JP6146962B2 - Magnetic data processing device - Google Patents

Description

  The present invention relates to a magnetic data processing apparatus including a three-axis magnetic sensor.

  Conventionally, a magnetic data processing device that detects a geomagnetic vector using a three-axis magnetic sensor mounted on a mobile device such as a mobile phone, and calculates the orientation of the mobile device and the orientation of the mobile device using the detected value It has been known.

  Mobile devices are equipped with various electronic components in addition to the three-axis magnetic sensor. Therefore, the triaxial magnetic sensor is affected by the magnetic field (internal magnetic field) generated from these electronic components, and the geomagnetic vector may not be measured accurately. In order to solve this problem, an azimuth measuring device and an electronic compass are known that correct the influence of an internal magnetic field so that the geomagnetic vector can be accurately measured and use the result (Patent Documents 1 and 2 below). reference).

In order to correct the influence of the internal magnetic field, as shown in FIG. 14, the center point O ′ of the azimuth sphere s drawn by the geomagnetic vector M is calculated in accordance with the attitude change of the portable device. Then, a deviation (offset) between the origin O of the three-axis magnetic sensor and the center point O ′ of the azimuth sphere s is corrected. That is, using the detected magnetic vector OM and the correction vector OO ′, a vector O′M starting from the center point O ′ is calculated from the following equation.
O'M = OM-OO '(1)

  As a method of calculating the center point O ′, for example, as described in Patent Document 2, as shown in FIG. 14, a plurality of detected values of the geomagnetic vector M are obtained, and each of them is obtained by, for example, the least square method. The point where the variation in the distance from the detected value is the smallest is calculated. This point is the center point O 'in FIG.

  When the center point O ′ is calculated in this way, it is desirable that the plurality of detected values of the geomagnetic vector M are not biased on the azimuth s. If the detected value is biased, the azimuth sphere s cannot be accurately calculated, and the calculation accuracy of the center point O ′ tends to be lowered. Therefore, in order to improve the calculation accuracy of the center point O ′ by this method, it is necessary to cause the user to point the mobile device in various directions and to acquire the detected value of the geomagnetic vector M uniformly on the azimuth sphere s. is there.

  Further, as another method for calculating the center point O ′, in Patent Document 1 below, four points separated from each other as much as possible that satisfy a predetermined condition on the azimuth sphere s are selected from the obtained detection values. A method of calculating the center point O ′ from the four points is disclosed.

JP 2008-241675 A JP-A-2005-195376

  However, in the method of statistically calculating the center point O ′ from a plurality of obtained detection value groups as in Patent Document 2, in order to improve the calculation accuracy, the user is caused to point the mobile device in various directions. There is a problem that if the operation is forced to improve the calculation accuracy, the burden on the user increases.

  On the other hand, in order to reduce the burden on the user, a method of obtaining a plurality of detected values of the geomagnetic vector M and obtaining the center point O ′ using the detected values while using the mobile device naturally. Is also possible. However, in this case, since the user does not consciously point the mobile device in various directions, the detected values tend to be biased on the azimuth s. Therefore, there is a problem that the calculation accuracy of the center point O ′ tends to be lowered. Further, in Patent Document 2, when sufficient calculation accuracy cannot be ensured such that only biased detection values are obtained, the obtained reference point is discarded, and a highly accurate reference point is obtained by discarding. That doesn't mean the problem is not solved.

  Further, in the method of obtaining the center point O ′ by selecting detection values that are as far as possible from each other and satisfying a predetermined condition as in Patent Document 1, the obtained detection values can be obtained without forcing the user to perform an operation. A detection value that can accurately calculate an offset can be selected from among them, but there is a problem that it may take time until a detection value that satisfies a predetermined condition is obtained.

  The present invention has been made in view of such a background, and does not force the user to perform a rotation operation, and even if the detected value is obtained only with a biased value, the center point of the azimuth sphere can be accurately determined. A magnetic data processing device capable of calculation is to be provided.

One aspect of the present invention includes a three-axis magnetic sensor that detects geomagnetism as a magnetic vector in a three-axis orthogonal coordinate system fixed to a portable device, and an azimuth sphere drawn by the magnetic vector as the posture of the portable device changes A magnetic data processing device that corrects the influence of the internal magnetic field of the portable device on the three-axis magnetic sensor by calculating a center point,
A memory for storing a plurality of detected values of the magnetic vector obtained during a posture change period in which the posture of the portable device is changing;
Rotation motion determination that determines whether or not a detection value group consisting of a plurality of detection values of the magnetic vector in one posture change period is obtained while rotating the posture of the portable device around one axis Means,
Rotation for calculating the rotation axis of the rotation when the rotation value determination means determines that the detected value group is obtained while rotating the attitude of the portable device around one axis. An axis calculation means;
From the detected value groups acquired during a plurality of different posture change periods, the intersections or approximate intersections of the plurality of rotation axes obtained by the rotational motion determination means and the rotation axis calculation means respectively are the center points of the azimuth sphere. A center point calculating means for calculating as
Offset correcting means for correcting an offset which is a deviation between the origin of the three-axis orthogonal coordinate system and the center point of the azimuth sphere;
A magnetic data processing apparatus comprising: (Claim 1).

In the magnetic data processing device, the rotation axis of the portable device is calculated using the detection value group. Then, using this rotation axis, the center point of the azimuth sphere is calculated.
If it does in this way, it will become easy to reduce a user's burden. That is, according to research by the inventors, when a user naturally uses a mobile device, for example, the user can change the direction while walking with the mobile device in a pocket, take out the mobile device from the pocket, or view the screen. In many cases, the movement of the mobile device is to rotate the posture of the portable device around one axis. Of course, since the rotation is not forced, it is rare that the rotation operation is 180 ° or more, but it can be sufficiently expected that the rotation operation is about 90 °. Therefore, if the center point of the azimuth sphere is calculated by calculating a plurality of the rotation axes using a plurality of detection value groups and obtaining the intersection or approximate intersection of the plurality of rotation axes, the detection value obtained is obtained. Even if the offset is biased when viewed from the whole azimuth and the conventional method cannot obtain the offset with high accuracy, the offset can be calculated with high accuracy without forcing the user to perform a rotating operation. It becomes possible.

In addition, the magnetic data detection device of this example includes rotational motion determination means for determining whether or not the detected value group is obtained while rotating the posture of the mobile device around one axis.
In this way, only when it is determined that the detected value group is obtained while rotating the posture of the mobile device around one axis, the detected value group is used for calculating the center point of the azimuth sphere. Can be used. Therefore, the center point of the azimuth sphere can be calculated with high accuracy.

In this example, a plurality of rotation axes are calculated, a combination of two rotation axes is selected from the plurality of rotation axes, and the intersection or approximate intersection is calculated as the center point of the azimuth sphere.
When the two selected rotation axes intersect, the intersection is the center point of the azimuth sphere. In addition, since the calculated rotation axis includes an error, the two rotation axes may not intersect each other, but even in this case, the approximate intersection of the two rotation axes is calculated as the center point of the azimuth sphere. This makes it possible to calculate the center point with certainty. Here, the approximate intersection can be, for example, the midpoint of the place where the two rotation axes are closest to each other.

  As described above, according to this example, it is possible to provide a magnetic data processing device that can accurately calculate the center point of the azimuth sphere without forcing the user to perform a rotation operation.

1 is a conceptual diagram of a magnetic data processing device in Embodiment 1. FIG. 1 is a perspective view of a mobile device in Embodiment 1. FIG. 1 is a block diagram of a magnetic data processing device in Embodiment 1. FIG. The conceptual diagram of the detected value of a magnetic vector, a circle | round | yen, and an orientation sphere in Example 1. FIG. The elements on larger scale of FIG. The detection value of the magnetic vector seen from the direction parallel to the plane containing a circle in Example 1. FIG. 3 is a diagram illustrating distances from the center of a circle to each detected value of a magnetic vector in the first embodiment. The figure showing rotation angle (theta) of the attitude | position of a portable apparatus in one attitude | position change period in Example 1. FIG. Explanatory drawing of the method of correct | amending the detected value of a magnetic vector using the center point of an azimuth | direction sphere in Example 1. FIG. 5 is a flowchart for calculating a center point of an azimuth sphere according to the first embodiment. The figure following FIG. The figure following FIG. The figure following FIG. Explanatory drawing of the method of measuring the center point of an orientation sphere in a prior art example.

  The magnetic data processing device displays, for example, an electronic compass that calculates the orientation of the portable device using the detected magnetic vector value, a magnetic gyro that calculates the rotational angular velocity of the portable device, and the orientation of the portable device together with map information It can be used for all devices that use a value obtained by correcting the geomagnetic vector with an offset, such as a navigation device.

In the magnetic data processing device, the rotational motion determination means determines whether the plurality of detection values included in one detection value group are collected on or around a circle. It is preferable to determine whether or not the value group is obtained while rotating the attitude of the portable device around one axis (claim 2).
In this case, it can be easily determined whether or not the detected value group is obtained while rotating the posture of the portable device around one axis. That is, when the attitude of the mobile device is rotated around one axis, the magnetic vector draws a circle on the azimuth sphere. Therefore, whether or not the detected value group is acquired while rotating the attitude of the portable device around one axis by determining whether or not the detected value group is gathered on or around the circumference. Can be easily determined.

Further, the rotational motion determination means calculates an equation of the circle from each of the detected values included in one detected value group by a least square method, and the plurality of detected values are calculated based on the calculated circumference of the circle. It is preferable to determine whether or not they are gathered at or near the top (claim 3).
In this case, since the least square method is used, the circle can be calculated with high accuracy. Therefore, it can be accurately determined whether or not the detected value group is acquired while the posture of the mobile device is rotated around one axis. Further, since a straight line passing through the center of the circle perpendicular to the plane including the circle becomes the rotation axis, the rotation axis can be accurately calculated by accurately obtaining the circle by the least square method. Therefore, the center point of the azimuth sphere can be calculated accurately.

Further, the rotational motion determination means has a variation in distance from a plane including the circle to each of the detected values not more than a predetermined value, and a variation in distance from the center of the circle to each of the detected values. When the value is equal to or less than a predetermined value, it is preferable to determine that the detected value group is obtained while the portable device is rotated about one axis (claim 4).
In this case, it is possible to accurately determine whether or not the detected value group is obtained while rotating the posture of the mobile device around one axis. That is, when the detection value group is obtained while rotating the mobile device around one axis, all of the individual detection values included in the detection value group exist near the plane, and It exists near the above circle. Therefore, the variation in distance from the plane to the individual detection values and the variation in distance from the center of the circle to the individual detection values are reduced. In addition, when the detection value group is not obtained while rotating the mobile device around one axis, the individual detection values included in the detection value group do not exist near the plane or circle, and the above variation Becomes bigger. Therefore, it is possible to accurately determine whether or not the detected value group is obtained while rotating the portable device around one axis by determining whether or not the variation is larger than a predetermined value. It becomes possible.
The variation can be, for example, the standard deviation σ.

Further, it is preferable that the rotation axis calculation means is configured to calculate the rotation axis only when the radius of the circle is equal to or less than a predetermined value.
In this case, the center point of the azimuth sphere can be accurately obtained. That is, if the radius of the circle is too large, the triaxial magnetic sensor may be affected by an external magnetic field, and the center point of the azimuth sphere may not be accurately calculated. Therefore, by calculating the rotation axis only when the radius of the circle is equal to or less than a predetermined value, it becomes possible to eliminate the influence of the external magnetic field and accurately determine the center point of the azimuth sphere. The predetermined value can be, for example, a value obtained by adding an amount corresponding to an error to the magnitude of geomagnetism in a place where the mobile device is used.

Further, it is preferable that the rotation axis calculation unit is configured to calculate the rotation axis only when the rotation angle of the attitude of the mobile device in one attitude change period is a predetermined value or more (claim). Item 6).
In this case, the calculation accuracy of the center point of the azimuth sphere can be further increased. That is, when the rotation angle is small, it is difficult to accurately calculate the circle, and thus it is difficult to accurately calculate the rotation axis. Therefore, the calculation accuracy of the center point of the azimuth sphere tends to be lowered. Therefore, if the rotation axis is calculated only when the rotation angle is greater than or equal to a predetermined value, the calculation accuracy of the center point can be improved. Since the rotation angle only needs to be large enough to calculate the rotation axis, the rotation of 180 ° or more as in the case of forcing the user to perform the rotation operation is not required, and the predetermined value can be set to 60 °, for example.

Example 1
An embodiment relating to the magnetic data processing apparatus will be described with reference to FIGS. As shown in FIG. 1, the magnetic data processing apparatus 1 of this example includes a three-axis magnetic sensor 2, a memory 3, a rotational motion determination unit 4, a rotational axis calculation unit 5, a center point calculation unit 6, and an offset correction. Means 7. The triaxial magnetic sensor 2 detects geomagnetism as a magnetic vector M in a triaxial orthogonal coordinate system fixed to the mobile device 10 (see FIG. 2). The magnetic data processing device 1 calculates the center point O ′ of the azimuth sphere s (see FIG. 4) drawn by the magnetic vector M in accordance with the change in the attitude of the mobile device 10, so that the triaxial magnetism due to the internal magnetic field of the mobile device 10 is obtained. The influence on the sensor 2 is corrected.

The memory 3 stores a plurality of detected values of the magnetic vector M obtained during the posture change period in which the posture of the mobile device 10 is changing.
The rotational motion determination unit 4 determines whether or not the detected value group d including a plurality of detected values of the magnetic vector M in one posture change period is obtained while rotating the posture of the mobile device 10 around one axis. Determine whether.
The rotation axis calculation means 5 rotates when the rotation value determination means 4 determines that the detected value group d is obtained while rotating the attitude of the mobile device 10 around one axis. The rotation axis A (see FIG. 4) is calculated.

The center point calculation means 6 includes two rotation axes out of the plurality of rotation axes A obtained by the rotation motion determination means 4 and the rotation axis calculation means 5 from the detection value groups d acquired during a plurality of different posture change periods. And the intersection or approximate intersection is calculated as the center point O ′ of the bearing sphere s.
The offset correction means 7 corrects an offset that is a deviation between the origin O (see FIG. 4) of the three-axis orthogonal coordinate system and the center point O ′ of the azimuth sphere s.

  In this example, the magnetic data processing apparatus 1 is used as an electronic compass. As shown in FIG. 2, in the mobile device 10, various electronic components such as a microcomputer 8 and a microphone 12 are mounted in addition to the triaxial magnetic sensor 2. An internal magnetic field is generated by these electronic components, and the three-axis magnetic sensor 2 is affected. The magnetic data processing device 1 corrects the influence of the internal magnetic field on the three-axis magnetic sensor 2 and obtains an accurate value of the geomagnetic vector M. Using this value, the correct orientation of the mobile device 10 is calculated and displayed on the display unit 11.

  As shown in FIG. 3, the microcomputer 8 includes a CPU 80, a ROM 81, a RAM (memory 3), an I / O 82, and a line 83 that connects them. The ROM 81 stores a program 81p. When the CPU 80 reads out and executes the program 81p, the rotational motion determination means 4, the rotational axis calculation means 5, the center point calculation means 6, and the offset correction means 7 are realized.

  If the position of the geomagnetism (magnetic vector M) does not change greatly on the earth, it can be considered that the orientation and the magnitude are always constant regardless of the orientation of the mobile device 10. As shown in FIG. 4, the end point of the geomagnetic vector M draws a sphere (azimuth sphere s) on a three-axis orthogonal coordinate system. Since the triaxial magnetic sensor 2 is affected by the internal magnetic field of the mobile device 10, the center O ′ of the azimuth s is deviated from the origin O of the triaxial magnetic sensor 2. The magnetic data processing apparatus 1 of this example corrects the influence of the internal magnetic field on the three-axis magnetic sensor 2 by calculating the center point O ′.

  The triaxial magnetic sensor 2 acquires a plurality of detection values (detection value group d) of the magnetic vector M while the posture of the mobile device 10 is changing (posture change period). The triaxial magnetic sensor 2 acquires a plurality of detection value groups d in different posture change periods. In this example, the center point O ′ of the azimuth sphere s is calculated using the plurality of detection value groups d.

  A method for calculating the center point O ′ will be described. First, as shown in FIG. 5, the rotational motion determination means 4 is used to calculate a circle C from individual detection values included in one detection value group d using the least square method. That is, the center point B and the radius where the variation in distance from each detected value becomes the smallest are obtained by the least square method, and the sum of squares of the distance from the plane including the circle C to each detected value is minimized. The circle C is calculated as follows. Then, by determining whether or not individual detection values are gathered on or around the circle C, the detection value group d is obtained while rotating the posture of the mobile device 10 around one axis. Judge whether or not

This determination is made as follows. First, as shown in FIG. 6, it is determined whether or not the variation (σ) of the distance x from the plane P including the circle C to each detected value is equal to or less than a predetermined value. Here, when the variation in the distance x exceeds a predetermined value, it is determined that the detected value group d is not obtained while the attitude of the mobile device 10 is rotated around one axis, and this detected value group The calculation of the center point O ′ using d is stopped. Then, another detection value group d is acquired.
When the variation in the distance x is equal to or smaller than the predetermined value, as shown in FIG. 7, it is determined whether or not the variation (σ) in the distance r from the center B of the circle C to each detected value is equal to or smaller than the predetermined value. to decide. Here, when the variation in the distance r exceeds a predetermined value, it is determined that the detected value group d is not obtained while the portable device 10 is rotated about one axis, and this detected value group d is used. The calculation of the center point O ′ that was present is stopped. Then, another detection value group d is acquired.

In this example, it is checked whether or not the calculated radius R of the circle C is equal to or smaller than a predetermined value. When the radius R is larger than a predetermined value, the triaxial magnetic sensor 2 may be affected by an external magnetic field. Therefore, in this case, the calculation of the center point O ′ using the detected value group d is stopped, and another detected value group d is acquired.
When not influenced by the external magnetic field, the maximum value of the radius R is set to the radius of the azimuth s (see FIG. 4), that is, the magnitude | M | equal. Since the magnitude | M | of the geomagnetic vector M can be considered to be substantially constant unless the measurement location moves greatly during acquisition of the detection value group d, the radius R is equal to | M | When the value is larger than the value, it can be determined that the triaxial magnetic sensor 2 may be affected by the external magnetic field. Since | M | differs depending on the region, it is preferable to set an appropriate value according to the planned use region of the device to which the present invention is applied.

  In this example, as shown in FIG. 8, the rotation angle θ of the posture of the mobile device 10 in one posture change period is calculated, and it is determined whether or not the rotation angle θ is a predetermined value or more. If the posture of the mobile device 10 is not sufficiently rotated and the rotation angle θ is small, the circle C cannot be accurately calculated, and the calculation accuracy of the center point O ′ may be lowered. Therefore, in this case, the calculation of the center point O ′ using this detection value group d is terminated, and another detection value group d is acquired.

Note that the rotation angle θ is almost 180 ° or less in the case of the present invention on the assumption that the user is not forced to perform the rotation operation, and therefore the end point of the magnetic vector M (n) obtained at the end. May be considered to be farthest from the end point of the magnetic vector M (1) obtained first, and the determination of whether or not it is larger than the predetermined value can be simplified as follows. For example, among a plurality of detection values included in the detection value group d, the distance h from the end point of the magnetic vector M (1) obtained first to the end point of the magnetic vector M (n) obtained last is calculated. To do. Then, it is determined whether or not a value h / R obtained by dividing the distance h by the radius R of the circle C is larger than a predetermined value.
Further, when the circle C is far away from the center point O ′ of the azimuth s, the radius R becomes small, and the calculation accuracy of the rotation axis A is lowered even if the rotation angle θ is the same value. It is also possible to check whether the distance h satisfies | M (n) −M (1) | ≧ h, and to prevent a reduction in calculation accuracy due to the small radius R of the circle C.

  By performing the processing described above, the detected value group d is obtained by the rotational motion determination means 4 while the attitude of the mobile device 10 is rotated about one axis, and a three-axis magnetic sensor 2 is not affected by the external magnetic field, and it can be confirmed that the rotation angle θ is sufficiently large to calculate the rotation axis A with high accuracy. Then, after this confirmation is completed, a straight line passing through the center of the circle C perpendicular to the plane P including the circle C is calculated as shown in FIG. This straight line is the rotation axis A of the rotational movement of the mobile device 10.

  Then, a plurality of detection value groups d (d1, d2,...) Are acquired during a plurality of different posture change periods, and a plurality of rotation axes A (A1, A2,. ) Is calculated. A combination of two rotation axes is selected from the plurality of rotation axes A, and the intersection or approximate intersection is calculated as the center point O 'of the azimuth s.

As shown in FIG. 9, after calculating the center point O ′, the deviation between the origin O of the three-axis orthogonal coordinate system and the center point O ′ is corrected. That is, from the magnetic vector OM detected by the three-axis magnetic sensor 2 and the correction vector OO ′, a vector O′M starting from the center point O ′ is calculated using the following equation.
O'M = OM-OO '(2)
Then, using the corrected vector O′M, an accurate orientation toward the portable device 10 is calculated and displayed on the display unit 11 (see FIG. 2).

  Next, a flowchart of the magnetic data processing apparatus 1 will be described with reference to FIGS. In this example, as shown in FIG. 10, first, the detected value of the magnetic vector M (n) by the triaxial magnetic sensor 2 is acquired (step S1). In step S2, it is determined whether the detected value is stored first or whether | M (n) −M (n−1) |> a is satisfied. Here, M (n−1) is the detection value of the geomagnetic vector stored last in step S3 to be described later at the time when the detection value of M (n) is acquired, and is not necessarily the detection value in the previous step. Is not limited. A is a predetermined constant. In step S2, if the end point of the newly detected magnetic vector M (n) is too close to the end point of the magnetic vector M (n-1) stored in step S3 described later, the circle C is not stored. It is provided for storing only when it is far enough that it can be determined that it is necessary to newly store it in order to obtain it. This is because even if a large number of detected values having almost the same end point are stored, they are not useful for later calculations. If YES is determined in step S2, the process proceeds to step S3, and the detected value and detection time of the geomagnetic vector M (n) are stored. Then, it returns to step S1. If it is determined No in step S2, the magnetic vector M (n) is not stored, and the process proceeds to step S4 for determining whether the rotation is finished.

  In step S4, it is determined whether or not the difference between the detection time of the detection value last stored in step S3 and the latest detection time is equal to or greater than a predetermined value. It is determined that the posture change in the middle posture change period has ended. If it is less than the predetermined value, it is determined that the posture change has not ended, and the process returns to step S1.

After step S5, the process proceeds to step S6. In step S6, it is determined whether or not the last stored geomagnetic vector M (n) and the first detected geomagnetic vector M (1) satisfy the following equation.
| M (n) -M (1) |> ε (3)
The above formula (3) is a formula for tentatively determining whether or not the detected posture change amount in the posture change period is large enough to obtain the rotation axis A with high accuracy. To make an accurate determination, it is necessary to determine the radius R of the circle C and then determine the rotation angle, which is provided here for provisional determination. However, if the circle C is far away from the center point O ′ of the azimuth s and the radius R is small and | M (n) −M (1) | Since the rotation axis A cannot be obtained well, the above formula (3) is useful in that the rotation axis A is not obtained in such a case.

  Further, ε in Equation (3) is a predetermined constant. Step S6 is provided to determine whether or not the last stored magnetic vector M (n) is sufficiently away from the first stored magnetic vector M (1). If it is determined Yes in step S6, the process proceeds to step S7. If it is determined No, it is determined that the rotation axis A cannot be calculated with high accuracy, and the detected data is erased (step S11), and then the data is again stored. Get it again. By repeating steps S1 to S6, a plurality of detection values (detection value group d) are stored.

  In step S7, a circle C is calculated from the individual detection values included in the detection value group d using the least square method. That is, a center point and a radius at which the variation in distance from each detected value is minimized are obtained, and a circle C in which the sum of squares of the distance from the plane including the circle C to each detected value is minimized (see FIG. 5). Is calculated. Also, a plane P including this circle C is calculated.

After step S7 is processed, step S8 is performed. Here, the variation of the distance from the individual detection values to the plane P x (see FIG. 6) σ (x) determines whether more than a predetermined value epsilon _x. If YES is determined in step S8, the process proceeds to step S9. When it is determined No in step S8, the stored detection value group d is not obtained while rotating the attitude of the mobile device 10 around one axis. Therefore, if it is determined No in step S8, the process proceeds to step S11, and the stored detection value group d is deleted. Then, returning to step S1, the detection value group d is stored again.

In step S9, the circle C variation in the distance r from the center B (see FIG. 7) to the individual detection values sigma (r) determines whether more than a predetermined value epsilon _r. If YES is determined in step S9, the process proceeds to step S10. When it is determined No in step S9, the stored detection value group d is not obtained while rotating the posture of the mobile device 10 around one axis. Therefore, when it is determined No in step S9, the process proceeds to step S11, and the stored detection value group d is deleted. Then, returning to step S1, the detection value group d is stored again.

In step S10, a circle radius R of the C (see FIG. 7) is equal to or less than the predetermined value epsilon _R. If it is determined Yes in step S10, the process proceeds to step S12. When it is determined No in step S10, the stored detection value group d is highly likely to be affected by the external magnetic field. Therefore, if it is determined No in step S10, the process proceeds to step S11, and the stored detection value group d is deleted. Then, returning to step S1, the detection value group d is stored again.

In step S12, the rotation angle θ (see FIG. 8) of the posture of the mobile device 10 in one posture change period is calculated. After processing step S12, the process proceeds to step S13. In step S13, the rotation angle theta is equal to or greater than a predetermined value epsilon _theta. If it is determined Yes in step S13, the process proceeds to step S14. The case where No is determined in step S13 is a case where the posture of the mobile device 10 is not sufficiently rotated during the posture change period. Therefore, if it is determined No in step S13, the process returns to step S11, and the stored detection value group d is deleted. Then, returning to step S1, the detection value group d is stored again.

Steps S8, S9, S10, S12, and S13 may be switched in order. The ε _x is about 1.3 μT, for example, and ε _r is about 0.45 μT, for example. Further, the epsilon _R is, for example, about 52MyuT, the epsilon _theta, for example, about 60 °.

  In step S14, a straight line passing through the center B of the circle C and orthogonal to the plane P is calculated as the rotation axis A. Thereafter, the process proceeds to step S15. In step S15, it is determined whether or not a plurality of rotation axes A have been calculated. If it is determined YES, the process proceeds to step S16. If NO is determined in step S15, the process returns to step S1, stores another detection value group d, and calculates another rotation axis A. Here, the number of rotation axes A to be obtained is at least two, and may be three or more. Also, the number setting can be changed according to the situation at the time of offset calculation.

  In step S16, a combination of the two rotation axes A having the largest intersection angle φ is selected from the plurality of rotation axes A that have already been obtained, and it is determined whether or not the intersection angle φ is greater than a predetermined value c. . This step S16 is a step in which the center point O ′ is not calculated in this case because the calculation accuracy of the center point O ′ is low even when the two rotation axes A having the intersection angle φ smaller than the predetermined value C are used. is there. If it is determined Yes in step S16, the process proceeds to step S17. If it is determined No, the process returns to step S1. In this embodiment, the two rotation axes having the largest intersection angle φ are selected. If the rotation axes are larger than the predetermined value C, the two rotation axes A other than the combination having the largest intersection angle φ are selected. Of course it is also possible. That is, it is possible to select a plurality of combinations of two rotation axes each having an intersection angle equal to or greater than a predetermined value.

  Since the rotation axis A includes an error, the two rotation axes A may not intersect at step S16. In this case, one rotation axis A is translated and the other rotation axis A is translated. The crossing angle φ in the state of crossing with is calculated.

  In step S17, the intersection of the two rotation axes A selected in step S16 is calculated. Here, when the two rotation axes A do not intersect, an approximate intersection is calculated. The approximate intersection can be defined as the midpoint of the closest points of the two rotation axes A, for example. When a plurality of combinations of two rotation axes A are selected, an offset is obtained by calculating an average value of approximate intersections obtained from each of the two combinations.

  After step S17 is processed, the process proceeds to step S18. In step S18, the calculated intersection or approximate intersection is set as the center point O 'of the azimuth sphere s.

  After processing step S18, the process proceeds to step S19. Here, the geomagnetic vector OM (see FIG. 9) detected by the three-axis magnetic sensor 2 is corrected, and a vector O′M starting from the center point O ′ is calculated.

The effect of this example will be described. The magnetic data processing apparatus 1 of this example calculates the rotation axis A of the portable device 10 using the detection value group d as shown in FIG. Then, using this rotation axis A, the center point O ′ of the azimuth sphere s is calculated.
If it does in this way, it will become easy to reduce a user's burden. That is, according to the inventors' research, when the user uses the mobile device 10 naturally, for example, while the mobile device 10 is in the pocket, the user can change the direction during walking, take out from the pocket, When the screen is brought to a position where it can be seen easily, the operation is often an operation of rotating the posture of the mobile device 10 around one axis. Therefore, the detection value group d is often obtained while rotating the posture of the mobile device 10 around one axis. Therefore, if the rotation axis A is calculated using the detected value group d, and the center point O ′ of the azimuth sphere s is calculated using the rotation axis A, the user's natural motion can be expressed as the center point O of the azimuth s. It becomes easy to use for the calculation of '.

  In the present invention, the following three points are important. The first is that although the rotation operation is a uniaxial rotation, the rotation angle uses a natural operation, and most of the rotation is, for example, about 60 to 90 ° and less than 180 °. Secondly, in the rotation of about 60 to 90 °, there is a possibility that the four points satisfying the selection condition cannot be selected in the invention of Patent Document 1 described above. In the invention of Patent Document 2, the detected value is an azimuth s Since there is no change in the existence at the above biased position, an accurate offset cannot be obtained. Third, although it is difficult to calculate an accurate offset with the conventional method as described above, it is possible to determine the rotation axis with a certain degree of accuracy. Therefore, the offset calculation by calculating the approximate intersection is compared with the conventional calculation method. It is a new finding that an improvement in accuracy can be expected.

However, in order to enable accurate offset calculation, it is necessary to calculate the rotation axis after accurately checking whether the detected value is obtained while rotating around one axis. Therefore, in the present invention, the check is performed as follows. Therefore, accurate offset calculation can be performed even if only biased data on the orientation sphere s is obtained on the premise of an accurate check.
That is, as shown in FIGS. 1 and 11, in the magnetic data processing device 1 of this example, whether or not the detected value group d is obtained while rotating the attitude of the mobile device 10 around one axis. Rotational motion determination means 4 for determining whether or not. Only when it is determined that the detected value group d is obtained while the attitude of the mobile device 10 is rotated around one axis, the detected value group d is set to the center point of the azimuth s. Used to calculate O ′. Therefore, the center point O ′ of the orientation sphere s can be calculated with high accuracy.

Further, as shown in FIGS. 5 to 7, the rotational motion determination means 4 of this example determines whether or not a plurality of detection values included in one detection value group d gather on or around the circle C. By determining, it is determined whether or not the detected value group d is obtained while rotating the posture of the mobile device 10 around one axis.
In this way, it is possible to easily determine whether or not the detected value group d is obtained while rotating the posture of the mobile device 10 around one axis. That is, when the attitude of the mobile device 10 is rotated around one axis, the magnetic vector M draws a circle C on the azimuth s. Therefore, by detecting whether or not the detection values of the magnetic vector M are gathered on the circumference of the circle C or in the vicinity thereof, while the detection value group d rotates the attitude of the mobile device 10 around one axis. It can be easily determined whether or not it has been acquired.

Further, as shown in FIGS. 5 to 7, the rotational motion determination unit 4 of this example calculates an equation of the circle C from each detection value included in one detection value group d using the least square method, It is determined whether or not a plurality of detection values are gathered on or around the calculated circle C.
In this way, since the least square method is used, the circle C can be calculated with high accuracy. Therefore, it can be accurately determined whether or not the detected value group d is acquired while rotating the posture of the mobile device 10 around one axis. Further, since a straight line passing through the center of the circle C perpendicular to the plane P including the circle C is the rotation axis A, the rotation axis A can be accurately calculated by accurately obtaining the circle C by the least square method. Therefore, the center point O ′ of the azimuth sphere s can be accurately calculated.

Further, the rotational motion determination means 4 of the present example has a variation in the distance x from the plane P (see FIG. 6) to each detected value that is not more than a predetermined value, and from the center B of the circle C (see FIG. 7). When the variation in the distance r to each detection value is equal to or less than a predetermined value, it is determined that the detection value group d is obtained while the portable device 10 is rotated about one axis.
In this way, it is possible to accurately determine whether or not the detected value group d is obtained while rotating the attitude of the mobile device 10 around one axis. That is, when the detection value group d is obtained while the portable device 10 is rotated about one axis, all the detection values included in the detection value group d are present near the plane P. And exists near the circle C. Therefore, variations in the distance x from the plane P to the individual detection values and variations in the distance r from the center B of the circle C to the individual detection values are reduced. Further, when the detection value group d is not obtained while rotating the mobile device 10 around one axis, the individual detection values included in the detection value group d exist near the plane P or the circle C. The variation becomes larger. Therefore, it is accurately determined whether or not the detected value group d is obtained while rotating the mobile device 10 around one axis by determining whether or not the variation is larger than a predetermined value. It becomes possible.

Further, as shown in FIG. 11, the rotation axis calculation means 5 of this example is configured to calculate the rotation axis A only when the radius R of the circle C is equal to or less than a predetermined value.
In this way, the center point O ′ of the azimuth sphere s can be accurately obtained. That is, when the radius R of the circle C is too large, the triaxial magnetic sensor 2 may be affected by the external magnetic field, and the center point O ′ of the azimuth sphere s may not be accurately calculated. Therefore, if the rotation axis A is calculated only when the radius R of the circle C is equal to or less than the predetermined value, the influence of the external magnetic field can be eliminated, and the center point O ′ of the azimuth sphere s can be accurately obtained.

Further, as shown in FIG. 12, the rotation axis calculation means 5 of this example calculates the rotation axis A only when the rotation angle θ of the posture of the mobile device 10 in one posture change period is equal to or larger than a predetermined value. It is configured.
As described above, the present invention is an invention that uses a posture change of a mobile device that is naturally performed without the user's awareness, and it is rare that the rotation angle is 180 ° or more as in the case of forcing a rotation operation. . However, in order to ensure high calculation accuracy of the rotation axis A, it is necessary to ensure a certain degree of rotation angle, and therefore the rotation axis A is calculated only when the above condition is satisfied.
In this way, the calculation accuracy of the center point O ′ of the azimuth sphere s can be further increased. That is, when the rotation angle θ is small, it is difficult to accurately calculate the circle C, and thus it is difficult to accurately calculate the rotation axis A. Therefore, the calculation accuracy of the center point O ′ of the azimuth sphere s tends to be lowered. Therefore, if the rotation axis A is calculated only when the rotation angle θ is greater than or equal to a predetermined value, the calculation accuracy of the center point O ′ can be increased.

  As shown in FIG. 4, in the present invention, the above condition is checked, and only when the condition is satisfied, the combination of the two rotation axes A1 and A2 is calculated, and the intersection of the two rotation axes A1 and A2 is calculated. Alternatively, the approximate intersection is calculated as the center point O ′ of the orientation sphere s. When these two rotation axes A intersect, the intersection is the center point of the azimuth sphere s. Further, since the calculated two rotation axes A include an error, when the plurality of rotation axes A do not intersect with each other, the approximate intersection is calculated as the center point O ′ of the azimuth sphere s. Since the intersection is obtained on the condition that the above condition is satisfied, the offset cannot be accurately obtained by the conventional method in a state where only the detected value biased on the azimuth sphere s is obtained. Even if it exists, an offset can be calculated | required with sufficient precision by application of this invention.

Next, the effect of the present invention will be described using experimental examples.
4 points used for offset calculation from the detection value of the magnetic sensor as described in the portable device A equipped with the offset calculation software that satisfies the conditions of the present invention, and the above-mentioned Patent Document 1 with a portable device equipped with a three-axis magnetic sensor The portable device B equipped with software for selecting the detected value and calculating the offset was prepared.
Then, for each of the portable device A and the portable device B, the rotation operation of horizontal rotation 75 ° and pitch rotation 75 ° is intentionally performed a plurality of times, the offset is obtained a plurality of times, and the direction is calculated using the obtained offset value. The variability of the orientation calculation was investigated.

  As a result, it was confirmed that the portable device A can determine the orientation with an accuracy of ± 5 °. On the other hand, in the portable device B, the rotation operation is small, and the condition for selecting the detection of four points is not satisfied, and thus the direction cannot be calculated.

  Therefore, for the portable device B, the horizontal rotation and pitch rotation angles were set to 120 °, and the rotation operation was intentionally performed a plurality of times again to obtain the offset a plurality of times, and the variation in the direction calculation was investigated. As a result, under this condition, it was possible to calculate the offset and calculate the azimuth, but the variation was ± 10 °, despite the fact that the angle of the rotation operation was increased. Only the calculation accuracy which does not reach the experimental result of the portable device A was obtained. The software installed in the mobile device B can be set so as to satisfy the selection condition even at a smaller rotation angle by changing the condition for selecting the four points, but the accuracy is improved as the selection condition is relaxed. In view of the fact that it is clearly reduced, it can be said that the present invention has the effect of greatly improving the accuracy of calculating the offset value using only magnetic data obtained by natural rotation operation.

DESCRIPTION OF SYMBOLS 1 Magnetic data processor 2 3-axis magnetic sensor 3 Memory 4 Rotation motion calculation means 5 Rotation axis calculation means 6 Center point calculation means 7 Offset correction means A Rotation axis B Circle center C Circle M Magnetic vector O ′ Center point of azimuth sphere s Azimuth sphere

Claims (7)

A three-axis magnetic sensor for detecting geomagnetism as a magnetic vector in a three-axis orthogonal coordinate system fixed to the portable device, and calculating a center point of an azimuth sphere drawn by the magnetic vector with a change in the posture of the portable device; A magnetic data processing device for correcting the influence of the internal magnetic field of the portable device on the three-axis magnetic sensor,
A memory for storing a plurality of detected values of the magnetic vector obtained during a posture change period in which the posture of the portable device is changing;
Rotation motion determination that determines whether or not a detection value group consisting of a plurality of detection values of the magnetic vector in one posture change period is obtained while rotating the posture of the portable device around one axis Means,
Rotation for calculating the rotation axis of the rotation when the rotation value determination means determines that the detected value group is obtained while rotating the attitude of the portable device around one axis. An axis calculation means;
From the detected value groups acquired during a plurality of different posture change periods, the intersections or approximate intersections of the plurality of rotation axes obtained by the rotational motion determination means and the rotation axis calculation means respectively are the center points of the azimuth sphere. A center point calculating means for calculating as
Offset correcting means for correcting an offset which is a deviation between the origin of the three-axis orthogonal coordinate system and the center point of the azimuth sphere;
A magnetic data processing apparatus comprising:   2. The magnetic data processing apparatus according to claim 1, wherein the rotational motion determination means determines whether or not a plurality of the detection values included in one detection value group are gathered on or around a circle. Thus, it is determined whether or not the detected value group is obtained while rotating the posture of the portable device around one axis.   3. The magnetic data processing device according to claim 2, wherein the rotational motion determination unit calculates an equation of the circle by a least square method from each of the detection values included in one detection value group, and A magnetic data processing apparatus for determining whether or not the detected values are gathered on or around the calculated circle.   4. The magnetic data processing device according to claim 2, wherein the rotational motion determination means has a variation in distance from a plane including the circle to each of the detected values not more than a predetermined value, and the center of the circle. Determining that the detected value group is obtained while rotating the portable device around one axis when the variation in distance from the detected value to each of the detected values is equal to or less than a predetermined value. A magnetic data processing device.   5. The magnetic data processing device according to claim 2, wherein the rotation axis calculation unit is configured to calculate the rotation axis only when a radius of the circle is a predetermined value or less. A magnetic data processing apparatus characterized by comprising:   6. The magnetic data processing device according to claim 2, wherein the rotation axis calculation unit is configured such that the rotation angle of the posture of the portable device in one posture change period is a predetermined value or more. The magnetic data processing apparatus is configured to calculate the rotational axis only for the first to second.   The magnetic data processing device according to any one of claims 1 to 6, wherein the center point calculating unit is configured to calculate the intersection point or the intersection point when the intersection angle of the plurality of rotation axes is larger than a predetermined value. A magnetic data processing apparatus configured to calculate the approximate intersection.

Images