the linear transformation using the conversion matrix H to obtain the acceleration vector (ax _{, i} , _{ay, i} , _{az, i} ) in the device coordinate system 601 and the acceleration vector (global coordinate system 602). Convert to α _{x, i} , α _{y, i} , α _{z, i} ) (i = 1 to n).

Thereby, acceleration vectors (α _{x, 1} , α _{y, 1} , α _{z, 1} ) to (α _{x, n} , α _{y, n} , α _{z, n} ) in the global coordinate system 602 are obtained.

The coefficient calculation unit 321 performs discrete Fourier transform using (α _{x, 1} , α _{y, 1} , α _{z, 1} ) to (α _{x, n} , α _{y, n} , α _{z, n} ), and The kth-order cosine coefficient (cx _{, k} , _{cy, k} , _{cz, k} ) and the kth-order sine coefficient (sx _{, k} , _{sy, k} , _{sz, k} ) are calculated.

When one cycle corresponds to P steps, the coefficient calculation unit 321 calculates Pth-order cosine coefficient and Pth-order sine coefficient as k = P. Then, the movement vector calculation unit 322 calculates the movement vector (pos _x , pos _y ) according to the following equation from the Pth-order cosine coefficient and the Pth-order sine coefficient.

The reason why the movement vector can be calculated by equation (9) will be described later. In the right side of the equation (9), nT, (d1 + d2 · nT), or (cz, k / | cz, k |) (d1 + d2 · nT) may be used instead of c _{z, k} . T represents the sampling time of acceleration data, and d1 and d2 represent predetermined constants.

The output unit 114 adds the movement vector (pos _x , pos _y ) starting from the end point of the movement vector 504 as the movement vector 505 of the fifth cycle, as shown in FIG. 7. From the movement locus represented by movement vector 501 to movement vector 505, it can be known that the user is currently at the end point of movement vector 505.

Next, the reason why the movement vector can be calculated by equation (9) will be described. Although cz, k on the right side of equation (9) is not the movement distance itself, it represents a length corresponding to the movement distance,
Represents a unit vector in the movement direction. The human walking model can be represented by a mathematical model in which the walking acceleration rotates on a plane spanned by two vectors of the moving direction and the gravity direction, so that the validity of equation (9) can be explained.

  FIG. 8 shows an example of the movement trajectory of the head when the user walks two steps. When the user starts walking with the user's head at the position of the origin 801 of the global coordinate system, the head moves along the movement trajectory 802. The walking acceleration 811 initially points upward, and during walking, rotates on a plane spanned by two vectors of the movement direction and the gravity direction, and points upward again when walking one step. Then, similarly in the second step, the walking acceleration 811 makes one rotation on the same plane.

  The walking acceleration 811 at each time can be decomposed into the xy plane component 812 and the z component 813. When the xy plane component 812 is moved on the xy plane, a vector 814 is obtained. Each vector 814 on the xy plane can be further decomposed into x component 815 and y component 816.

  FIG. 9 shows an example of temporal change of the z component 813, the x component 815, and the y component 816 while the user walks two steps. The curve 901 represents the time change of the z component 813, the curve 902 represents the time change of the x component 815, and the curve 903 represents the time change of the y component 816. Each of the curves is a periodic waveform for two cycles, the curve 901 is a cosine waveform, and the curves 902 and 903 are sine waveforms.

In this case, the k-th cosine coefficient c _{z, k} (k = 2) in Expression (7) represents the amplitude 911 of the second-order component of the curve 901. Further, the k-th sine coefficient s _{x, k} (k = 2) in equation (8) represents the amplitude 912 of the second-order component of the curve 902, and the k-th sine coefficient s _{y, k} (k = 2) represents the curve The amplitude 913 of the second-order component 903 is shown.

In more detail, by fitting the method described in Japanese Patent Application No. 2014-189103 is prior application, the acceleration vector in the global coordinate system _{(α x, 1, α y} , 1, α z, 1) ~ (α x, _A continuous time acceleration curve can be determined from _n , α _{y, n} , α _{z, n} ). At this time, the continuous time acceleration curve is a Fourier series, and the second order cosine coefficient c _{z, 2} of the Fourier series is a coefficient of a second order cosine curve, and the second order sine coefficients s _{x, 2} and s _{y, 2} are 2 It is the coefficient of the next sine curve.

Since the curves 901 to 903 in FIG. 9 are ideal shapes, they are clean quadratic cosine curves or quadratic sine curves, but in practice they also include cosine curves or sine curves other than quadratic components. However, since components other than the second order can be regarded as noise, they are omitted in the calculation of the movement vector (pos _x , pos _y ) in equation (9).

As shown in FIG. 8, when starting to walk from the position of the origin 801, the walking acceleration 811 initially points upward and does not include the xy plane component, so the magnitude of the walking acceleration 811 as shown by the curve 901. Can be represented by a second-order cosine coefficient c _{z, 2} . And c _{z, k} (k = 2) of the right side of Formula (9) represents not the movement distance itself but the magnitude | size of walking acceleration.

Also, on the right side of equation (9)
As shown in FIG. 10, (k = 2) represents a unit vector 1001 of the moving direction on the xy plane. Since the magnitude of the walking acceleration and the movement distance can be considered to be proportional, it is possible to define the movement vector by equation (9). The same applies to cases other than k = 2.

The k-th cosine coefficient c _{z, k in} equation (7) and the approximate value of the k-th sine coefficient s _{x, k} and s _{y, k} in equation (8) can be obtained, for example, by Taylor expansion, and The movement vector (pos _x , pos _y ) of) can be obtained by the Newton method or the like. In this case, the movement vector can be calculated only by the four arithmetic operations, and the calculation load is reduced, so that the movement information calculation process of FIG. 4 can be executed even by a CPU with low processing capacity.

  Further, in step 401, by obtaining the number of time series data included in one cycle using the acceleration vector at each time, a set of acceleration vectors according to the number of steps such as 2 steps and 4 steps is determined. Therefore, the movement vector can be calculated accurately.

  Incidentally, instead of the magnetic field detected by the geomagnetic sensor 302, it is also possible to calculate a global coordinate system based on the angular velocity of the moving object detected by the angular velocity sensor.

  FIG. 11 shows a second specific example of the movement information calculation apparatus 101 of FIG. 1 using an angular velocity sensor. The movement information calculation apparatus 101 of FIG. 11 has a configuration in which a gyro sensor 1101 is added to the movement information calculation apparatus 101 of FIG. 3. The gyro sensor 1101 detects the angular velocity of the moving object at each time, and the storage unit 111 stores information on the detected angular velocity.

  FIG. 12 shows an example of a geomagnetic mode in which a global coordinate system is calculated based on a magnetic field, and a gyro mode in which a global coordinate system is calculated based on an angular velocity. For example, as shown by the movement trajectory 1201, when the user moves from the outdoors where the geomagnetism is stable to the indoor where the geomagnetism is unstable, the movement information calculation apparatus 101 operates in the geomagnetic mode outdoors, and the gyro is indoors. Operate in mode.

  In addition, the movement information calculation apparatus 101 does not fix the coordinate axes of the global coordinate system to the east, west, north, south, and as the movement locus 1202 indicates, the gyro mode is based on the position of the starting point 1204 and the moving direction 1203 at the starting point 1204 The global coordinate system may be calculated by

  FIG. 13 is a flowchart showing an example of coordinate system calculation processing performed by the coordinate system calculation unit 312 of FIG. 11 in step 402 of FIG. First, based on the magnetic field detected by the geomagnetic sensor 302, the coordinate system calculation unit 312 checks whether or not the geomagnetism is stable (step 1301). The coordinate system calculation unit 312 can determine, for example, whether or not the geomagnetism is stable by comparing values such as the strength or inclination of the magnetic field with a predetermined threshold.

  If the geomagnetism is stable (step 1301, YES), the coordinate system calculation unit 312 calculates a global coordinate system using the magnetic field detected by the geomagnetic sensor 302 (step 1302).

  On the other hand, if the geomagnetism is not stable (step 1301, NO), the coordinate system calculation unit 312 calculates a global coordinate system using the angular velocity detected by the gyro sensor 1101 (step 1303). In this case, the coordinate system calculation unit 312 calculates the transformation matrix H by updating the transformation matrix H calculated in the previous cycle based on the angular velocity.

  FIG. 14 is a flowchart showing an example of the first update process in step 1303 of FIG. The rotation matrix R obtained by transposing the transformation matrix H is expressed by the following equation.

  At this time, the normalized vectors x, y, and z are defined by the following equations.

Vector x = (x1, x2, x3) ^T (11)
Vector y = (y1, y2, y3) ^T (12)
Vector z = (z1, z2, z3) ^T (13)

Further, n angular velocities detected by the gyro sensor 1101 during one cycle are represented by angular velocity vectors b _i (i = 1 to n) of the following equation.

b _i = (b _{x, i} , b _{y, i} , b _{z, i} ) ^T (14)

  First, the coordinate system calculation unit 312 obtains a gravity direction vector g normalized using the equation (1) from n acceleration vectors (step 1401).

Next, using the inner product dθ of the _first angular velocity vector b1 and the gravity direction vector g, the coordinate system calculation unit 312 calculates the rotation angle φ according to the following equation (step 1402).

φ = dθΔT = angular velocity vector b ₁ · gravity direction vector gΔT (15)

  ΔT represents the sampling interval of each sensor, dθ represents the angular velocity around the vertical axis, and φ represents the rotation angle around the vertical axis during time ΔT.

  Next, the coordinate system calculation unit 312 calculates the rotation matrix Q around the z axis according to the following equation using the rotation angle φ (step 1403).

  Next, using the rotation matrix Q, the coordinate system calculation unit 312 calculates the rotation matrix R = R2 after update from the rotation matrix R = R1 before update according to the following equation (step 1404).

R2 = R1Q (17)

  As the rotation matrix R1 of Expression (17), a matrix obtained by replacing the vector z of the rotation matrix R in the previous cycle with the gravity direction vector g obtained in step 1401 can be used. Since the vector z does not change even if the rotation matrix R1 is multiplied by the rotation matrix Q, the vector z of the updated rotation matrix R2 also matches the gravity direction vector g.

  Next, the coordinate system calculation unit 312 checks whether or not n angular velocity vectors corresponding to one cycle have been processed (step 1405). If an unprocessed angular velocity vector remains (step 1405, NO), the coordinate system calculation unit 312 repeats the processing of step 1402 and subsequent steps using the next angular velocity vector. At this time, the updated rotation matrix R2 calculated in step 1404 is used as the new rotation matrix R1 of equation (17).

  Then, if all angular velocity vectors have been processed (step 1405, YES), the coordinate system calculation unit 312 ends the processing. The transposed matrix H is obtained by transposing the rotation matrix R2 calculated last.

  According to the update process of FIG. 14, when the movement information calculation apparatus 101 does not rotate much except around the vertical axis, it is possible to obtain an appropriate conversion matrix H. However, when the user holds the movement information calculation apparatus 101 and turns while tilting, the updated conversion matrix H does not necessarily represent the correct conversion.

FIG. 15 is a flow chart showing an example of the second updating process in step 1303 of FIG. First, the coordinate system calculation unit 312, by using a pre-update rotation matrix R1 1 th and the angular velocity vector b _1, the following equation to calculate the angular velocity vector r in the global coordinate system (step 1501).

Angular velocity vector r = R1 ^T b ₁ (21)

  As the rotation matrix R1 of equation (21), the rotation matrix R in the previous cycle can be used.

  Next, the coordinate system calculation unit 312 calculates the difference of the rotation matrix R using the angular velocity vector r (step 1502). At this time, the coordinate system calculation unit 312 generates the vectors v1 to v3 of the following equations using the elements of the rotation matrix R of equation (10).

Vector v1 = (x1, y1, z1) ^T (22)
Vector v2 = (x2, y2, z2) ^T (23)
Vector v3 = (x3, y3, z3) ^T (24)

  Next, using the angular velocity vector r, the coordinate system calculation unit 312 performs an outer product calculation such as the following equation.

Vector w1 = (w11, w12, w13) ^T
= Angular velocity vector r × vector v1 (25)
Vector w2 = (w21, w22, w23) ^T
= Angular velocity vector r × vector v 2 (26)
Vector w3 = (w31, w32, w33) ^T
= Angular velocity vector r × vector v3 (27)

  The outer product calculation of Equations (25) to (27) corresponds to performing differentiation of the vectors v1 to v3. The coordinate system calculation unit 312 generates a difference dR as expressed by the following equation using elements of the vectors w1 to w3.

  Next, using the difference dR, the coordinate system calculation unit 312 calculates the post-update rotation matrix R2 from the pre-update rotation matrix R1 according to the following equation (step 1503).

R2 = R1 + dR (29)

  The addition of equation (29) corresponds to the integration of the rotation matrix R.

  Next, the coordinate system calculation unit 312 checks whether or not n angular velocity vectors corresponding to one cycle have been processed (step 1504). If an unprocessed angular velocity vector remains (step 1504, NO), the coordinate system calculation unit 312 repeats the processing of step 1501 and subsequent steps using the next angular velocity vector. At this time, the updated rotation matrix R2 calculated in step 1503 is used as the new rotation matrix R1 of the equations (21) and (29).

  When all angular velocity vectors have been processed (YES in step 1504), the coordinate system calculation unit 312 checks whether or not a certain period has elapsed since the rotation matrix R2 was previously corrected (step 1505).

  If the fixed period has elapsed (YES in step 1505), the coordinate system calculation unit 312 corrects the rotation matrix R2 using the gravity direction vector g (step 1506). At this time, the coordinate system calculation unit 312 obtains the gravity direction vector g normalized using the equation (1) from the n acceleration vectors, and sets the vector z of the rotation matrix R2 after updating as the gravity direction vector g. replace. Then, the coordinate system calculation unit 312 sets the outer product of the vector z of the rotation matrix R2 and the vector x as the vector y, sets the outer product of the vector y and the vector z as the vector x, and normalizes the vector x and the vector y. , To generate a corrected rotation matrix R2.

  On the other hand, when the fixed period has not elapsed (step 1505, NO), the coordinate system calculation unit 312 ends the process. By transposing the rotation matrix R2 last calculated in step 1503 or the rotation matrix R2 corrected in step 1506, a transformation matrix H is obtained.

  According to the update processing of FIG. 15, the error of the transformation matrix H can be reduced by processing the rotation angle in a three-dimensional manner. Also, by periodically correcting the rotation matrix R2 in step 1506, it is possible to prevent the accumulation of errors.

  The configuration of the movement information calculation device 101 of FIGS. 1, 3 and 11 is merely an example, and some components may be omitted or changed according to the application and conditions of the movement information calculation device 101. For example, when the movement information calculation apparatus 101 of FIG. 11 operates only in the gyro mode, the geomagnetic sensor 302 can be omitted.

  When the storage unit 111 previously stores information on acceleration, magnetic field, and angular velocity, the acceleration sensor 301, the geomagnetic sensor 302, and the gyro sensor 1101 shown in FIGS. 3 and 11 can be omitted. Further, when the number of times corresponding to one cycle is determined in advance, the cycle calculation unit 311 of FIGS. 3 and 11 can be omitted.

  The flowcharts of FIG. 2, FIG. 4, and FIG. 13 to FIG. 13 are merely examples, and part of the processing may be omitted or changed according to the configuration and conditions of the movement information calculation apparatus 101. For example, when the number of times corresponding to one cycle is determined in advance, the process of step 401 in FIG. 4 can be omitted.

  In step 406 of FIG. 4, the output unit 114 may transmit information indicating the movement vector to another information processing apparatus via the communication network instead of displaying the movement vector on the screen. In this case, the information processing apparatus that has received the movement vector can display the movement vector on the screen.

  The movement trajectories in FIGS. 5 and 7 are merely examples, and the output unit 114 may display the movement trajectories on the screen in another format. The moving object may be an animal other than a human being, or may be an object such as a vehicle or a robot.

Expressions (1) to (29) are merely examples, and the movement information calculation process may be performed using another calculation expression or coordinate system. For example, the gravity direction vector (g _x , g _y , g _z ) and the magnetic field vector (m _x , m _y , m _z ) are calculated using a low pass filter operation instead of the equations (1) and (2) May be Also, instead of Fourier series, another frequency domain representation may be used to calculate the motion vector.

  The global coordinate system of FIGS. 5 to 10 is only an example, and another global coordinate system may be used to calculate the movement vector. For example, a global coordinate system based on the start point 1204 and the moving direction 1203 of FIG. 12 may be used.

  FIG. 16 shows a configuration example of an information processing apparatus for realizing the movement information calculation apparatus 101 of FIG. 1, FIG. 3 and FIG. The information processing apparatus of FIG. 16 includes a CPU 1601, a memory 1602, an input device 1603, an output device 1604, an auxiliary storage device 1605, a medium drive device 1606, and a network connection device 1607. These components are connected to one another by a bus 1608. The acceleration sensor 301, the geomagnetic sensor 302, and the gyro sensor 1101 of FIGS. 3 and 11 may be connected to the bus 1608.

  The memory 1602 is, for example, a semiconductor memory such as a read only memory (ROM), a random access memory (RAM), or a flash memory, and stores programs and data used for the movement information calculation process. The memory 1602 can be used as the storage unit 111 in FIGS. 1, 3 and 11.

  The CPU 1601 (processor) operates as the conversion unit 112 and the calculation unit 113 in FIGS. 1, 3 and 11 by executing a program using, for example, the memory 1602. The CPU 1601 also operates as the cycle calculation unit 311, the coordinate system calculation unit 312, the acceleration conversion unit 313, the coefficient calculation unit 321, and the movement vector calculation unit 322 in FIGS. 3 and 11.

  The input device 1603 is, for example, a keyboard, a pointing device, etc., and is used for inputting an instruction or information from an operator or a user. The output device 1604 is, for example, a display device, a printer, a speaker, or the like, and is used to inquire or instruct an operator or a user, and to output a processing result. The processing result may be a motion vector. The output device 1604 can be used as the output unit 114 of FIGS. 1, 3 and 11.

  The auxiliary storage device 1605 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, a tape device or the like. The auxiliary storage device 1605 may be a hard disk drive or a flash memory. The information processing apparatus can store programs and data in the auxiliary storage device 1605, load them into the memory 1602, and use them. The auxiliary storage device 1605 can be used as the storage unit 111 in FIGS. 1, 3 and 11.

  The medium drive device 1606 drives a portable recording medium 1609 and accesses the recorded contents. The portable recording medium 1609 is a memory device, a flexible disk, an optical disk, a magneto-optical disk or the like. The portable recording medium 1609 may be a Compact Disk Read Only Memory (CD-ROM), a Digital Versatile Disk (DVD), a Universal Serial Bus (USB) memory, or the like. An operator or a user can store programs and data in this portable recording medium 1609, load them into the memory 1602, and use them.

  Thus, the computer readable recording medium for storing the program and data used for the movement information calculation process is physically (non-temporary, such as the memory 1602, the auxiliary storage device 1605, or the portable recording medium 1609). ) Recording medium.

  The network connection device 1607 is a communication interface that is connected to a communication network such as a Local Area Network, a Wide Area Network, etc., and performs data conversion involved in communication. The information processing device can receive programs and data from an external device via the network connection device 1607, load them into the memory 1602, and use them. The network connection device 1607 can be used as the output unit 114 of FIGS. 1, 3 and 11.

  The information processing apparatus can also receive a processing request from the user terminal via the network connection device 1607, perform movement information calculation processing, and transmit the processing result to the user terminal.

  Note that the information processing apparatus need not include all the components shown in FIG. 16, and some of the components may be omitted depending on the application and conditions. For example, when the information processing apparatus receives a processing request from a user terminal via a communication network, the input device 1603 and the output device 1604 may be omitted. In addition, when the portable recording medium 1609 or the communication network is not used, the medium drive device 1606 or the network connection device 1607 may be omitted.

  When the information processing apparatus is a portable terminal apparatus having a call function, it may include a call apparatus such as a microphone and a speaker, or may include an imaging apparatus such as a camera.

  While the disclosed embodiments and their advantages have been described in detail, those skilled in the art can make various changes, additions, and omissions without departing from the scope of the present invention as set forth in the claims. I will.

The following appendices will be further disclosed regarding the embodiment described with reference to FIGS. 1 to 16.
(Supplementary Note 1)
The computer is
Converting a plurality of acceleration vectors representing accelerations of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system into a plurality of acceleration vectors in the second coordinate system;
Converting the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations;
The moving distance and the moving direction of the moving object are obtained from the coefficients of terms of a predetermined order in the series;
Outputting movement information indicating the movement distance and the movement direction;
A movement information calculation method characterized in that.
(Supplementary Note 2)
The computer converts the plurality of acceleration vectors in the second coordinate system into the series by Fourier transformation, obtains the movement distance from the cosine coefficient of the predetermined order in the series, and the sine of the predetermined order in the series The movement information calculation method according to claim 1, wherein the movement direction is determined from a coefficient.
(Supplementary Note 3)
The computer uses the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times and the plurality of magnetic field vectors representing the magnetic fields at the plurality of times detected by the geomagnetic sensor, A conversion matrix from the second coordinate system to the second coordinate system, and using the conversion matrix, the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are determined by the plurality of accelerations in the second coordinate system The movement information calculation method according to appendix 1 or 2, characterized by converting into a vector.
(Supplementary Note 4)
The computer determines a transformation matrix from the first coordinate system to the second coordinate system using a plurality of angular velocity vectors representing the angular velocity of the moving object at the plurality of times detected by the angular velocity sensor. The plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are converted into the plurality of acceleration vectors in the second coordinate system, using a transformation matrix. Movement information calculation method.
(Supplementary Note 5)
The computer is
The acceleration vector representing the acceleration of the moving object detected at each time by the acceleration sensor is separated into a gravity acceleration vector and a movement acceleration vector generated by the movement of the moving object,
Calculating an inner product of the gravity acceleration vector and the movement acceleration vector;
Counting the number of accelerations detected in a period from when the sign of the inner product is inverted to when the sign is inverted to the sign before the inversion;
The number of the plurality of times is determined based on the number of the counted accelerations,
The movement information calculation method according to any one of appendices 1 to 4, characterized in that:
(Supplementary Note 6)
A storage unit storing accelerations of moving objects at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system;
A conversion unit configured to convert a plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times into a plurality of acceleration vectors in a second coordinate system;
A calculating unit that converts the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations, and obtains a moving distance and a moving direction of the moving object from coefficients of terms of a predetermined order in the series;
An output unit that outputs movement information indicating the movement distance and the movement direction;
A movement information calculation apparatus comprising:
(Appendix 7)
The calculation unit converts the plurality of acceleration vectors in the second coordinate system into the series by Fourier transformation, obtains the movement distance from the cosine coefficient of the predetermined order in the series, and calculates the moving distance in the series. The movement information calculation apparatus according to appendix 6, wherein the movement direction is determined from a sine coefficient.
(Supplementary Note 8)
The conversion unit uses the plurality of acceleration vectors representing the accelerations of the moving object at the plurality of times and the plurality of magnetic fields vectors representing the magnetic fields at the plurality of times detected by the geomagnetic sensor. A transformation matrix from the system to the second coordinate system is determined, and using the transformation matrix, the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are determined by the plurality of the plurality of acceleration vectors in the second coordinate system. The movement information calculation apparatus as set forth in claim 6 or 7, characterized in that it is converted into an acceleration vector.
(Appendix 9)
The conversion unit obtains a conversion matrix from the first coordinate system to the second coordinate system using a plurality of angular velocity vectors representing the angular velocity of the moving object at the plurality of times detected by the angular velocity sensor. The plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are converted into the plurality of acceleration vectors in the second coordinate system, using the conversion matrix. Movement information calculation device.
(Supplementary Note 10)
The conversion unit is
The acceleration vector representing the acceleration of the moving object detected at each time by the acceleration sensor is separated into a gravity acceleration vector and a movement acceleration vector generated by the movement of the moving object,
Calculating an inner product of the gravity acceleration vector and the movement acceleration vector;
Counting the number of accelerations detected in a period from when the sign of the inner product is inverted to when the sign is inverted to the sign before the inversion;
The number of the plurality of times is determined based on the number of the counted accelerations,
The movement information calculation apparatus according to any one of appendices 6 to 9, characterized in that
(Supplementary Note 11)
Converting a plurality of acceleration vectors representing accelerations of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system into a plurality of acceleration vectors in the second coordinate system;
Converting the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations;
The moving distance and the moving direction of the moving object are obtained from the coefficients of terms of a predetermined order in the series;
Outputting movement information indicating the movement distance and the movement direction;
A movement information calculation program that causes a computer to execute processing.
(Supplementary Note 12)
The computer converts the plurality of acceleration vectors in the second coordinate system into the series by Fourier transformation, obtains the movement distance from the cosine coefficient of the predetermined order in the series, and the sine of the predetermined order in the series The movement information calculation program according to appendix 11, wherein the movement direction is determined from a coefficient.
(Supplementary Note 13)
The computer uses the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times and the plurality of magnetic field vectors representing the magnetic fields at the plurality of times detected by the geomagnetic sensor, A conversion matrix from the second coordinate system to the second coordinate system, and using the conversion matrix, the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are determined by the plurality of accelerations in the second coordinate system The movement information calculation program according to appendix 11 or 12, characterized by converting into a vector.
(Supplementary Note 14)
The computer determines a transformation matrix from the first coordinate system to the second coordinate system using a plurality of angular velocity vectors representing the angular velocity of the moving object at the plurality of times detected by the angular velocity sensor. The plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are converted into the plurality of acceleration vectors in the second coordinate system, using a transformation matrix. Movement information calculation program.
(Supplementary Note 15)
The computer is
The acceleration vector representing the acceleration of the moving object detected at each time by the acceleration sensor is separated into a gravity acceleration vector and a movement acceleration vector generated by the movement of the moving object,
Calculating an inner product of the gravity acceleration vector and the movement acceleration vector;
Counting the number of accelerations detected in a period from when the sign of the inner product is inverted to when the sign is inverted to the sign before the inversion;
The number of the plurality of times is determined based on the number of the counted accelerations,
The movement information calculation program according to any one of appendices 11 to 14, characterized in that

101 movement information calculation apparatus 111 storage unit 112 conversion unit 113 calculation unit 114 output unit 301 acceleration sensor 302 geomagnetic sensor 311 cycle calculation unit 312 coordinate system calculation unit 313 acceleration conversion unit 321 coefficient calculation unit 322 movement vector calculation unit 501 to 505 movement vector 601 device coordinate system 602 global coordinate system 801 origin 802, 1201, 1202 movement locus 811 walking acceleration 812 xy plane component 813 z component 814 vector 815 x component 816 y component 901 to 903 curve 911 to 913 amplitude 1001 unit vector 1101 gyro sensor 1203 Movement direction 1204 Starting point 1601 CPU
1602 Memory 1603 Input Device 1604 Output Device 1605 Auxiliary Storage Device 1606 Media Drive Device 1607 Network Connection Device 1608 Bus 1609 Portable Storage Media

Claims (7)

The computer is
Converting a plurality of acceleration vectors representing accelerations of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system into a plurality of acceleration vectors in the second coordinate system;
Converting the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations;
The moving distance and the moving direction of the moving object are obtained from the coefficients of terms of a predetermined order in the series;
Outputting movement information indicating the movement distance and the movement direction;
A movement information calculation method characterized in that.   The computer converts the plurality of acceleration vectors in the second coordinate system into the series by Fourier transformation, obtains the movement distance from the cosine coefficient of the predetermined order in the series, and the sine of the predetermined order in the series The movement information calculation method according to claim 1, wherein the movement direction is determined from a coefficient.   The computer uses the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times and the plurality of magnetic field vectors representing the magnetic fields at the plurality of times detected by the geomagnetic sensor, A conversion matrix from the second coordinate system to the second coordinate system, and using the conversion matrix, the plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are determined by the plurality of accelerations in the second coordinate system The movement information calculation method according to claim 1 or 2, wherein the movement information is converted into a vector.   The computer determines a transformation matrix from the first coordinate system to the second coordinate system using a plurality of angular velocity vectors representing the angular velocity of the moving object at the plurality of times detected by the angular velocity sensor. The plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times are converted into the plurality of acceleration vectors in the second coordinate system using a transformation matrix. How to calculate movement information. The computer is
The acceleration vector representing the acceleration of the moving object detected at each time by the acceleration sensor is separated into a gravity acceleration vector and a movement acceleration vector generated by the movement of the moving object,
Calculating an inner product of the gravity acceleration vector and the movement acceleration vector;
Counting the number of accelerations detected in a period from when the sign of the inner product is inverted to when the sign is inverted to the sign before the inversion;
The number of the plurality of times is determined based on the number of the counted accelerations,
The movement information calculation method according to any one of claims 1 to 4, characterized in that: A storage unit storing accelerations of moving objects at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system;
A conversion unit configured to convert a plurality of acceleration vectors representing the acceleration of the moving object at the plurality of times into a plurality of acceleration vectors in a second coordinate system;
A calculating unit that converts the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations, and obtains a moving distance and a moving direction of the moving object from coefficients of terms of a predetermined order in the series;
An output unit that outputs movement information indicating the movement distance and the movement direction;
A movement information calculation apparatus comprising: Converting a plurality of acceleration vectors representing accelerations of the moving object at a plurality of times within a predetermined period detected by the acceleration sensor in the first coordinate system into a plurality of acceleration vectors in the second coordinate system;
Converting the plurality of acceleration vectors in the second coordinate system into a series of frequency domain representations;
The moving distance and the moving direction of the moving object are obtained from the coefficients of terms of a predetermined order in the series;
Outputting movement information indicating the movement distance and the movement direction;
A movement information calculation program that causes a computer to execute processing.