JP2019097732A - Exercise performance evaluation device, exercise performance evaluation method and exercise performance evaluation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique capable of accurately determining a skill level. An exercise performance evaluation device of the present invention estimates the exercise performance of a subject in a batting exercise utilizing the rotational movement of the body. The exercise performance evaluation device of the present invention includes an exercise information acquisition unit and a performance estimation unit. The exercise information acquisition unit acquires time-series information of the rotational movement of the subject's body. The performance estimation unit estimates the motor performance of the subject from the rotational movement time-series information acquired by the motion information acquisition unit based on the relationship between the rotational movement time-series information and the evaluation of the motor performance. [Selection diagram] Fig. 6

Description

  The present invention relates to an exercise performance evaluation apparatus, exercise performance evaluation method, and exercise performance evaluation program for estimating the exercise performance of a subject in batting exercise using rotational movement of the body.

  In order to properly acquire the basic motions of sports, Non-Patent Document 1, as a technology that enables a target person (actor) to objectively grasp the difference between his or her physical motion and the physical motion of a skilled person, Non-patent document 2 is known. In Non-Patent Document 1, it is shown that it is possible to distinguish the experienced person and the unexperienced person based on the difference between the peak time of the acceleration at the butt tip and the impact time. Specifically, for each of a plurality of persons, the difference between the peak time of the acceleration at the tip of the bat and the impact time (the time when the ball contacts the bat) in each of a plurality of trials is determined. It is disclosed that the distribution of the average of the inexperienced person and the distribution of the average of the skilled person can be grouped when the mean value and the variance of the difference are determined, and that the skilled person tends to have a smaller variance than the beginner. It is done. Non-Patent Document 2 discloses that the maximum value of the acceleration of the expert tends to be larger than the maximum value of the acceleration of the beginner when the maximum value of the waist acceleration in the batting motion of baseball is determined. There is.

Akihiro Okuda, Koji Okamoto, Keiji Namba: "Three-dimensional analysis of batting motion of baseball using high-speed video camera", The 56th Annual Meeting of Applied Mechanics (2007). Katsumi Yamaguchi, Takeshi Murakami, Koichi Nakayama: "Development of a Batting Skill Acquisition Support System for Baseball Using an Acceleration Sensor", 2013 Annual Conference of Student Research Presentations (2013).

  The ideal batting action is performed by stepping on the foot and rotating the body. People who are not used to batting can not turn their body in a timely manner, and tend to try batting only with arm movements. Non-Patent Document 1 attempts to determine the skill level of the subject based on the acceleration of the bat tip. In this case, it is determined whether batting is performed only by arm movement or whether the body is properly used. It is difficult to accurately determine the proficiency level.

  Non-Patent Document 2 discloses that there is a correlation between the magnitude of the maximum value of acceleration of the waist and the skill level, focusing on the waist that produces rotational movement of the body. However, it is difficult to properly determine the level of skill in situations where the waist is usable but the ball does not hit the bat.

  In view of the above problems, it is an object of the present invention to provide a technique capable of accurately determining the skill level.

  The exercise performance evaluation device of the present invention estimates the exercise performance of the subject in batting motion utilizing the rotational motion of the body. The exercise performance evaluation device of the present invention includes an exercise information acquisition unit and a performance estimation unit. The exercise information acquisition unit acquires time-series information (information related to time and size) of rotational motion of the subject's body. The performance estimation unit estimates the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit, based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance.

  According to the exercise performance evaluation device of the present invention, since the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, it is possible to accurately determine the skill level.

The figure which shows the change of the angular velocity of the waist based on the time of the pitcher of a softball releasing a ball. The figure which shows the change of the angular velocity of the waist on the basis of impact time. The figure which shows the change of the angular velocity of the waist on the basis of the impact time when knowing a ball type. The figure which shows the change of the angular velocity of the waist on the basis of the impact time when not knowing a ball type. The chart shows the average peak time based on the impact time, with top and young players, and with and without knowledge of the ball type. FIG. 1 is a diagram showing an example of the configuration of an exercise performance evaluation apparatus according to a first embodiment. The figure which shows the process flow example of the exercise information collection of the exercise | movement performance evaluation apparatus of Example 1, 5. The figure which shows the processing flow of evaluation of the exercise | movement performance evaluation apparatus of Examples 1-4. FIG. 7 is a diagram showing an example of the configuration of an exercise performance evaluation apparatus according to a second embodiment. The figure which shows the process flow example of the exercise information collection of the exercise | movement performance evaluation apparatus of Example 2, 5. FIG. 7 is a diagram showing an example of the configuration of an exercise performance evaluation apparatus according to a third embodiment. The figure which shows the process flow example of the exercise | movement information collection of the exercise | movement performance evaluation apparatus of Example 3, 5. FIG. 16 is a diagram showing an example of the configuration of an exercise performance evaluation apparatus according to a fourth embodiment. The figure which shows the process flow example of the exercise | movement information collection of the exercise | movement performance evaluation apparatus of Example 4,5. FIG. 18 is a diagram showing an example of the configuration of an exercise performance evaluation apparatus according to a fifth embodiment. FIG. 18 is a diagram showing a processing flow of learning of the exercise performance evaluation apparatus of the fifth embodiment. FIG. 18 is a diagram showing a processing flow of evaluation of the exercise performance evaluation device of the fifth embodiment.

  Hereinafter, embodiments of the present invention will be described in detail. Note that components having the same function will be assigned the same reference numerals and redundant description will be omitted.

<Analysis to be the basis of the present invention>
FIG. 1 shows the change in the angular velocity of the waist relative to when the softball pitcher releases the ball. FIG. 1 (A) shows the case of the top player, and FIG. 1 (B) shows the case of the young player. The ball type was straight (fast ball) and change-up, and the player was measured without knowing the ball type. The horizontal axis is time (seconds) and time 0 is when the pitcher releases the ball. The vertical axis is the waist angular velocity (degree / second). When the solid line is straight, the dotted line indicates the change-up case, and one line indicates one trial. In the case of the top player, it can be seen that the peak time, which is the time at which the magnitude of the angular velocity based on the time of release is maximum, differs depending on the ball type. On the other hand, it can be seen that young players have no clear difference in peak time depending on the ball type. That is, it can be seen that the highly skilled player adjusts the timing to rotate the hips according to the ball type.

  FIG. 2 shows the change in the angular velocity of the hip based on the impact time (the time when the ball contacts the bat). FIG. 2 shows the result of FIG. 1 normalized so that the impact time is 0 seconds, FIG. 2 (A) is for the top player, and FIG. 2 (B) is for the young player. . From FIG. 2, top players concentrate (bias) in a time section where there is a difference between the peak time, which is the time when the hip angular velocity is maximum, and the impact time, regardless of the ball type. It can be seen that the difference with the time varies.

  FIG. 3 shows the change in the angular velocity of the hip based on the impact time when the ball type is known, and FIG. 4 shows the change in the angular velocity of the hip based on the impact time when the ball type is unknown. Similar to FIG. 2, these figures are obtained by normalizing changes in the hip angular velocity in each trial, with the impact time as zero. FIG. 5 is a graph in which the average value of peak times based on the impact time is graphed with and without knowing the top player, the young player, and the ball type. FIGS. 3 (A), 4 (B), and 5 (A) are for top players, and FIGS. 3 (B), 4 (B), and 5 (B) are for young players.

  From this result, the top athlete concentrates (biases) on the time section where there is a difference between the peak time and the impact time regardless of whether he knows the ball type, but when young players do not know the ball type and Even if the ball type is known, in the case of change-up, it can be said that the difference between the peak time and the impact time tends to be large.

It is as follows in summary.
(Trend 1) The difference in the average ball speed (peak speed) of the peak angular velocity of the waist relative to the ball release time when the top player tried multiple times was larger than that of the young players (Tend 2) The top player is less biased than the young players in the distribution of peak time of the hip angular velocity based on the impact time when trying multiple times than the young players (trend 3) The top players are better than the young players Distribution of peak time of hip angular velocity with reference to impact time when batting for up is small (Trend 4) Younger athletes are based on impact time when they know the ball type than top players There is a large difference between the distribution of peak time of the hip angular velocity and the distribution of peak time of the hip angular velocity based on the impact time when you do not know the ball type. The player can properly move his body against the impact. In other words, the difference between the peak time and the impact time (time width) can be made uniform to a fixed time width regardless of the type of ball.

  The present invention acquires biological information by an inertial sensor attached to a subject based on the above findings, and performs exercise performance (skill level) based on the relationship between time series information of rotational movement and evaluation of exercise performance. It is something to evaluate. Alternatively, even without using an inertial sensor, a reflection marker or the like may be attached on the rotation axis of the target, and calculation may be made from position information obtained from motion capture or video, or human motion such as a Kinect is identified. The existing video information that can be used may be used. In the above analysis, batting against a flying ball is analyzed, but the present invention can also be applied to batting movements using rotational movements of other bodies (such as the waist and body). For example, even in tennis stroke, toss batting, tee batting, golf, etc., exercise performance (skill level) can be evaluated based on the relationship between time series information on rotational movement and evaluation of exercise performance.

<Specific configuration>
FIG. 6 shows a configuration example of the exercise performance evaluation apparatus of the first embodiment, FIG. 7 shows a process flow example of exercise information collection of the exercise performance evaluation apparatus, and FIG. 8 shows a process flow of evaluation of the exercise performance evaluation apparatus. The exercise performance evaluation device 200 estimates the exercise performance of the subject in batting motion using the rotational motion of the body. The exercise performance evaluation device 200 includes an exercise information acquisition unit 210, a performance estimation unit 220, a characteristic information acquisition unit 230, and a recording unit 290. Further, the reference time detection unit 260 may be provided as needed.

  The exercise information acquisition unit 210 acquires time series information (information on time and size) of rotational motion of the subject's body, and records peak time, which is the time at which the magnitude of rotational motion of the body is maximum. (S210). More specifically, the exercise information acquisition unit 210 acquires time-series information of rotational movement of the subject's body (S211). The exercise information acquisition unit 210 may record the acquired time series information of the rotational movement in the recording unit 290. The magnitude of the rotational movement is, for example, the speed of rotation (angular velocity) or acceleration. In short, any information that can acquire an index corresponding to the size (or speed) of the rotational movement may be used. Such information can be acquired by the biological sensor attached to the subject. For example, when acquiring angular acceleration, it is an inertial sensor or the like. As for the attachment position of the sensor, although the sensor is attached to the subject's waist in the above-mentioned experiment, it may be any place where it is possible to obtain an index corresponding to the size of the rotational movement of the body. It is not limited. In addition, as described above, even if a biometric sensor is not attached to a subject, a reflection marker or the like may be attached on the rotational axis of the subject, and calculation may be performed from position information obtained from motion capture or video. It is possible to use existing video information that can identify such human motion.

  The exercise information acquisition unit 210 further detects a peak time that is a time at which the magnitude of the rotational movement is maximum (S212), and records the peak time in the recording unit 290 (S213). For example, the exercise information acquisition unit 210 acquires information on the size of the rotational movement for each time interval. In this case, the motion information acquisition unit 210 indicates that the information on the magnitude of the rotational motion in the acquired time interval t indicates that the rotational motion is larger than the predetermined threshold, and the magnitude of the rotational motion in the time interval t is The time when the magnitude of the rotational movement in the last time interval t-1 is smaller is detected (SS212), and the peak time may be recorded in the recording unit 290 (S213). Although the maximum value is acquired by this, when there are a plurality of peak times in one trial, the peak time at which the magnitude of the rotational movement becomes maximum is recorded in the recording unit 290 as the final peak time (S213) . That is, the exercise information acquisition unit 210 records, as the peak time, a time near the time when the magnitude of the rotational movement is maximized. The threshold may be updated as appropriate. Step S210 is a collection of steps S211 to S213. Note that the performance estimation unit 220 described later may perform steps S212 and S213. In this case, steps S212 and S213 may be included in the exercise performance estimation (S220) and executed.

  The characteristic information acquisition unit 230 acquires the movement characteristic of the flying object, and records it in the recording unit 290 (S230). The characteristic information acquisition unit 230 acquires, for example, information on the ball type of the ball in each trial (in the above example, batting) of the target person, and records the information in the recording unit 290. The movement characteristic may input the result determined by image analysis etc., and may be input by hand. The “exercise characteristic” is a characteristic that can uniquely identify the type, for example, straight and change-up in the case of softball, straight, slider, curve, etc. in the case of baseball. It may be a difference in the speed of a simple flight. For example, the first motion characteristic may be determined as straight and the second motion characteristic as change-up, and the characteristic information acquisition unit 230 may acquire information indicating which movement characteristic the ball has.

  The reference time detection unit 260 detects a reference time, which is a time of a predetermined timing before the arrival of a flying object in each trial of the object person, and records the reference time in the recording unit 290 (S260). The "predetermined timing" is, for example, the timing at which the pitcher releases the ball. However, the timing is not limited to the release timing. For example, when the ball passes a position 10 m away from the home base may be set as the predetermined timing. More specifically, the reference time detection unit 260 acquires reference time information (S261). For example, video information of the position of the ball and the position of the pitcher's hand is repeatedly acquired. The reference time detection unit 260 confirms whether or not the state to be the reference time has been detected (S262). For example, check if the ball has been released from the pitcher's hand. When the state to be the reference time is detected, the reference time detection unit 260 records the reference time in the recording unit 290 (S263).

  When evaluating the exercise performance by defining “peak time” as the time when the magnitude of the rotational movement based on the predetermined timing before the arrival of the flying objects is maximum, the peak acquired by the exercise information acquisition unit 210 The reference time may be subtracted from the time to obtain the defined “peak time”. However, if the sensor is activated according to the ball release timing and acquisition of exercise information is started, the start time becomes the reference time, so the peak time acquired by the exercise information acquisition unit 210 It can be used as "peak time" of the above definition as it is. Therefore, in this case, the reference time detection unit 260 is not necessary.

  The exercise performance evaluation device 200 confirms whether the trial necessary for the evaluation of exercise performance has been performed (whether the repetition condition is satisfied) (S205). If not satisfied, the trial is repeated and the process is ended if satisfied.

  The performance estimation unit 220 estimates the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance (S220) ). More specifically, the performance estimation unit 220 calculates the average of peak times in the striking motion for the flying object having the first motion characteristic and the average of the peak times in the striking motion for the flying object having the second motion characteristic. The first evaluation value is output as the evaluation result of the exercise performance when the magnitude of the difference is the first value. Then, the performance estimation unit 220 determines the difference between the average of peak times in the striking motion with respect to the flying object having the first motion characteristic and the average of the peak times in the striking motion with respect to the flying object having the second motion characteristic. Is a second value greater than the first value, a second evaluation value corresponding to the higher exercise performance than the first evaluation value is output as an exercise performance estimation result. For example, in relation to the performance estimation unit 220 of the present embodiment, “the average of peak times in the striking motion for the flying object having the first motion characteristic and the peak in the striking motion for the flying object having the second motion characteristic are The larger the difference from the average time, the higher the evaluation of exercise performance. This association is based on the trend 1 shown in “Analysis underlying the present invention”. A threshold is defined for the difference between the average of peak times in the striking motion for the airborne object having the first motion characteristic and the average of the peak times in the impacting motion for the airborne object having the second motion characteristic For example, if two levels of evaluation and four threshold values are determined, five levels of evaluation are possible. Note that, the smaller the difference between the average of peak times in the striking motion for the flying object having the first motion characteristic and the average of the peak times in the striking motion for the flying object having the second movement characteristic, the more the exercise performance is The relation of “low evaluation” is equivalent to the relation described above.

  To describe the process of the performance estimation unit 220 in more detail, first, the performance estimation unit 220 calculates an average value of peak times of rotational movement for each movement characteristic of the flying object. In addition, when performing step S212 and S213 by the performance estimation part 220, rotation of the body for every time section acquired by the exercise information acquisition part 210 about each of the several times of trials for every exercise | movement characteristic is said before the said process. The peak time for each trial is obtained by detecting the time (peak time) at which the magnitude of the rotational movement takes a peak value after the reference time (for example, release time) from the information on the size of the movement.

  Next, the performance estimation unit 220 obtains and outputs an estimation result of performance based on the average value of peak times for each motion characteristic. In the case of the two-stage evaluation, if the absolute value of the difference between the average of peak times for the first motion characteristic and the average of peak times for the second motion characteristic is equal to or greater than a predetermined threshold, exercise performance Outputs an evaluation result indicating that the skill level is high (the skill level is high), and the absolute value of the difference between the average value of peak times for the first motion characteristic and the average value of peak times for the second motion characteristic is predetermined If it is smaller than the threshold of, an evaluation result indicating that exercise performance is low (skill level is low) is output.

  According to the exercise performance evaluation device 200, the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, so that the skill level can be determined with high accuracy. Here, the first exercise characteristic and the second exercise characteristic can be easily evaluated by combining two kinds of exercise characteristics, such as straight and change-up, which cause a clear difference in peak time in a top athlete with high exercise performance.

  In the above description, the batting motion in a real environment is described, but motion information acquired from a batting motion performed in a VR environment may be used. For example, the subject mounts a head mount display, acquires exercise information when actually performing an exercise to repel a virtual ball thrown by a pitcher on a VR under a virtual environment, and based on the exercise information Exercise performance may be assessed.

  In addition to peak time, it is also possible to replace with peak value.

[Modification 1]
The performance estimation unit 220 of the first embodiment is the difference between the average of peak times in the striking motion for the flying object having the first motion characteristic and the average of the peak times in the striking motion for the flying object having the second movement characteristic. However, an index indicating the degree of variation in peak time may be calculated for each movement characteristic, and an evaluation of movement performance for each movement characteristic of the flying object may be added. For example, standard deviation of differences for all trials, kurtosis, absolute value of difference between "maximum value of difference" and "minimum value of difference" (Mmax with maximum value of difference as Mmin, minimum value of difference as Mmin, Mmax Calculate -Mmin as a "index indicating the degree of variation". When the "index indicating the degree of variation" indicates that "the variation is large", it indicates that the exercise performance is low. This makes it possible to grasp the proficiency of each movement characteristic (ball type), such as high straight proficiency but low changeup proficiency (not good), and low movement proficiency (ball type) Can be used to create a training menu that should be trained intensively.

  FIG. 9 shows a configuration example of the exercise performance evaluation apparatus of the second embodiment, FIG. 10 shows a process flow example of exercise information collection of the exercise performance evaluation apparatus, and FIG. 8 shows a process flow of evaluation of the exercise performance evaluation apparatus. The exercise performance evaluation device 201 estimates the exercise performance of the subject in batting motion using the rotational motion of the body. The exercise performance evaluation device 201 includes an exercise information acquisition unit 210, a performance estimation unit 221, an impact time acquisition unit 270, and a recording unit 290. Further, the reference time detection unit 260 may be provided as needed.

  The exercise information acquisition unit 210 is the same as that of the first embodiment. In this embodiment, it is assumed that the peak time is a time at which the magnitude of the rotational movement is maximum with reference to a predetermined timing, and the impact time is a time at which an impact is given with reference to the predetermined timing. In this embodiment, an arbitrary time may be set as a predetermined timing (the arbitrary time may be set as time 0), the peak time and the impact time may be obtained, and a reference time detection unit 260 is also provided. The peak time and the impact time may be determined by setting the release time as 0).

  The impact time acquisition unit 270 acquires the impact time (S270). As the impact time, a result estimated from a sensor attached to a bat or the like may be used, an impact time may be estimated by image processing, or the impact time may be given manually. In the case of a sensor, for example, there is a method of using the time at which the pressure takes the maximum value by the pressure sensor as the impact time, or setting the time at which the square sum of the acceleration acquired by the inertia sensor takes the maximum value as the impact time. For example, the impact time acquisition unit 270 acquires impact information that is information (such as video information) for detecting an impact (S271). The impact time acquisition unit 270 confirms from the impact information whether it is an impact (S272), and in the case of an impact, records the time at that time as the impact time in the recording unit 290 (S273).

  The performance estimation unit 221 estimates the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit 210 based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance ( S221). More specifically, when the variation in the difference between the peak time and the impact time is a first value, the performance estimation unit 221 outputs the first evaluation value as the evaluation result of the exercise performance. Then, when the variation of the difference between the peak time and the impact time is a second value smaller than the first value, the performance estimation unit 221 corresponds to the second corresponding to the higher exercise performance than the first evaluation value. The evaluation value of is output as an estimation result of exercise performance. For example, in the relevance of the performance estimation unit 221 of the present embodiment, "the smaller the variation in the difference between the peak time and the impact time, the higher the evaluation of exercise performance is." This relevancy is the relevancy based on the trend 2 shown in the "analysis underlying the present invention". One threshold value for an index (for example, standard deviation, kurtosis, absolute value of difference between "maximum difference" and "minimum difference") indicating the degree of variation in difference between peak time and impact time If it is defined, evaluation of two stages is possible, and if four thresholds are defined, evaluation of five stages is possible. In addition, the relationship that "the evaluation of exercise performance is lower as the variation in the difference between the peak time and the impact time is larger" is equivalent to the above-described relationship.

  The process of the performance estimation unit 221 will be described in more detail. The performance estimation unit 221 calculates the difference between the impact time and the peak time for each trial. Then, an index indicating the degree of variation of the difference between the impact time and the peak time is calculated for all the trials. For example, standard deviation of differences for all trials, kurtosis, absolute value of difference between "maximum value of difference" and "minimum value of difference" (Mmax with maximum value of difference as Mmin, minimum value of difference as Mmin, Mmax Calculate -Mmin as a "index indicating the degree of variation". In the case where the performance estimation unit 221 performs steps S212 and S213, before the above process, the performance estimation unit 221 sets a time at which the size of the exercise takes a peak value after the release time (peak time) By detecting), we obtain the peak time for each trial.

  The performance estimation unit 221 estimates that the exercise performance is low when the “index indicating the degree of variation” indicates that “the variation is large”, and estimates that the exercise performance is high otherwise. For example, when the standard deviation is less than a predetermined threshold, an evaluation result indicating that exercise performance is high is output, and when the standard deviation is larger than the predetermined threshold, an evaluation result indicating that exercise performance is high is output. Alternatively, an evaluation result indicating that exercise performance is high is output if Mmax-Mmin is less than or equal to a predetermined threshold, and an evaluation result indicating that exercise performance is low is output if Mmax-Mmin is greater than the predetermined threshold.

  According to the exercise performance evaluation device 201, the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, so that the skill level can be determined with high accuracy. In addition, since the exercise performance evaluation apparatus 201 does not consider the movement characteristics of the object, it is possible to evaluate impact movements such as toss batting, tee batting, and golf. Furthermore, when the movement performance evaluation device 201 is used for a set of trials of the same movement characteristic after acquiring and recording the movement characteristics of the coming movement, the movement performance for the trial in which the movement characteristics of the flying object are limited When the evaluation device 201 is used, exercise performance with limited movement characteristics of flying objects can be evaluated. Also, by notifying the subject of the motion characteristics of all the flying objects and using the motion performance evaluation device 201, it is possible to evaluate the motion performance with knowledge of the motion characteristics, and if the motion characteristics are not notified, exercise Can assess exercise performance when you do not know the characteristics.

  In the above description, the batting motion in a real environment is described, but motion information acquired from a batting motion performed in a VR environment may be used. For example, the subject mounts a head mount display, acquires exercise information when actually performing an exercise to repel a virtual ball thrown by a pitcher on a VR under a virtual environment, and based on the exercise information Exercise performance may be assessed.

  FIG. 11 shows a configuration example of the exercise performance evaluation apparatus of the third embodiment, FIG. 12 shows a process flow example of exercise information collection of the exercise performance evaluation apparatus, and FIG. 8 shows a process flow of evaluation of the exercise performance evaluation apparatus. The exercise performance evaluation device 202 estimates the exercise performance of the subject in batting motion using the rotational motion of the body. The exercise performance evaluation device 202 includes an exercise information acquisition unit 210, a performance estimation unit 222, a characteristic information acquisition unit 230, an impact time acquisition unit 270, and a recording unit 290. Further, the reference time detection unit 260 may be provided as needed.

  The exercise information acquisition unit 210 and the characteristic information acquisition unit 230 are the same as in the first embodiment, and the impact time acquisition unit 270 is the same as the second embodiment. In this embodiment, it is assumed that the peak time is a time at which the magnitude of the rotational movement is maximum with reference to a predetermined timing, and the impact time is a time at which an impact is given with reference to the predetermined timing. In this embodiment, an arbitrary time may be set as a predetermined timing (the arbitrary time may be set as time 0), the peak time and the impact time may be obtained, and a reference time detection unit 260 is also provided. The peak time and the impact time may be determined by setting the release time as 0).

  The performance estimation unit 222 estimates the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit 210 based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance ( S222). More specifically, the performance estimation unit 222 determines the degree of variation of the difference between the peak time and the impact time at the time of the flying object having the first motion characteristic, and at the time of the flying object having the second motion characteristic. When the difference between the peak time and the impact time is the first value, the first evaluation value is output as the evaluation result of the exercise performance. Then, the performance estimation unit 222 determines the degree of variation of the difference between the peak time and the impact time at the time of the flying object having the first motion characteristic, and the peak time and the impact at the time of the flying object having the second motion characteristic. The second evaluation value corresponding to the fact that the exercise performance is higher than the first evaluation value when the difference between the time and the degree of variation of the difference is a second value smaller than the first value Output as the estimation result. For example, in the relevancy of the performance estimation unit 222 of the present embodiment, “the degree of variation of the difference between the peak time and the impact time at the time of the flying object having the first movement characteristic and the second movement characteristic The smaller the difference between the peak time and the impact time of the object, the higher the evaluation of exercise performance. This relevancy is the relevancy based on the trend 3 shown in the "analysis underlying the present invention". Index indicating the degree of variation of the difference between the peak time and the impact time for the first and second motion characteristics (eg, standard deviation, kurtosis, absolute value of difference between "maximum difference" and "minimum difference") For the difference in value), if one threshold is determined, two-stage evaluation can be performed, and if four thresholds are determined, five-stage evaluation can be performed. Note that “the degree of variation in the difference between the peak time and the impact time for an airborne object having the first movement characteristic, and the difference between the peak time and the impact time for an airborne object having the second movement characteristic” The relationship that the evaluation of exercise performance is lower as the difference with the degree of variation is larger is equivalent to the above-described relationship.

  The process of the performance estimation unit 222 will be described in more detail. The performance estimation unit 222 calculates the difference between the impact time and the peak time for each trial. Then, for each exercise characteristic, an index indicating the degree of variation of the difference between the impact time and the peak time is calculated. For example, standard deviation of differences for all trials, kurtosis, absolute value of difference between "maximum value of difference" and "minimum value of difference" (Mmax with maximum value of difference as Mmin, minimum value of difference as Mmin, Mmax Calculate -Mmin as a "index indicating the degree of variation". In the case where the performance estimation unit 222 performs steps S212 and S213, before the above process, the performance estimation unit 222 determines the time at which the size of the exercise takes a peak value after the release time (peak time) By detecting), we obtain the peak time for each trial.

  The performance estimation unit 222 estimates exercise performance based on the “index indicating the degree of variation” for each exercise characteristic. Specifically, if the absolute value of the difference between the “index indicating the degree of variation” for the first motion characteristic and the “index indicating the degree of variation” for the second motion characteristic is equal to or less than a predetermined threshold value An evaluation result indicating that exercise performance is high (skilled degree is high) is output, and the “index indicating the degree of variation” for the first exercise characteristic and the “index indicating the degree of variation” for the second exercise characteristic And the evaluation result indicating that the exercise performance is low (the skill level is low) is output if the absolute value of the difference between the and the is greater than the predetermined threshold.

  According to the exercise performance evaluation device 202, the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, so that the skill level can be determined with high accuracy.

  In the above description, the batting motion in a real environment is described, but motion information acquired from a batting motion performed in a VR environment may be used. For example, the subject mounts a head mount display, acquires exercise information when actually performing an exercise to repel a virtual ball thrown by a pitcher on a VR under a virtual environment, and based on the exercise information Exercise performance may be assessed.

  FIG. 13 shows a configuration example of the exercise performance evaluation apparatus of the fourth embodiment, FIG. 14 shows a process flow example of exercise information collection of the exercise performance evaluation apparatus, and FIG. 8 shows a process flow of evaluation of the exercise performance evaluation apparatus. The exercise performance evaluation device 203 estimates the exercise performance of the subject in batting motion using the rotational motion of the body. The exercise performance evaluation device 203 includes an exercise information acquisition unit 210, a performance estimation unit 223, a characteristic information acquisition unit 230, a characteristic presentation information acquisition unit 240, an impact time acquisition unit 270, and a recording unit 290. Further, the reference time detection unit 260 may be provided as needed.

  The exercise information acquisition unit 210 and the characteristic information acquisition unit 230 are the same as in the first embodiment, and the impact time acquisition unit 270 is the same as the second embodiment. In this embodiment, it is assumed that the peak time is a time at which the magnitude of the rotational movement is maximum with reference to a predetermined timing, and the impact time is a time at which an impact is given with reference to the predetermined timing. In this embodiment, an arbitrary time may be set as a predetermined timing (the arbitrary time may be set as time 0), the peak time and the impact time may be obtained, and a reference time detection unit 260 is also provided. The peak time and the impact time may be determined by setting the release time as 0).

  The characteristic presentation information acquisition unit 240 acquires information on whether or not the movement characteristic of the flying object has been presented (S240). The exercise performance evaluation device 203 may present the motion characteristic of the flying object to the target person, or the character or information may be presented to the target person as a sound or a signal, and then the characteristic presentation information acquisition unit 240 presents it. You may acquire the information of whether or not.

  The performance estimation unit 223 estimates the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit 210 based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance ( S223). More specifically, the performance estimation unit 223 knows the degree of variation of the difference between the peak time and the impact time when the target person knows the movement characteristics of the flying object, and the target person knows the movement characteristics of the flying object. The first evaluation value is output as the evaluation result of the exercise performance when the difference between the peak time and the impact time when there is no difference is the first value. Then, the performance estimation unit 223 determines the degree of variation of the difference between the peak time and the impact time when the target person knows the movement characteristics of the flying object, and the peak when the target person does not know the movement characteristics of the flying object When the difference between the time of day and the degree of difference in the difference between the impact times is a second value smaller than the first value, a second evaluation value corresponding to higher exercise performance than the first evaluation value is Output as an estimation result of exercise performance. For example, in the relevance of the performance estimation unit 223 of the present embodiment, “the degree of variation of the difference between the peak time and the impact time when the target person knows the motion characteristics of the The smaller the difference between the peak time and the impact time when you do not know exercise characteristics, the higher the evaluation of exercise performance. This relevancy is the relevancy based on the trend 4 shown in the "analysis underlying the present invention". Index indicating the degree of variation between the peak time and the impact time when the motion characteristics of the flying object are known and unknown (for example, standard deviation, kurtosis, "maximum difference" and "minimum difference" If one threshold value is set for the difference between the absolute values of the difference of the value “2”, it is possible to make two-stage evaluation, and if four threshold values are set, five-stage evaluation is possible. Note that “the degree of variation in the difference between the peak time and the impact time when the subject knows the motion characteristics of the flying object, and the peak time and the impact time when the subject does not know the motion characteristics of the flying object The relation that the evaluation of the exercise performance is lower as the difference with the degree of the difference of the difference is larger is equivalent to the above-mentioned relation.

  The process of the performance estimation unit 223 will be described in more detail. The performance estimation unit 223 is a first index, which is an index indicating the degree of variation of the difference between the impact time and the peak time when the target person knows the movement characteristic of the flying object, and the movement characteristic of the target person The second index, which is an index indicating the degree of variation of the difference between the impact time and the peak time when not knowing The “index indicating the degree of variation” is the same as in the third embodiment. In the case where the performance estimation unit 223 performs steps S212 and S213, before the above process, the performance estimation unit 223 determines the time at which the size of exercise reaches a peak value after the release time (peak time) By detecting), we obtain the peak time for each trial.

  The performance estimation unit 223 estimates exercise performance based on the first index and the second index. Specifically, when the absolute value of the difference between the first index and the second index is less than or equal to a predetermined threshold, an evaluation result indicating that the exercise performance is high (skill level is high) is output, and the first index When the absolute value of the difference between the second index and the second index is larger than a predetermined threshold, an evaluation result indicating that exercise performance is low (skill level is low) is output.

  According to the exercise performance evaluation device 203, the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, so that the skill level can be determined with high accuracy.

  In the above description, the batting motion in a real environment is described, but motion information acquired from a batting motion performed in a VR environment may be used. For example, the subject mounts a head mount display, acquires exercise information when actually performing an exercise to repel a virtual ball thrown by a pitcher on a VR under a virtual environment, and based on the exercise information Exercise performance may be assessed.

  15 shows a configuration example of the exercise performance evaluation apparatus of the fifth embodiment, FIG. 7, 10, 12 and 14 show a processing flow example of exercise information collection of the exercise performance evaluation apparatus, and FIG. 16 shows a learning process of the exercise performance evaluation apparatus. The flow is shown in FIG. 17 which shows the process flow of evaluation of the exercise performance evaluation apparatus. The exercise performance evaluation device 204 estimates the exercise performance of the subject in batting motion using the rotational motion of the body.

<Learning>
In this embodiment, a plurality of persons whose motion performance evaluations are known perform a striking motion, the following first to fifth feature quantities are extracted, and a set of the exercise performance and the extracted feature quantities is learned. It will be data. The first to fifth feature amounts can be acquired by the processing described in the first to fourth embodiments.
(1) First feature quantity A set of an average value of peak times for each motion characteristic of the flying object calculated by the performance estimation unit 220 of the first embodiment and information indicating the motion characteristic (2) Second feature quantity Index value indicating the degree of variation of the difference between the impact time and peak time for all trials calculated by the performance estimation unit 221 of Example 2 (3) Third feature quantity Calculation by the performance estimation unit 222 of Example 3 A set of an index value indicating the degree of variation of the difference between the impact time and the peak time for each movement characteristic of the flying object and the information indicating the movement characteristic (4) Fourth feature quantity The performance estimation unit 223 of the fourth embodiment Of the index value indicating the degree of variation of the difference between the impact time and the peak time when the motion characteristics of the flying object are known and the information indicating that the motion characteristics of the flying object are known (5) Fifth feature index value indicating the degree of variation of the difference between the impact time and the peak time when the motion characteristic of the flying object is not known, calculated by the performance estimating unit 223 of the fourth embodiment; Pair with information indicating that you do not know the movement characteristics of

  In the present embodiment, feature amounts selected from the first to fifth feature amounts may be used. All five may be used, any one may be used, and some may be used. In the process of learning, a learned model is obtained using learning data that is a set of a set of feature quantities obtained from a person whose exercise performance is known and the person's evaluation, and recorded in the recording unit 290 (S290) . The technology for obtaining the learned model may use an existing technology. For example, it can be generated by a method such as machine learning or neural network. In the case of SVM (Support Vector Machine), a hyperplane that accurately separates a feature vector corresponding to a high exercise performance included in the learning data and a feature vector corresponding to a low exercise performance is included. learn. By this, when a feature quantity vector whose motion performance is unknown is given, it corresponds to a category corresponding to high motion performance and a low motion performance from the positional relationship between the hyperplane and the feature quantity vector. By estimating to which of the categories the exercise performance can be estimated.

  Similarly, a model that classifies (discriminates) a feature quantity vector consisting of at least one feature quantity among the first feature quantity to the fifth feature quantity into a plurality of classes according to the height of exercise performance The same can be achieved by using a discriminative learning method to learn. In addition, types of exercise performance to classify are high / middle / low three stages, a plurality according to high of exercise performance such as A / B / C / D / E even if it is not two stages of high / low It is good also as a model classified into a level. In Examples 1 to 4, the exercise performance was estimated by providing a threshold in advance, but in this example, a learned model learned using data (learning data) of people whose exercise performance evaluation is known is used. Therefore, the setting corresponding to the setting of the threshold in the first to fourth embodiments is performed by learning. Further, in the case of the fifth embodiment, the setting corresponding to the setting of the threshold is performed by learning even for the combination of the first to fifth feature amounts, so that the motion with a plurality of feature amounts (feature amount vectors) as an input Performance evaluation is possible.

  In the case of a neural network, an estimated value of exercise performance is obtained by inputting each feature amount vector of learning data to the neural network, and the estimated value approaches the height of the correct exercise performance included in the learning data By repeatedly updating the parameters of the neural network, it is possible to learn a model of the neural network that outputs an estimated value of motion performance, using an arbitrary feature quantity vector as an input. That is, the learned model is obtained by learning the relationship between a feature quantity vector including at least one feature quantity among the first to fifth feature quantities and the height of exercise performance.

<Evaluation>
The exercise performance evaluation device 204 includes an exercise information acquisition unit 210, a performance estimation unit 224, a feature quantity extraction unit 280, and a recording unit 290. In addition, according to the selected feature amount, the characteristic information acquisition unit 230, the characteristic presentation information acquisition unit 240, the reference time detection unit 260, and the impact time acquisition unit 270 may be provided. In addition, the recording unit 290 records a learned model learned using time-series information of the rotational motion of the body acquired from the batting motion of the subject whose motion performance evaluation is known. That is, a feature quantity including at least one or more of the first to fifth feature quantities (a feature quantity vector, hereinafter also referred to as "feature quantity used for learning") is input, and an estimated value of exercise performance is output. It records learned models.

  The processing flow of the movement information collection is the same as in the first to fourth embodiments, and any of the processing flow of the movement information collection of the movement performance evaluation apparatus in FIGS. 7, 10, 12, 14 according to the feature used for learning. Or a process flow combining these may be executed. When the flow of the motion information collection ends, the information of the trial necessary for the evaluation of the feature amount used for the learning of the person to be evaluated is recorded in the recording unit 290. Note that the feature amounts used for learning may include at least one of the first to fifth feature amounts, and feature amounts other than the first to fifth feature amounts may be additionally used.

  The feature amount extraction unit 280 obtains a feature amount to be used for learning from the information on the trial of the subject recorded in the recording unit 290 (S280). The first to fifth feature quantities may be determined by the methods described in the first to fourth embodiments. The performance estimation unit 224 estimates the exercise performance of the subject using the learned model (S224). That is, the performance estimation unit 224 obtains the estimation result of the height (evaluation value) of the exercise performance of the subject by inputting the feature value obtained from the information of the trial of the subject to the learned model.

  According to the exercise performance evaluation device 204, the exercise performance of the subject is estimated based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance, so that the skill level can be determined with high accuracy.

  In the above description, the batting motion in a real environment is described, but motion information acquired from a batting motion performed in a VR environment may be used. For example, the subject mounts a head mount display, acquires exercise information when actually performing an exercise to repel a virtual ball thrown by a pitcher on a VR under a virtual environment, and based on the exercise information Exercise performance may be assessed.

[Program, recording medium]
The various processes described above may be performed not only in chronological order according to the description, but also in parallel or individually depending on the processing capability of the apparatus that executes the process or the necessity. It goes without saying that other modifications can be made as appropriate without departing from the spirit of the present invention.

  Further, when the above configuration is realized by a computer, the processing content of the function that each device should have is described by a program. The above processing function is realized on the computer by executing this program on the computer.

  The program describing the processing content can be recorded in a computer readable recording medium. As the computer readable recording medium, any medium such as a magnetic recording device, an optical disc, a magneto-optical recording medium, a semiconductor memory, etc. may be used.

  Further, this program is distributed, for example, by selling, transferring, lending, etc. a portable recording medium such as a DVD, a CD-ROM or the like in which the program is recorded. Furthermore, this program may be stored in a storage device of a server computer, and the program may be distributed by transferring the program from the server computer to another computer via a network.

  For example, a computer that executes such a program first temporarily stores a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. Then, at the time of execution of the process, the computer reads the program stored in its own recording medium and executes the process according to the read program. Further, as another execution form of this program, the computer may read the program directly from the portable recording medium and execute processing according to the program, and further, the program is transferred from the server computer to this computer Each time, processing according to the received program may be executed sequentially. In addition, a configuration in which the above-described processing is executed by a so-called ASP (Application Service Provider) type service that realizes processing functions only by executing instructions and acquiring results from the server computer without transferring the program to the computer It may be Note that the program in the present embodiment includes information provided for processing by a computer that conforms to the program (such as data that is not a direct command to the computer but has a property that defines the processing of the computer).

  Further, in this embodiment, although the present apparatus is configured by executing a predetermined program on a computer, at least a part of the processing contents may be realized as hardware.

200, 201, 202, 203, 204 Motion Performance Evaluation Device 210 Motion Information Acquisition Unit 220, 221, 222, 223, 224 Performance Estimation Unit 230 Characteristic Information Acquisition Unit 240 Characteristic Presentation Information Acquisition Unit 260 Reference Time Detection Unit 270 Impact Time Acquisition Unit 280 Feature extraction unit 290 Recording unit

Claims (10)

An exercise performance evaluation device for estimating a subject's exercise performance in a batting motion using a rotational motion of a body,
An exercise information acquisition unit for acquiring time-series information of rotational movement of the subject's body;
A performance estimation unit for estimating the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance Exercise performance evaluation device provided. The exercise performance evaluation apparatus according to claim 1, wherein
The peak time is assumed to be the time at which the magnitude of the rotational movement is maximum with reference to the predetermined timing before the arrival of the airborne object,
It also has a characteristic information acquisition unit that acquires movement characteristics of flying objects,
The performance estimation unit may be configured to calculate a difference between an average of peak times in the striking motion for the flying object having the first motion characteristic and an average of peak times in the striking motion for the flying object having the second motion characteristic. When it is the first value, the first evaluation value is output as the evaluation result of the exercise performance, and the flight time having the average of the peak time in the striking movement for the flying object having the first movement characteristic and the second movement characteristic A second evaluation corresponding to higher exercise performance than the first evaluation value when the magnitude of the difference from the average of the peak time in the striking motion with respect to the object is a second value larger than the first value An exercise performance evaluation device characterized by outputting a value as an estimation result of exercise performance. The exercise performance evaluation apparatus according to claim 1, wherein
The time at which the magnitude of the rotational movement is maximum with reference to the predetermined timing as the peak time, and the impact time is the time when an impact is applied to the reference with the predetermined timing,
It also has an impact time acquisition unit that acquires the impact time,
The performance estimation unit outputs the first evaluation value as an evaluation result of exercise performance when the variation in the difference between the peak time and the impact time is the first value, and the variation in the difference between the peak time and the impact time When the second value is smaller than the first value, the second evaluation value corresponding to the higher exercise performance than the first evaluation value is output as the estimation result of the exercise performance. Performance evaluation device. The exercise performance evaluation apparatus according to claim 1, wherein
The time at which the magnitude of the rotational movement is maximum with reference to the predetermined timing as the peak time, and the impact time is the time when an impact is applied to the reference with the predetermined timing,
A characteristic information acquisition unit that acquires movement characteristics of a flying object;
It also has an impact time acquisition unit that acquires the impact time,
The performance estimation unit determines the degree of variation of the difference between the peak time and the impact time of the flying object having the first motion characteristic, and the peak time and the impact time of the flying object having the second motion characteristic. The first evaluation value is output as an evaluation result of exercise performance when the difference with the degree of variation of the difference is the first value, and the peak time and the impact time at the time of the flying object having the first exercise characteristic When the difference between the degree of variation of the difference and the degree of variation of the difference between the peak time and the impact time at the time of the flying object having the second motion characteristic is a second value smaller than the first value An exercise performance evaluation device characterized by outputting, as an estimation result of exercise performance, a second evaluation value corresponding to exercise performance being higher than the first evaluation value. The exercise performance evaluation apparatus according to claim 1, wherein
The time at which the magnitude of the rotational movement is maximum with reference to the predetermined timing as the peak time, and the impact time is the time when an impact is applied to the reference with the predetermined timing,
The missile has a first motion characteristic or a second motion characteristic,
A characteristic information acquisition unit that acquires movement characteristics of a flying object;
A characteristic presentation information acquisition unit that acquires information on whether or not motion characteristics of a flying object are presented;
It also has an impact time acquisition unit that acquires the impact time,
The performance estimation unit determines the degree of variation of the difference between the peak time and the impact time when the target person knows the movement characteristics of the flying object, and the peak when the target person does not know the movement characteristics of the flying object When the difference between the time of day and the time of impact difference is the first value, the first evaluation value is output as the evaluation result of the exercise performance, and the target person knows the movement characteristics of the flying object. The difference between the degree of variation of the peak time and the impact time at the time of arrival and the degree of variation of the difference between the peak time and the impact time when the target person does not know the movement characteristics of the flying object is a first value A second evaluation value corresponding to higher exercise performance than the first evaluation value is output as an estimation result of the exercise performance when the second value is smaller than the second evaluation value. Exercise performance evaluation device. The exercise performance evaluation apparatus according to claim 1, wherein
A recording unit that records a learned model learned using time-series information of rotational movement of the body acquired from the batting movement of a subject whose evaluation of exercise performance is known;
A feature quantity extraction unit for extracting a feature quantity to be input to the learned model;
The exercise performance evaluation apparatus according to claim 1, wherein the performance estimation unit estimates exercise performance of the subject using the learned model. The exercise performance evaluation apparatus according to claim 1, wherein
A recording unit that records a learned model that has been learned so as to output an estimated value of exercise performance, using a feature of time series information of rotational movement as an input;
A feature quantity extraction unit for extracting a feature quantity to be input to the learned model;
The exercise performance evaluation apparatus according to claim 1, wherein the performance estimation unit estimates exercise performance of the subject using the learned model. The exercise performance evaluation apparatus according to claim 6 or 7, wherein
The feature quantity extraction unit
A first feature value which is a combination of an average value of peak times for each movement characteristic of a flying object and information indicating the movement characteristic;
A second feature value which is an index value indicating the degree of variation of the difference between the impact time and the peak time for all trials;
A third feature value that is a combination of an index value indicating the degree of variation of the difference between the impact time and the peak time for each movement characteristic of the flying object, and information indicating the movement characteristic;
A fourth set of an index value indicating the degree of variation of the difference between the impact time and the peak time when the motion characteristics of the flying object are known, and information indicating that the motion characteristics of the flying object are known; Feature value,
A fifth feature value that is a combination of an index value indicating the degree of variation of the difference between the impact time and the peak time when the motion characteristics of the flying object are not known, and information indicating that the motion characteristics of the flying object are not known An exercise performance evaluation device characterized by extracting one or more of them. An exercise performance evaluation method for estimating a subject's exercise performance in a batting motion using a rotational motion of a body,
An exercise information acquisition step of acquiring time-series information of rotational movement of the subject's body;
A performance estimation step of estimating the exercise performance of the subject from the time series information of the rotational movement acquired by the exercise information acquisition unit based on the relationship between the time series information of the rotational movement and the evaluation of the exercise performance How to evaluate exercise performance.   An exercise performance evaluation program for causing a computer to function as the exercise performance evaluation device according to any one of claims 1 to 8.

Images