JP2013202066A - Motion analysis device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a motion analysis device capable of detecting that an object is practically at the same position before and after a motion, acquiring the start point and end point of the motion, and acquiring a reference position being the start point and the end point in a motion space.SOLUTION: A motion analysis device 1 includes: an inertial sensor 100 attached to a measurement object 10; an imaging part 40 for imaging the measurement object; a start point acquisition part 202 for acquiring a time of a standstill state of the measurement object as a start point t1 of the motion of the measurement object; an end point acquisition part 203 for acquiring an end point t2 of the motion of the measurement object on the basis of a difference between a reference image imaged in the imaging part at the start point and a comparison image imaged in the imaging part after the motion; and a data acquisition part 204 for acquiring an output of the inertial sensor at least from the start point to the end point.

Description

  The present invention relates to a motion analysis device that analyzes a motion of a measurement object from a start point to an end point.

  There is a need for an apparatus for analyzing the motion of an object to be examined in various fields. For example, a tennis racket or a swing trajectory of a golf club, an exercise form such as baseball pitching or batting, etc. can be analyzed, and an improvement point can be clarified from the analysis result, thereby improving the competitiveness.

  At present, an optical motion capture device is known as a practical motion analysis device. This device generally analyzes the motion by continuously photographing the measurement target with the marker attached thereto using an infrared camera or the like, and calculating the movement trajectory of the marker using the photographed continuous images (patent) Reference 1).

  On the other hand, in recent years, a device has been proposed in which a small inertia sensor is attached to an object to be inspected, and the motion of the object to be inspected is analyzed from output data of an inertia sensor that detects a physical quantity to be measured (Patent Document 2). . This apparatus has the advantage that it is easy to handle because it does not require an infrared camera.

JP 2010-110382 A JP 2008-073210 A

  In motion sensing using an inertial sensor such as a gyro sensor or an acceleration sensor, alignment during measurement is required. This is because the inertial sensor measures the amount of change such as angular velocity and acceleration using the inertia of the measurement object and does not have absolute position information. Even if position information can be detected by integrating the output signal of the inertial sensor, position information with high accuracy cannot be acquired due to an integration error or the like. For this reason, the position information indicating at which position the measured information is acquired becomes important information in the motion analysis of the measurement target.

  Therefore, for example, it is conceivable to acquire the positions before the start of exercise and after the end of exercise from information other than the inertial sensor.

  At that time, if the position at the start of exercise and the position at the end of exercise are respectively set to positions specified in advance, the degree of freedom of movement of the measurement target is limited, and in particular, the subject who operates the measurement target or the test subject itself The burden of.

  Some aspects of the present invention detect that the object is substantially in the same position before and after the movement, acquire the starting point and the ending point of the movement, and acquire the reference position as the starting point and the ending point in the movement space. It is an object of the present invention to provide a motion analysis apparatus that can perform the above.

(1) One aspect of the present invention is
An inertial sensor attached to the measurement object;
An imaging unit for imaging the measurement object;
A start point acquisition unit for acquiring the time when the measurement target is stationary as a start point of the movement of the measurement target;
An end point acquisition unit that acquires an end point of the movement of the measurement target based on a difference between a reference image captured by the imaging unit at the start point and a comparative image captured by the imaging unit after the exercise; ,
A data acquisition unit for acquiring an output of the sensor between at least the start point and the end point;
The present invention relates to a motion analysis apparatus having

  In one embodiment of the present invention, an imaging unit that images a measurement target is provided in addition to an inertial sensor that detects a physical quantity of the measurement target. The start point acquisition unit acquires the time when the measurement target is in a stationary state as the start point of the movement of the measurement target. The fact that the measurement object is stationary may be acquired based on an external signal that informs that it is before an instruction for instructing the start of exercise, for example, or a difference in imaging information from the imaging unit or You may acquire from a coincidence degree. The end point of the exercise is acquired from the difference or coincidence of the imaging information from the imaging unit, and the difference or coincidence between the reference image at the start point and the comparison image after the start point is determined. The data acquisition unit acquires the output of the inertial sensor from at least the start point to the end point based on information from the start point acquisition unit and the end point acquisition unit. Thus, the start point and end point of the motion can be acquired, and information on the substantially same position at the start point and the end point can be used as the reference position in the motion space.

  (2) In one aspect of the present invention, the imaging unit can image a region including the inertial sensor attached to the measurement target at a start position of the movement of the measurement target.

  When an exercise apparatus such as a golf club tennis racket is a measurement target, a start position of an exercise, for example, a swing exercise can be determined in advance. If the region including the inertial sensor at the start position is a region of interest, the inertial sensor can be a feature point in the captured image. Therefore, an inertial sensor always exists in the reference image, and the difference or coincidence of images can be determined based on the presence / absence and position of the inertial sensor in the comparative image.

(3) In one aspect of the present invention, the end point acquisition unit includes:
A first storage unit that stores the reference image from the imaging unit;
A second storage unit for storing the comparison image from the imaging unit;
The reference image read from the first storage unit and the comparison image read from the second storage unit are input, and each difference between two pixel values on the same address is acquired. A subtraction unit;
A comparison unit for comparing each difference output from the subtraction unit with a threshold value;
A determination unit that determines the end point based on the number of addresses equal to or greater than the threshold;
Can have.

  In this way, the determination unit determines the difference or coincidence between the reference image and the comparison image based on the number of addresses where the difference on the same address is equal to or greater than a threshold value, and the number of addresses is less than a predetermined value. Since the degree of coincidence increases sometimes (difference decreases), it can be determined that the end point is reached. Note that the number of addresses to be counted is not limited to an address greater than or equal to a threshold value, and may be an address less than the threshold value. Regardless of which is counted, since the degree of coincidence increases (difference decreases) when the number of addresses equal to or greater than the threshold value is smaller than the predetermined value, it can be determined that the end point is reached.

  (4) In one aspect of the present invention, the subtracting unit can acquire each difference between two pixel values on the same address in a part of the region of interest to be measured.

  It is possible to determine the difference or coincidence between the reference image and the comparison image by acquiring each difference between each two pixel values on the same address in the region of interest without detecting the difference for all addresses in the image. High-speed processing is achieved.

  (5) In one aspect of the present invention, the start point acquisition unit includes the first storage unit, the second storage unit, the subtraction unit, the comparison unit, and the determination unit provided in the end point acquisition unit. Can also be used. That is, the images from the imaging unit before the start point acquisition are alternately stored in the first storage unit and the second storage unit. The subtraction unit receives the images read from the first storage unit and the second storage unit, and acquires each difference between the two pixel values on the same address. The comparison unit compares each difference output from the subtraction unit with a threshold value, and the determination unit can determine the starting point based on the number of addresses equal to or greater than the threshold value. Thereby, for example, the subject who operates the measurement target or the test subject who is the measurement target itself can freely determine the stationary position within the region of interest captured by the imaging unit.

  (6) In one mode of the present invention, it can further have a start point notification part which notifies the start of movement based on acquisition of the start point in the start point acquisition part. Thereby, the subject who operates the measuring object or the subject who is the measuring object itself can check the start time of the exercise.

  (7) In one mode of the present invention, it can further have an end point notification part which notifies the end of exercise based on acquisition of the end point in the end point acquisition part. Thereby, the subject who operates the measuring object or the subject who is the measuring object itself can check the end time of the exercise.

  (8) In one aspect of the present invention, the data acquisition unit can add a code to outputs corresponding to the start point and the end point among outputs from the inertial sensor. As a result, the physical quantity output from the inertial sensor is associated with the start point and the end point, and the physical quantity of the measurement target corresponding to the start point and the end point is acquired at the same position in real time. Can do. Further, by correcting the position information obtained by second-order integration of the output of an inertial sensor, for example, an acceleration sensor, based on the position information of the start point and the end point, the position of the measurement target during exercise can be corrected.

It is a figure showing the outline of the motion analysis device concerning the embodiment of the present invention. It is a figure which shows the golf club which is a measuring object, and the charger which charges the sensor unit attached to the golf club. FIG. 3 is a block diagram of the sensor unit shown in FIGS. 1 and 2. It is a figure which shows the charge state which connected the golf club and the charger with the contact. It is a timing chart which shows the sensor output signal and imaging frame sampled in the 1st-3rd period. 1 is a block diagram of a motion analysis apparatus according to an embodiment of the present invention. It is a block diagram of the end point acquisition part (start point acquisition part) which concerns on embodiment of this invention. It is a figure which shows the difference image between two images with a low coincidence degree (a big difference). It is a difference image of the region of interest in FIG. It is a figure which shows the difference image between two images with a high coincidence degree (small difference). It is a difference image of the region of interest in FIG. It is a flowchart which shows the motion analysis method which concerns on embodiment of this invention. It is a figure which shows an example of the data structure in which the start point / end point was encoded. It is explanatory drawing of the linear approximation method which calculates | requires a bias estimated value for a starting point determination.

  Hereinafter, preferred embodiments of the present invention will be described in detail. The present embodiment described below does not unduly limit the contents of the present invention described in the claims, and all the configurations described in the present embodiment are indispensable as means for solving the present invention. Not necessarily.

1. Overview of Motion Analysis Device As shown in FIG. 1, the motion analysis information collection device 100 includes a sensor unit 20 including an inertial sensor attached to a measurement target, for example, a golf club 10, an imaging unit 40, and a host terminal 200. Have

1.1. Sensor Unit Here, the target of the motion analysis information of the motion analysis device 1 is the golf club 10 operated by the subject shown in FIG. FIG. 1 shows a swing locus A of a club head 11 of a golf club 10 operated by a subject. The swing locus A includes a swing activation position (address position) P1, a top position P2, an impact position P3, and a follow-through top position P4. Note that FIG. 1 shows a swing trajectory of swinging, and the swing trajectory position P1 (address position) and the impact position P3 are shown apart from each other. In the swing for hitting the golf ball, the positions P1 and P3 are close to each other at the position where the golf ball is teeed up.

  FIG. 2 shows a golf club 10 to which the sensor unit 20 used in the present embodiment is mounted, and a holder for the golf club 10, for example, a charger 30. The charger 30 charges the secondary battery in the sensor unit 20 that supplies power to the sensor unit 20.

  FIG. 2 schematically shows the golf club 10 in which the sensor unit 20 is mounted on the club shaft 12, for example. The charger 30 is a stand type, and can hold the club shaft 12 in a stationary state, and can charge a secondary battery in the sensor unit 20 via a contact described later. The position where the sensor unit 20 is attached is not limited to the club shaft 12 but may be the club head 11 or the like, for example. Further, the sensor unit 20 is not necessarily limited to the one attached to the golf club 10 and may be attached to the subject. In this case, the subject is the measurement target.

  The sensor unit 20 can detect a given physical quantity using the inertia of the golf club 10 and output a signal (data) corresponding to the magnitude of the detected physical quantity (for example, acceleration, angular velocity, etc.). In the present embodiment, the sensor unit 20 includes an inertial sensor 100 as shown in FIG. The inertial sensor 100 includes, for example, three-axis acceleration sensors 102x to 102z (an example of an inertial sensor) that detect acceleration in the X-axis, Y-axis, and Z-axis directions, and three that detect angular velocities in the X-axis, Y-axis, and Z-axis directions. A 6-axis motion sensor including 104 g to 104 z is provided with an axis gyro sensor (another example of an angular velocity sensor and an inertial sensor).

  Here, the triaxial acceleration sensors 102x to 102z detect, for example, the amount and direction of the inertial movement of the weight part configured by MEMS (Micro Electro Mechanical Systems) with acceleration based on a change in capacitance value. It is possible to detect the acceleration in three axis directions. As the three-axis gyro sensors (angular velocity sensors) 104x to 104z, vibration angular velocity sensors can be used. The vibration type angular velocity sensor vibrates the vibrating body at a constant frequency. When an angular velocity is applied to the vibrating body, a Coriolis force is generated, and the vibrating body vibrates in different directions due to the Coriolis force. The angular velocity can be detected by detecting the displacement due to the Coriolis force.

  The control unit 110 illustrated in FIG. 3 is connected to the inertial sensor 100, the communication unit 120, and the secondary battery 130. The control unit 110 can be further connected to the charger 30.

  The control unit 110 includes a data processing unit 110A that outputs the output signals of the sensors 102x to 102z and 104x to 104z together with time information, for example, as a packet and outputs the packet to the communication unit 120. The communication unit 118 illustrated in FIG. 3 performs processing for transmitting the packet data received from the data processing unit 110A to the host terminal 300. The control unit 110 can include a charge control unit 110B that controls charging of the secondary battery 130.

  FIG. 4 shows a basic configuration example of the charger 30 and the exercise apparatus (golf club) 10 to be measured shown in FIG. The charger 30 functioning as a holder includes a grounding portion 31, a shaft holding portion 32 extending upward from the grounding portion 31, a charging circuit 33 provided in the grounding portion 31, and 2 provided in the shaft holding portion 32. Two charging terminals 34 and 35 are provided. The golf club 10 has to-be-charged terminals 13 and 14 in contact with the charging terminals 34 and 35 of the charger 30 on the club shaft 12. The sensor unit 20 provided in the golf club 10 is not shown in FIG.

1.2. Imaging Unit An imaging unit 40 illustrated in FIG. 1 is an imaging camera that images the golf club 10. The imaging unit 40 is configured so that when the golf club 10 is set at the swing path position (address position) P1 in the swing path A shown in FIG. Can be set.

  As shown in FIG. 5, from the start time t0, for example, the sensor unit 30 of the golf club 10 is imaged by the imaging unit 40 at a frame frequency of a predetermined period P1. The start time t0 can be, for example, the time when the golf club 10 is detached from the charger 30. Similarly, data detected by the sensor unit 20 is sampled at a sampling frequency of a predetermined period P2 from the start time t0 in FIG. The sampling frequency of the sensor output signal may or may not match the frame frequency.

  Here, the time t1 shown in FIG. 5 is set as the starting point of the movement of the golf club 10, and the time t2 is set as the ending point of the movement of the golf club 10. A second period T2 from the start point t1 to the end point t2 is an exercise period of the golf club 10. A period from the measurement start time t0 to the start point t1 is a first period T1, and a period from the end point t2 to the measurement end point t3 is a third period T3. The sensor unit 20 and the imaging unit 40 measure over a measurement period T (= T1 + T2 + T3). However, simultaneously with the detection of the end point t2, the sensor unit 20 and the imaging unit 40 may measure the power of a cow. Further, at least at the start point t1 and the end point t2, the golf club 10 is in a stationary state, and in this embodiment, the swing activation position (address position) P1 in FIG. 1 is the start point and the end point. In other words, the subject rests the golf club 10 at the swing activation position (address position) P1 (start point), swings (exercise period), and again rests at the swing activation position (address position) P1 (end point). .

1.3. Host Terminal FIG. 6 shows the host terminal 200. The host terminal 200 includes a processing unit (CPU) 201, a communication unit 210, an operation unit 220, a ROM 230, a RAM 240, a nonvolatile memory 250, a display unit 260, and a notification unit 270.

  The communication unit 210 performs processing of receiving signals transmitted from the sensor unit 20 and the imaging unit 40 by wire or wirelessly and sending them to the processing unit 201. The operation unit 220 performs a process of acquiring operation data from the user and sending it to the processing unit 201. The ROM 230 stores programs for the processing unit 201 to perform various calculation processes and control processes, various programs and data for realizing application functions, and the like. The RAM 240 is used as a work area of the processing unit 201, and temporarily stores programs and data read from the ROM 230, data input from the operation unit 220, calculation results executed by the processing unit 201 according to various programs, and the like. It is a storage unit. In the present embodiment, in particular, known data about the initial position P0 and the impact position P3 shown in FIG. 7 to be described later can be stored in the ROM 230 or the RAM 240. The non-volatile memory 250 is a storage unit that records data that needs to be stored for a long time among the data generated by the processing of the processing unit 201. The display unit 260 displays the processing result of the processing unit 201 as characters, graphs, or other images. The display unit 260 is, for example, a CRT, LCD, touch panel display, HMD (head mounted display), or the like. In addition, you may make it implement | achieve the function of the operation part 220 and the display part 260 with one touchscreen type display. The notification unit 270 notifies the subject of the start and / or end of exercise.

  The processing unit 201 performs various calculation processes on the data received from the sensor unit 20 via the communication unit 210 and various control processes (display control for the display unit 260, etc.) in accordance with a program stored in the ROM 230. In particular, in the present embodiment, the processing unit 201 includes a start point acquisition unit 202, an end point acquisition unit 203, a data acquisition unit 204, a calculation unit 205, and a data correction unit 206. The starting point acquisition unit 202 acquires the time when the golf club 10 is in a stationary state at the swing activation position (address position) P1 in FIG. 1 as the starting point t1 (FIG. 5) of the golf club 10. The end point acquisition unit 203 determines that the golf club 10 is again based on the difference or the degree of coincidence between the reference image captured by the imaging unit 40 at the start point t1 and the comparison image captured by the imaging unit 40 after the start point t1. When it is returned to the swing activation position (address position) P1 of FIG. 1 and is in a stationary state, the movement end point t2 (FIG. 5) of the golf club 10 is acquired. The data acquisition unit 204 acquires the outputs of the inertial sensors 102x to 102z and 104x to 104z at least from the start point t1 to the end point t2. The computing unit 205 obtains motion analysis data such as the speed, angle, and position of the golf club 10 by integrating the data (acceleration, angular velocity) obtained by the data obtaining unit 204 with n (n is a natural number). The data correction unit 206 corrects the motion analysis data, which is the calculation result of the calculation unit 205, based on known information such as acceleration, angular velocity, and speed at the stationary position (all the above physical quantities are zero).

  The data acquisition unit 204 further linearly approximates the output signal acquired from the angular velocity sensors 104x to 104z during a period including, for example, the start point t1 when the golf club 10 is in a stationary state, for example, by the least square method, and outputs the output signal 104x of the angular velocity sensor. The bias value included in ˜104z that changes with time can be acquired as a bias estimated value that is a function of time. Instead of this, the data acquisition unit 204 may calculate the average of the output signals acquired from the angular velocity sensors 104x to 104z that are stationary at the start point and the end point to obtain the estimated bias value during the exercise period. The data correction unit 206 removes the bias estimation value corresponding to the output signal and the time axis from the output signal acquired from the angular velocity sensors 104x to 104z during the exercise period following the stationary period of the golf club 10, and generates correction data. Can be generated.

1.4. End point acquisition unit (start point acquisition unit)
FIG. 7 is a block diagram illustrating an example of the end point acquisition unit 203. The end point acquisition unit 203 can also be used as the start point acquisition unit 202. However, the start point acquisition unit 202 is not limited to the one that also uses the configuration of the end point acquisition unit 203.

  In the present embodiment, the start point acquisition unit 202 and the end point acquisition unit 203 share the configuration shown in FIG. The first storage unit 203A stores an image from the imaging unit 40. The second storage unit 203B stores other images from the imaging unit 40. The first and second storage sections 203A and 203B can be different areas of the RAM 240 shown in FIG. The subtraction unit 203C receives the image read from the first storage unit 203A and the image read from the second storage unit 203B, and acquires each difference between the two pixel values on the same address. To do. The comparison unit 203D compares each difference output from the subtraction unit 203C with a threshold value. The output of the comparison unit 203D can be set to, for example, “0” if the output is less than the threshold, and can be set to “1” if the output is greater than or equal to the threshold. The determination unit 203E determines the end point or the start point based on the number of addresses equal to or greater than the threshold value.

  In order to determine the end point by the determination unit 203E, the reference image captured by the imaging unit 40 at the start point t1 in FIG. 5 is stored in the first storage unit 202A, and in the second period T2 in FIG. The comparison image captured at 40 is stored in the second storage unit 202B.

  In order to determine the starting point by the determining unit 203E, the imaging unit 40 is stored in the first storage unit 202A and the second storage unit 202B in the first period T1 before the acquisition of the starting point t1 shown in FIG. Images from are stored alternately.

  8 to 11 show binary difference images that are output from the comparison unit 203D. FIG. 8 illustrates a difference image that is determined by the determination unit 203E to have a low degree of coincidence (a large difference). 8 to 11, in the difference image, the output from the comparison unit 203D is “0”, that is, the pixel having a high degree of coincidence between pixels (small difference) is displayed in black, and the output from the comparison unit 203D is A pixel that is “1”, that is, a pixel with a low degree of coincidence between pixels (a large difference) is displayed in white.

  The difference image shown in FIG. 8 has a high degree of coincidence between the images of the subject, but the images of the golf club 10, the sensor unit 20, and the arm of the subject do not coincide. The determination unit counts the number of white display pixels (the number of addresses on the storage unit) that are equal to or greater than the threshold value in the difference image illustrated in FIG. 8. Judged as inconsistent.

  Here, the subtraction unit 203 </ b> C is not limited to obtaining the difference between each two pixels on the same address of the entire image. The subtraction unit 203C may obtain each difference between each two pixel values on the same address in the region of interest that is a part of the entire image.

  FIG. 9 shows a difference image for a part of the region of interest in FIG. The region of interest shown in FIG. 9 is an imaging region including the sensor unit 20 attached to the golf club 10 when the club head 12 is positioned at the swing path position (address position) P1 shown in FIG. Since the number of white display pixels in the total number of pixels in the difference image in FIG. 9 is relatively large, the determination system in the determination unit 203E is high.

  On the other hand, FIG. 10 shows a difference image determined by the determination unit 203E as having a high degree of matching (difference is small). The difference image shown in FIG. 10 has a high degree of coincidence between the images of the subject and the degree of coincidence between the images of the golf club 10, the sensor unit 20, and the arm of the subject. The determination unit counts the number of white display pixels (the number of addresses on the storage unit) that are equal to or greater than the threshold value among the difference images illustrated in FIG. 10. It can be determined that they match.

  FIG. 11 shows a difference image for the same region of interest as FIG. 9 in the image of FIG. That is, the region of interest shown in FIG. 11 is an imaging region including the sensor unit 20 attached to the golf club 10 when the club head 11 is positioned at the swing trajectory position (address position) P1 shown in FIG.

  In order to determine the end point by the determination unit 203E, the reference image captured by the imaging unit 40 at the start point t1 in FIG. 5 is stored in the first storage unit 202A, and in the second period T2 in FIG. The comparison image captured at 40 is stored in the second storage unit 202B. Therefore, as shown in FIG. 10 or FIG. 11, when the degree of coincidence between the reference image and the comparative image increases (the difference decreases), the subject moves the golf club 10 to the swing activation position (address position) P1 in FIG. This means that the image is stopped (start point) and then stopped again at the swing activation position (address position) P1. Therefore, the end point t2 in FIG. 5 can be determined when the degree of coincidence between the reference image and the comparative image is high (the difference is small).

  On the other hand, in order to determine the start point by the determination unit 203E, the first storage unit 202A and the second storage unit 202B are imaged in the first period T1 before the acquisition of the start point t1 shown in FIG. Images from the unit 40 are stored alternately. Therefore, as shown in FIG. 10 or FIG. 11, the degree of coincidence between the reference image and the comparative image is high (the difference is small). This means that the subject moves the golf club 10 to the charger in FIG. This means that the robot is released from 30 and moved to the swing activation position (address position) P1 in FIG. Therefore, the starting point t1 in FIG. 5 can be determined by increasing the degree of coincidence between the two images (decreasing the difference).

2. Motion Analysis Method FIG. 12 is a flowchart showing the motion analysis method according to the embodiment of the present invention. In step S1 of FIG. 12, it is monitored whether or not the golf club 10 is detached from the charger 30 of FIG. Whether or not the golf club 10 has been detached from the charger 30 can be determined by the charging control unit 110B in FIG.

  If it is acquired that the golf club 10 has been removed from the charger 30 (YES in step S1 in FIG. 12), the process proceeds to step S2. In step S2, from the start time t0 in FIG. 5, the imaging unit 40 starts imaging the subject, the golf club 10, and the sensor unit 20 at a frame frequency of a predetermined period P1. Further, a sensor unit signal from the sensor unit 20 is acquired.

  FIG. 13 shows an example of data acquired by the data acquisition unit 204 of FIG. By executing step S2 in FIG. 12, first output signals X0, Y0, Z0 are obtained as sensor output signals X, Y, Z from the sensors 102x-102z and 104x-104z shown in FIG. The output signal is transmitted to the host terminal 200 at the sampling period P2 of 5. On the other hand, the first imaging frame signal F0 is obtained as the imaging frame signal F from the imaging unit 40, and thereafter, the output signal is transmitted to the host terminal 200 in the frame period P1 of FIG. In FIG. 13, for example, the frame period P1 and the sampling period P2 coincide.

  Next, in step S3 of FIG. 12, it is determined whether or not the start point has been acquired. As described above, the test subject removes the golf club 10 from the charger 30 in FIG. 2 at time t0 in FIG. 5, moves it to the swing activation position (address position) P1 in FIG. Accordingly, the difference between the two images (for example, FIG. 13) captured by the imaging unit 40 during the first period T1 in FIG. 5 and alternately stored in the first and second storage units 203A and 203B shown in FIG. The determination unit 203E determines that the degree of coincidence of F0-F1) shown in FIG. As a result, the start point acquisition unit 202 in FIG. 6 can acquire the start point. Note that “0” in the “determination” column shown in FIG. 13 indicates that the degree of matching is low, and “1” indicates that the degree of matching is high. The determination unit 203E can set, for example, the acquisition time of an image having a first high coincidence after time t0 as the start point t1. Then, as shown in FIG. 13, a symbol is attached in association with the sensor output signal corresponding to the start point t <b> 1, for example, “1” is attached as the start point / end point flag.

  When the start point is acquired by the start point acquisition unit 202 (YES in step S3 in FIG. 12), the start of exercise can be notified from the notification unit 270 in FIG. 6 (step S4 in FIG. 12).

Next, in step S5 of FIG. 12, it is determined whether or not an end point has been acquired. As described above, the subject stops the golf club 10 at the swing activation position (address position) P1 in FIG. 1 (starting point), and then swings along the swing locus A shown in FIG. (Address position) Still at P1. The reference image captured by the imaging unit 40 at the start point t1 of the second period T2 in FIG. 5 is stored in the first storage unit 203A illustrated in FIG. Images are sequentially updated in the second storage unit 203B. Then, it is determined to the determination unit 203E that the degree of coincidence of the difference between the two images stored in the first and second storage units 203A and 203B (for example, F _n−1 −F _n shown in FIG. 13) has increased. Is determined. As a result, the end point acquisition unit 203 in FIG. 6 can acquire the end point. The determination unit 203E can set, for example, the acquisition time of the image having the first high coincidence after time t1 as the end point t2. Then, as shown in FIG. 13, a code is attached in association with the sensor output signal corresponding to the end point t <b> 2, for example, “1” is added as the start point / end point flag.

  When the end point is acquired by the end point acquisition unit 203 (YES in step S5 in FIG. 12), the end of the exercise can be notified from the notification unit 270 in FIG. 6 (step S6 in FIG. 12). And measurement can be complete | finished at the time t3 of FIG. 5 (step S7 of FIG. 12).

3. Although the present embodiment has been described in detail as described above, it will be easily understood by those skilled in the art that many modifications can be made without substantially departing from the novel matters and effects of the present invention. Accordingly, all such modifications are intended to be included in the scope of the present invention. For example, the measurement target object of the present invention can be preferably applied to an exercise apparatus such as a golf club or a tennis racket, but is not limited thereto.

  In the above embodiment, the start point acquisition unit 202 shares the configuration of the end point acquisition unit 203 illustrated in FIG. 7, but is not limited thereto. For example, when the notification unit 270 in FIG. 6 instructs the start of exercise when a certain time has elapsed after the golf club 10 is detached from the charger 30, the notification unit 270 indicates a signal indicating that it is before the instruction. Based on this, the starting point t1 may be acquired. In this case, it is only necessary for the subject to be informed in advance that the camera stops at the swing activation position (address position) P1 before the start of exercise is notified.

  Another example in which the start point acquisition unit 202 acquires the start point t1 will be described with reference to FIG. As described above, the data acquisition unit 204 in FIG. 6 can acquire the bias estimation value based on the output signals from the angular velocity sensors 104x to 104x in a stationary state before the start of exercise. When the estimated bias value is confirmed by the data acquisition unit 204, the start point acquisition unit 102 may acquire the start point t1.

  The data acquisition unit 204 including a bias estimation unit linearly approximates data sampled on the time axis as output signals from the three-axis gyro sensors 104x to 104z, for example, on a two-dimensional plane including the time axis, and obtains a bias estimation value. Can be sought.

  FIG. 14 shows a technique for obtaining a bias estimated value by linear approximation. In FIG. 14, the horizontal axis represents time t, and the vertical axis represents the angular velocity as an angle per second (dps). FIG. 14 shows that the angular velocities y1 to yn are acquired at the time of sampling n times from time x1 to xn.

The data acquisition unit 204 converts n data (x1, y1), (x2, y2),... (Xn, yn) sampled on the time axis as output signals from the three-axis gyro sensors 104x to 104z, Y = a × x + b, which is linearly approximated by a two-dimensional plane (xy) including the axis, is obtained as a bias estimation value using time as a function. In other words, since the output signals y1 to yn from the three-axis gyro sensors 104x to 104z in the stationary state indicate the bias values corresponding to the times x1 to xn, the above equation expresses the bias estimated value as Vias (t ) And t is the time, the bias estimated value Vias (t), which is the angular velocity per second (dps), is
Vias (t) = a × t + b (1)
It can be expressed as. The estimated bias value Vias (t) is obtained for each of the X, Y, and Z axes.

  Here, the coefficients a and b in Expression (1) are obtained as follows in the least square method based on n pieces of data (x1, y1), (x2, y2),... (Xn, yn). .

  According to the inventor's experiment, if a bias estimation value is obtained by securing a stationary period set before the exercise period for at least 2 seconds, an appropriate bias estimation value can be obtained without setting the stationary period after the exercise period. Was found to be required.

  The length of the stationary period (bias estimation time) set before the exercise period depends on the bias stability of the three-axis gyro sensors 104x to 104z. As the triaxial gyro sensors 104x to 104z, as described above, for example, vibration-type angular velocity sensors with high bias stability in which angular velocity sensor elements are sealed under reduced pressure can be used.

  Here, the length of the stationary period (bias estimation time) set before the exercise period depends on the bias stability of the three-axis gyro sensors 104x to 104z, etc., or depends on external factors, and is therefore variable. May be.

  According to the experiment of the present inventor, the data collection period (stationary period) shown in FIG. 14 is continued until the slope a in the expression (1) satisfies a ≦ 0.010, and the above inequality is satisfied. Sometimes the bias estimate of equation (1) can be determined. According to the experiment, the estimated bias value obtained in the stationary period of 1 second is not adopted, and the estimated bias value obtained by setting the stationary period to 2 seconds becomes the definite value. Alternatively, according to the experiment of the present inventor, the data collection period (stationary period) shown in FIG. 14 is continued until the slope a in the expression (1) satisfies a ≦ 0.005, and the above inequality is satisfied. When done, the bias estimate of equation (1) can be determined. According to the experiment, the estimated bias value obtained in each stationary period of 1 second and 2 seconds is not adopted, and the estimated bias value obtained by setting the stationary period to 3 seconds is a definite value.

  As described above, the bias estimation value is evaluated by continuing the first period T1 shown in FIG. 5 until the bias estimation value is determined. For example, the magnitude of the slope a in the equation (1) is set as the threshold value. It can be evaluated in comparison. By doing in this way, the reliability of a bias estimated value can be improved.

  In this case, the data correction unit 206 in FIG. 6 obtains a bias estimation value Vias (t) corresponding to the output signal and the time axis from the output signals acquired from the three-axis gyro sensors 104x to 104z in the second period T2. Remove. The estimated bias value Vias (t) corresponding to the output signal on the time axis is obtained by substituting the time t at which the output signal was sampled into the equation (1). At this time, if the estimated bias value as a function of time is a linear function as shown in Equation (1), the data correction unit 202 can easily and quickly calculate the bias value divided from the angular velocity sensor output. it can.

  As described above, since the data acquisition unit 204 includes the bias estimation unit described above, the start point acquisition unit 202 can acquire the time when the bias estimation value is determined as the start point t1, and the data correction unit 206 performs bias correction. be able to.

  DESCRIPTION OF SYMBOLS 1 Motion analysis apparatus, 10 Measurement object, 20 Sensor unit, 30 Charger (holding tool), 40 Imaging part, 100, 102x-102z, 104x-104z Inertial sensor, 200 Host terminal, 201 Processing part, 202 Start point acquisition part, 203 end point acquisition unit, 203A first storage unit, 203B second storage unit, 203C subtraction unit, 203D comparison unit, 203E determination unit, 204 data acquisition unit, 205 calculation unit, 206 data correction unit, 270 notification unit (start point) Notification section, end point notification section), T1 first period, T2 second period, T3 third period

Claims (8)

An inertial sensor attached to the measurement object;
An imaging unit for imaging the measurement object;
A start point acquisition unit for acquiring the time when the measurement target is stationary as a start point of the movement of the measurement target;
An end point acquisition unit that acquires an end point of the movement of the measurement target based on a difference between a reference image captured by the imaging unit at the start point and a comparative image captured by the imaging unit after the exercise When,
A data acquisition unit for acquiring an output of the inertial sensor between at least the start point and the end point;
A motion analysis apparatus comprising: In claim 1,
The motion analysis apparatus characterized in that the imaging unit images a region including the inertial sensor attached to the measurement target at a movement start position of the measurement target. In claim 1 or 2,
The end point acquisition unit
A first storage unit that stores the reference image from the imaging unit;
A second storage unit for storing the comparison image from the imaging unit;
The reference image read from the first storage unit and the comparison image read from the second storage unit are input, and each difference between two pixel values on the same address is acquired. A subtraction unit;
A comparison unit for comparing each difference output from the subtraction unit with a threshold value;
A determination unit that determines the end point based on the number of addresses equal to or greater than the threshold;
A motion analysis apparatus comprising: In claim 3,
The subtraction unit acquires each difference between two pixel values on the same address in a part of the region of interest to be measured. In claim 3 or 4,
The start point acquisition unit is also used as the first storage unit, the second storage unit, the subtraction unit, the comparison unit, and the determination unit provided in the end point acquisition unit,
Images from the imaging unit before the start point acquisition are alternately stored in the first storage unit and the second storage unit,
The motion analysis apparatus, wherein the determination unit determines the start point based on the number of addresses equal to or greater than the threshold obtained for a difference image between two images before the start point is acquired. In claim 5,
A motion analysis apparatus further comprising a start point notifying unit for notifying start of exercise based on the acquisition of the start point by the start point acquiring unit. In any one of Claims 1 thru | or 6,
The motion analysis apparatus further comprising an end point notifying unit for notifying the end of the exercise based on the acquisition of the end point by the end point acquiring unit. In any one of Claims 1 thru | or 7,
The data acquisition unit adds a code to the output corresponding to the start point and the end point among outputs from the inertial sensor.

Images