JP2010017525A - Action determining apparatus and action determining method - Google Patents

Abstract

Provided is a technique capable of measuring the amount of activity by classifying the types of user activities.
A behavior determination unit 31 of a behavior determination meter 10 classifies a user's activity state into five types based on a measurement result of an acceleration sensor 20 that is a three-axis accelerometer. Furthermore, the activity age calculation unit 32 calculates the activity age based on the composition ratio of the activity states classified by the behavior determination unit 31, and the obesity activity index calculation unit 33 calculates the obesity activity index of the user. The calculation result is displayed on the display unit 15.
[Selection] Figure 2

Description

  The present invention relates to a behavior determination device and a behavior determination method, and more particularly to a behavior determination device and a behavior determination method for estimating an activity amount by classifying types of user behavior, activities, actions, and the like.

  Conventionally, the importance of prevention and improvement of lifestyle-related diseases has been evoked in various situations, but in recent years, various approaches have been implemented based on specific data and indicators. According to the Ministry of Health, Labor and Welfare's “Exercise Standard 2006 for Health Promotion”, it is recommended that physical activity of 3MET (Metabolic Equivalent) or more, including daily life activities, be performed for approximately 60 minutes a day as a preventive measure for lifestyle-related diseases. Therefore, the establishment of evaluation of daily activities is more demanded than ever.

  In the case of Japanese with a standard body shape, it is said that the values of 60% basal metabolism and 10% diet-induced body heat production are almost fixed in the breakdown of the total daily energy consumption. In addition, according to a certain research result, the cause of the individual difference in the total daily energy consumption is a part called nonexercise activity thermogenesis (NEAT) including housework activities. It accounts for 20 to 30%, and the individual difference is ± 200 to 300 kcal, which causes a large individual difference in physical activity level.

  Currently, there is an activity meter using an accelerometer, for example, as a means for evaluating a daily physical activity amount. The activity meter can estimate physical activity intensity METs, activity time, and energy consumption of each action. And the technique which improved the precision of the estimation is proposed (for example, refer patent document 1). In the technique disclosed in Patent Document 1, an exercise amount measuring device including a three-axis accelerometer is used for the purpose of accurately measuring a user's exercise intensity.

JP 2006-204446 A

  By the way, in the technique disclosed in Patent Document 1, the energy consumption per unit time can be estimated from the magnitude of acceleration, but there is a problem that the type of activity cannot be grasped. In addition, since the type of activity cannot be grasped, the estimation of energy consumption in one day and the state of activity (behavior) could not be grasped. In other words, it was difficult to know what kind of activities in daily life generate high energy consumption, and it was difficult to improve daily activities.

  This invention is made | formed in view of the above situations, The objective is to provide the technique which can measure the amount of activity by classifying the kind of user's activity.

The device according to the present invention relates to a behavior determination device. The behavior determination device includes: an acceleration acquisition unit that acquires a horizontal acceleration and a vertical acceleration; a behavior determination unit that classifies a user's activity state based on the horizontal and vertical accelerations; Is provided.
The acceleration acquisition unit may include a three-axis acceleration sensor that measures the horizontal acceleration and the vertical acceleration.
The acceleration acquisition means may acquire a measurement result of an external acceleration sensor.
The behavior determination unit may refer to a ratio of acceleration magnitudes in the horizontal direction and the vertical direction when classifying the activity state of the user.
In addition, the behavior determination apparatus may include an activity amount calculating unit that calculates user energy consumption based on the classified user activity state.
Further, the behavior determination device compares the composition ratio of the categorized activity state with the first reference configuration pattern in which the age and the activity state are associated with each other, and estimates the activity age of the user May be provided.
In addition, the behavior determination apparatus calculates the obesity activity index for calculating the obesity activity index of the user by comparing the composition ratio of the classified activity states with the second reference configuration pattern in which the body type and the activity state are associated with each other. Means may be provided.
The behavior determination apparatus may include an interface that communicates with an external device.
In addition, an angular velocity acquisition unit that acquires the angular velocity in the horizontal direction and the angular velocity in the vertical direction is provided, and the behavior determination unit may reflect the angular velocity acquired by the angular velocity acquisition unit in a process of classifying the user's activity state. Good.
In addition, the behavior determination unit may reflect the variation coefficient of the angular velocity acquired by the angular velocity acquisition unit in the process of classifying the user's activity state.
The method according to the present invention relates to a behavior determination method. This method includes a horizontal acceleration, an acceleration acquisition step of acquiring a vertical acceleration, and an action determination step of classifying a user's activity state based on each of the horizontal and vertical accelerations. .
The behavior determining step may refer to a ratio of magnitudes of the horizontal acceleration and the vertical acceleration in order to classify the user activity state.
In addition, the behavior determination method may include an activity amount calculation step of calculating user energy consumption based on the classified user activity state.
Further, the behavior determination method includes an activity age calculation step of estimating the activity age of the user by comparing the composition ratio of the classified activity states with a first reference configuration pattern in which the age and the activity state are associated with each other. May be provided.
Further, the behavior determination method includes calculating the obesity activity index for calculating the obesity activity index of the user by comparing the composition ratio of the classified activity states with a second reference configuration pattern in which the body type and the activity state are associated with each other. A process may be provided.
In addition, an angular velocity acquisition step of acquiring a horizontal angular velocity and a vertical angular velocity may be provided, and the behavior determination step may reflect the angular velocity acquired in the angular velocity acquisition step in a process of classifying a user's activity state. Good.
Further, the behavior determination step may reflect the variation coefficient of the angular velocity acquired in the angular velocity acquisition step in the process of classifying the user's activity state.

  ADVANTAGE OF THE INVENTION According to this invention, the technique which classifies the kind of activity and can measure an activity amount can be provided.

It is a figure which shows the schematic external appearance of the action determination meter based on 1st Embodiment. It is a functional block diagram which shows schematic structure of the action determination meter based on 1st Embodiment. It is the table which showed the typical example of the activity content and activity time seen according to the physical activity level. It is an example of the table which showed the relationship between the obesity activity index and the physical activity level based on 1st Embodiment. It is the figure which showed the example of a display of the composition ratio of the active state, the physical activity level, and the obesity activity index which are displayed on the display part of the action determination meter based on 1st Embodiment. The example which showed the user's energy consumption based on 1st Embodiment as a graph is shown. It is the flowchart which showed the measurement process of the acceleration in the action determination meter based on 1st Embodiment, and the determination process of operation | movement. It is a flowchart which shows the operation | movement which displays the log | history of an active state with respect to the user based on 1st Embodiment. It is the figure which showed the state to which the computer for management, the application server, and the body composition meter which function in cooperation with the action determination meter based on 2nd Embodiment were connected. It is the figure which showed the measurement result of the acceleration sensor and angular velocity sensor in five types of active states based on 3rd Embodiment in time series. It is the figure which showed the measurement example of the output of a progress pitch and an acceleration sensor at the time of the walk and driving | running | working based on 3rd Embodiment. It is the graph which showed the relationship between the estimated consumption energy at the time of a walk and driving | running | working and measured consumption energy based on 3rd Embodiment. It is the figure which showed the measurement result of the desk work and sitting position (rest) based on 3rd Embodiment continuously in time series. It is a functional block diagram which shows schematic structure of the action determination meter based on 3rd Embodiment. It is the flowchart which showed the measurement process of the acceleration in the action determination meter, and the determination process of an operation | movement based on 3rd Embodiment.

  Next, the best mode for carrying out the present invention (hereinafter simply referred to as “embodiment”) will be specifically described with reference to the drawings.

<First Embodiment>
FIG. 1 is a diagram showing a schematic appearance of a behavior determination meter 10 according to the present embodiment. FIG. 1A shows a state in which nothing is displayed on the display unit 15, and FIGS. 1B to 1D show states in which determination results described later are displayed. FIG. 2 is a functional block diagram showing a schematic configuration of the behavior determination meter 10. The behavior determination meter 10 includes an acceleration sensor 20 (triaxial acceleration watch) capable of measuring triaxial acceleration, and by using the acceleration sensor 20, the horizontal acceleration and the vertical acceleration are measured, The type of user activity is classified based on the measurement result, and the estimation accuracy of the activity amount is improved.

  In particular, in the present embodiment, by calculating the ratio between the magnitude of the acceleration in the horizontal direction and the magnitude of the acceleration in the vertical direction (referred to as “horizontal / vertical component ratio”), the user's action state (“activity state”) is also calculated. Are classified into five types: “sitting position”, “standing position”, “walking”, “exercising”, and “housework activities”. Although specific classification processing will be described later, first, the behavior determination meter 10 determines that the activity state is relatively low when the magnitude of the acceleration is smaller than the first predetermined value, and further determines the horizontal / vertical component ratio. The activity state is classified into “sitting position” or “standing position” based on the above. Subsequently, when the magnitude of the acceleration is between the first predetermined value and the second predetermined value, the behavior determination meter 10 determines that the activity state is a medium intensity, and further, the horizontal / vertical component is determined. Based on the ratio, the activity state is classified as “housework activity” or “walking”. Furthermore, when the magnitude of the acceleration is larger than the second predetermined value, the behavior determination meter 10 classifies the activity state as “exercise”. Note that the types of classification are not limited to the above five types.

  Then, based on the above-described behavior state categorization results, the total daily energy consumption, the activity time for each type of behavior, and the time breakdown by each behavior are displayed. Moreover, since qualitative changes in physical activity occur with aging (Tsuneyoshi et al., Japan Physical Fitness Society 2004), the age of activity is calculated and presented to the user in comparison with the standard behavior pattern of each age. . Furthermore, since the proportion of each posture state of the sitting position and the standing position in the daily behavior pattern is different between the obese person and the non-obese person, the actions such as sitting position, standing position, housework activity, walking are discriminated, and the user's To what extent the behavior pattern approximates the behavior pattern of an obese person is presented as an obesity activity index.

  Here, the obesity activity index will be described. The obesity activity index is calculated according to the physical activity level (total daily energy consumption / basal metabolism; PAL). FIG. 3 is a table showing typical examples of activity contents and activity times viewed by “individual physical activity level”. This table shows average data for cases of ages 15-69. “Individual activity classification” is classified into five types, and the numbers in parentheses in each classification indicate the metabolic value in each classification when the basal metabolism of sleep state is “1”. Yes. The time for classifying each activity is shown for each of the three types of physical activity levels. For example, if the “individual physical activity level” is “normal (II)”, the content of daily life is “sitting-centered work, but moving / working in the workplace, standing-up work, commuting, etc. It indicates that the activity includes “shopping / housework or light sports”. Furthermore, when classifying “individual activities” in the case of normal (II), “sleeping time” (PAL = 1.0) is 7 to 8 hours, “long-term sustainable exercise / work etc. The activity (including normal walking) "(PAL = 4.5: 3.0 to 5.9) is 2 hours. Based on the relationship between the metabolic value in the “individual activity classification” and the time of each classification, the “individual physical activity level” is classified into “low (I)”, “normal (II)”, and “high (III)”. Each PAL value is 1.50 (1.40 to 1.60), 1.75 (1.60 to 1.90), 2.00 (1.90 to 2) as the average value for each day. 20).

  Then, the physical activity level is calculated from the time of each action (sitting, standing, domestic activity, exercise, etc.) of the user and the metabolic value of each action, and the obesity activity index is calculated in 10 stages from the calculation result. The larger the obesity activity index, the more obese body pattern, the more likely it is to become obese. FIG. 4 is an example of a table showing the relationship between the obesity activity index and the physical activity level. An obesity activity index is calculated based on a standard as shown. For example, when the physical activity level is “1.75”, the obesity activity index is 3, and when the physical activity level is “1.25”, the obesity activity index is “8”. FIGS. 5A and 5B show display examples of the composition ratio of the active state, the physical activity level, and the obesity activity index. Here, FIG. 5A shows the physical activity level “1.75”. (B) is the physical activity level “1.25”. In FIG. 5 and FIG. 6 to be described later, although the sleep state is not displayed, the sleep state may naturally be displayed. Further, although “walking” and “standing position” are combined, they may be displayed individually.

  Returning to the description of FIG. 2, details of the behavior determination meter 10 will be described. The behavior determination meter 10 includes a main control unit 11, a calculation unit 12, a storage unit 13, an operation unit 14, a display unit 15, an input / output interface 16, an acceleration sensor 20, and a determination unit 30. It is prepared for. Here, the main control unit 11 comprehensively controls each component of the behavior determination meter 10. The operation unit 14 is a user interface such as a button and accepts an operation by a user. When the operation by the operation unit 14 is received, the main control unit 11 executes various processes described later in cooperation with each component. The display unit 15 is a display unit such as a liquid crystal panel, and displays measurement results and determination contents as shown in FIGS. FIG. 1B shows the activity time (minutes) and energy consumption (kcal) of the day for five types of activities. FIG. 1C shows the activity age of the user. In the figure, the fact that it is determined as “45 years old” is displayed. FIG. 1D shows that the obesity activity index is determined to be “7”. FIGS. 6A and 6B show examples in which the user's energy consumption is displayed in a graph. FIG. 6A shows a ratio (composition ratio) of the total daily energy consumption, and FIG. 6B shows a transition of the ratio of the total daily energy consumption. What display mode is used may be appropriately selected according to the processing capability of the calculation unit 12, the storage unit 13, the display unit 15, and the like of the behavior determination meter 10.

  The input / output interface 16 is communicatively connected to an external device such as a personal computer or a mobile phone. The communication connection may be a wired wireless connection, a USB (Universal Serial Bus) interface connection, an infrared communication connection, or a Bluetooth (registered trademark) connection. When the behavior determination meter 10 does not cooperate with an external device, the input / output interface 16 is not necessary. A configuration and operation in cooperation with an external device will be described later in the second embodiment.

  The acceleration sensor 20 includes a X-axis acceleration sensor 21, a Y-axis acceleration sensor 22, and a Z-axis acceleration sensor 23 to form a three-axis accelerometer, and measures acceleration in the three-axis direction at a predetermined sampling period. In the present embodiment, the horizontal and vertical accelerations are used. The acceleration in the horizontal direction may be calculated by the acceleration sensor 20 and output to the main control unit 11, or the calculation unit 12 may calculate the acceleration in the horizontal direction based on the data of the acceleration sensor 20 output via the main control unit 11. It may be calculated. In this embodiment, the calculating part 12 shall calculate the acceleration of a horizontal direction. The measurement result is digitally converted by the acceleration sensor 20, amplified as necessary, and output to the main controller 11.

  The calculation unit 12 is composed of an LSI (Large Scale Integrated Circuit) such as a CPU (Central Processing Unit), and calculates the magnitude of the combined acceleration from the magnitude of each acceleration in the horizontal and vertical directions and the magnitude of those accelerations. To do. Hereinafter, for convenience, the magnitude of acceleration in the horizontal direction is referred to as “horizontal component acceleration Ah”, and the magnitude of acceleration in the vertical direction is referred to as “vertical component acceleration Av”. In this embodiment, when the output value of each sampling of the acceleration sensor 20 exceeds a predetermined threshold, it is determined that the user has moved in the vertical direction or the horizontal direction, and the vertical component acceleration Av or the horizontal component acceleration is determined. Ah is incremented by 1 count. That is, the magnitude of the vertical component acceleration Av or the horizontal component acceleration Ah is represented by the count value. Further, the predetermined threshold differs depending on the specification of the acceleration sensor 20, and may be set as appropriate. Naturally, acceleration values corresponding to the output values of the respective samplings of the acceleration sensor 20 may be calculated and used as long as desired performance and specifications can be realized with respect to the processing load, cost, and power consumption of the CPU. .

  Furthermore, the storage unit 13 is a memory such as a ROM or a RAM, and stores a measurement result of the acceleration sensor 20 for a predetermined period and a determination result by the determination unit 30 described later for the calculation process of the calculation unit 12. accumulate. In addition, the storage unit 13 holds reference data serving as a reference for various determinations as illustrated in FIG. In FIG. 3, there are three types of physical activity levels for users who are 15 to 69 years old. As reference data, this table may be used in common for all ages, general types, and men and women, and further divided into chronological patterns and body-specific patterns (obesity patterns, standard anatomical patterns, etc.) Data may be used. In that case, when the use of the behavior determination meter 10 is started, the user inputs and sets his / her age, body type, and gender.

  The determination unit 30 is realized by an LSI such as a CPU, a memory, or an arbitrary program, and includes an action determination unit 31, an activity age calculation unit 32, an obesity activity index calculation unit 33, and an activity amount calculation unit 34. . The behavior determination unit 31 uses the horizontal axis acceleration Ah and the vertical component acceleration Av stored in the storage unit 13 for a predetermined period (for example, the immediately preceding 10 seconds), the three-axis composite acceleration Atotal, The vertical component ratio Ah / Av is calculated, and the user's activities are classified into five types of “sitting position”, “standing position”, “walking”, “exercise”, and “housework activity” according to the flowchart of FIG. Stored in the unit 13 and accumulated as determination history data.

  The activity amount calculation unit 34 calculates the accumulated time for a predetermined period for each type of activity, the ratio of the type of activity, and the energy consumption based on the user activity status accumulated in the storage unit 13. The calculation result is displayed on the display unit 15 as necessary as shown in FIG. As for the energy consumption, not only the total energy consumption but also the energy of each activity is calculated. Further, not only the energy consumption during the actual integration time but also the estimated daily energy consumption may be calculated. In addition, the activity amount calculation unit 34 accumulates the calculation result of the above-mentioned activity type ratio and energy consumption every day as the energy consumption history data in the storage unit 13.

  The activity age calculation unit 32 compares the energy consumption history stored in the storage unit 13 with the reference data, and calculates which age's behavior pattern the user's behavior corresponds to. This calculation is performed based on, for example, a user instruction, and the calculation result is displayed on the display unit 15 as shown in FIG.

  The obesity activity index calculation unit 33 uses, as the obesity activity index, how close the user's behavior pattern is to the obesity pattern based on the energy consumption history accumulated in the storage unit 13 and the reference data. Calculate. The calculation result is displayed on the display unit 15 as shown in FIG.

  The operation of the behavior determination meter 10 having the above configuration will be described with reference to the flowcharts of FIGS. FIG. 7 shows acceleration measurement and action determination processing in the action determination meter 10. When the behavior determination meter 10 is on, the main control unit 11 reads the data measured by the acceleration sensor 20 (X-axis acceleration sensor 21, Y-axis acceleration sensor 22, Z-axis acceleration sensor 23) and stores it in the storage unit 13. (S10).

  Subsequently, the calculation unit 12 calculates a horizontal component acceleration Ah, a vertical component acceleration Av, and a three-axis combined acceleration Atotal based on measurement data for a predetermined period (S12). Then, the behavior determination unit 31 determines whether or not the triaxial synthetic acceleration Atotal is less than 100 counts based on the calculation result (S14). When the triaxial composite acceleration Atotal is less than 100 counts (Y in S14), the behavior determination unit 31 determines that the activity is relatively mild, and further calculates the horizontal / vertical component ratio Ah / Av. It is determined whether or not the value is less than 1 (S16). When the horizontal / vertical component ratio Ah / Av is less than 1 (Y of S16), it indicates that the user's activity is less up and down, and the action determination unit 31 is in the state where the user's activity state is “sitting” (S18). If the horizontal / vertical component ratio Ah / Av is 1 or more (N in S16), it is determined that the user's activity state is the "standing position" (S20).

  If it is determined in step S14 that the triaxial synthetic acceleration Atotal is 100 counts or more (N in S14), the behavior determination unit 31 determines whether the triaxial synthetic acceleration Atotal is less than 200 ( S22). When the triaxial synthetic acceleration Atotal is equal to or greater than 200 counts (N in S22), the behavior determination unit 31 determines that the user's activity state is an “exercise” state such as running (S24). When the triaxial synthetic acceleration Atotal is less than 200 counts (Y in S22), the behavior determination unit 31 determines whether the horizontal / vertical component ratio Ah / Av is less than 5 (S26). When the horizontal / vertical component ratio Ah / Av is 5 or more (N in S26), the behavior determination unit 31 determines that the user's activity state is the “housework activity” state (S28). If the horizontal / vertical component ratio Ah / Av is less than 5 (Y in S26), the behavior determination unit 31 determines that the user's activity state is “walking” (S30).

  When a determination result is obtained in the processing of S18, S20, S24, S28, and S30, the determination result of the action determination unit 31 is stored and stored as determination history data in the storage unit 13 in association with the date and time information (S32). .

  FIG. 8 is a flowchart showing an operation of displaying an activity history for the user. First, the main control unit 11 of the behavior determination meter 10 acquires an operation instruction for display from the user through the operation unit 14 (S50).

  Then, based on the determination history data accumulated in the storage unit 13, the activity amount calculation unit 34 calculates the energy consumption of each activity, the accumulated time of each activity, and the composition ratio by type of activity time (S52). ). In addition, when the above-described calculation processing is not performed on the previous day, for example, when the behavior determination meter 10 is turned on, the activity amount calculation unit 34 may calculate and record and accumulate in the storage unit 13.

  Next, the activity amount calculation unit 34 reads various activity patterns (period-specific patterns and obese person patterns) stored in the storage unit 13 (S54).

  Subsequently, the activity age calculation unit 32 calculates the calculated energy consumption of each activity, and compares the accumulated time of each activity, the composition ratio of each type of activity time, and the age-specific pattern read from the storage unit 13. Then, it is determined which age pattern the user's activity state is close to and is calculated as the user's activity age (S56).

  Further, the obesity activity index calculation unit 33 calculates, as an obesity activity index, how close the user activity state is to the obesity activity pattern compared to the obesity pattern (S58). Note that, as described above, the determination result in the action determination unit 31 includes time information. When calculating the activity age or the obesity activity index, processing for estimating the type of user's action may be performed for a time when there is no determination result. For example, “PAL = 1” may be used as the value of the physical activity level as data of time when the user is assumed to be sleeping.

  When the calculation process by the activity age calculation unit 32 and the obesity activity index calculation unit 33 is completed, the main control unit 11 displays the calculation result as shown in FIGS. 1B to 1D, FIG. 15 (S60), and the calculation result is stored in the storage unit 13 (S62). The calculation results that can be displayed are, as described above, calculation of energy consumption calculated for each activity, cumulative time of each activity, composition ratio of each type of activity time, activity age, and obesity activity index. May be displayed alone, or all items may be displayed at once. Furthermore, it may be displayed together with a typical pattern (standard pattern, obesity pattern, pattern to be targeted, etc.).

  As described above, according to the behavior determination meter 10 having the configuration and operation as in the present embodiment, the user's activity state can be grasped more accurately. Furthermore, it is possible to present to the user what pattern the activity state of the user belongs to. As a result, the user can accurately grasp his / her activity state.

<Second Embodiment>
In the first embodiment, the state where the behavior determination meter 10 functions alone has been described. In the present embodiment, a case will be described in which the measurement result of the behavior determination meter 10 is used by a body composition meter on an application on a personal computer or a network.

  FIG. 9 is a diagram illustrating a state in which the management computer 50, the application server 40, and the body composition meter 70 that function in cooperation with the behavior determination meter 10 are connected. The behavior determination meter 10 transmits the measurement data to the management computer 50, the application server 40, and the body composition meter 70. Then, the user activates predetermined health management applications provided in the management computer 50, the application server 40, and the body composition meter 70, respectively, and functions similar to the functions realized by the behavior determination meter 10 in the first embodiment. Is used.

  First, the cooperation between the management computer 50 and the behavior determination meter 10 will be described. The configuration of the behavior determination meter 10 is the same as that of the first embodiment. The input / output interface 16 includes a USB interface and can communicate with an external device based on the USB standard.

  On the other hand, the management computer 50 includes a main control unit 51, an action determination management unit 52, a data storage unit 53, a model storage unit 54, a display unit 55, and an input / output interface 56. The main control unit 51 comprehensively controls the management computer 50. The behavior determination management unit 52 performs the same function as the behavior determination meter 10 together with the data storage unit 53 and the model storage unit 54.

  First, the user sets the behavior determination meter 10 and the management computer 50 in a communicable state. More specifically, each of the input / output interfaces 16 and 56 of the behavior determination meter 10 and the management computer 50 includes, for example, a USB interface and a USB interface, and communication according to the USB standard is possible. It has become.

  The behavior determination management unit 52 acquires the determination history data from the behavior determination meter 10 and stores it in the data storage unit 53. At this time, the determination history data may be stored in association with an authentication code for individual identification of the behavior determination meter 10. Then, the data of the same authentication code is added to the data stored in the data storage unit 53. At this time, when data of a plurality of behavior determination meters 10 exist in the data storage unit 53, the designation may be received from the user, or the data matching the authentication code may be updated, Both authentication processes may be performed.

  In response to the user's operation, the management computer 50 calculates the total energy consumption and the energy of each activity based on the determination history data. The calculation result is stored in the data storage unit 53 as the energy consumption history data as the ratio of the above-mentioned activity types and the energy consumption.

  Subsequently, in response to the user's operation, the behavior determination management unit 52 calculates the energy consumption of each activity, the accumulated time of each activity, and the composition ratio of each type of activity time, as shown in the first embodiment. The activity age and the obesity activity index are calculated. Then, the calculation result is displayed on the display unit 55.

  When the behavior determination meter 10 is configured to accumulate the measurement data of the acceleration sensor 20, the behavior determination management unit 52 may acquire the measurement data and record it in the data accumulation unit 53. Then, based on the measurement data, the data storage unit 53 classifies the user's activities into five types of “sitting position”, “standing position”, “walking”, “exercise”, and “housework activity”. Energy consumption, the accumulated time of each activity, the composition ratio of each activity time, and the activity age and the obesity activity index are calculated. And the model memory | storage part 54 is equipped with more patterns than the action determination meter 10 about various patterns (a standard pattern, an obesity pattern, a pattern according to age), etc., and the action determination management part 52 is a user's body shape, A reference pattern to be compared may be selectable in consideration of age, sex, and the like.

  And when displaying each calculation result on the display part 55, if a graphic display and a model used as a reference | standard are displayed in parallel, grasping | ascertaining of a user's own activity state will become easier. In addition to the above functions, the behavior determination management unit 52 may be capable of correcting data acquired from the behavior determination meter 10, for example. For example, the user sometimes removes the behavior determination meter 10 when performing intense exercise or swimming. Therefore, the behavior determination management unit 52 displays the type of activity in time series, and accepts correction of the period and the type of behavior. By providing such a function, the user's activity amount calculation accuracy can be improved.

  Furthermore, as shown in FIG. 9, when the management computer 50 is connected to a predetermined application server 40 via the network line 90, the application server 40 includes, for example, a user data management unit 41 and user data input / output. Unit 42, model storage unit 43, and user data storage unit 44.

  The user data management unit 41 controls connection from the management computer 50 and executes processing based on a request instruction from the management computer 50. The user data input / output unit 42 performs authentication processing of the management computer 50 and data input / output control. The model storage unit 43 stores various patterns (standard pattern, obesity pattern, age-specific pattern) and can be transmitted to the management computer 50. The user data storage unit 44 acquires and stores data from many users via the management computer 50 or directly from the behavior determination meter 10. In such a configuration, the user can manage his / her activity amount on the network.

  For example, it is assumed that the function of the behavior determination meter 10 is mounted on a mobile phone or the like. In recent years, mobile phones include products in which a three-dimensional sensor (three-axis acceleration watch) such as a motion sensor is mounted. In such a case, the mobile phone can be made to function as the behavior determination meter 10 by installing an application for calculating the amount of activity. Furthermore, it is possible to easily connect to the application server 40 using the communication function of the mobile phone, manage user activity, and upgrade the application. Further, the obesity activity index up to a predetermined time is calculated, and the behavior determination meter 10 uses the function of the mobile phone, for example, when the obesity activity index is high, the color of the small display of the mobile phone is displayed in red. May be. This can prompt the user to exercise.

  In the case where the behavior determination meter 10 and the body composition meter 70 can be linked, the body composition meter 70 acquires the measurement result or the determination result of the behavior determination meter 10, and the weight and body measured by the body composition meter 70. By managing together with the fat percentage and muscle mass, the user's physical health management can be executed more effectively. Furthermore, if the body composition meter 70 can cooperate with the management computer 50 and the application server 40, the user's physical health management can be performed in more detail and effectively. An authentication key device having a storage area may be used to exchange various data between the behavior determination meter 10 and other devices (application server 40, management computer 50, body composition meter 70).

<Third Embodiment>
In the present embodiment, in order to improve the discrimination accuracy of the above-described embodiment, an angular velocity sensor (two-axis gyroscope) 25 described later with reference to FIG. In addition, the user's action is discriminated using the measurement data of the angular velocity sensor 25. Specifically, in the above, according to the flowchart of FIG. 7, the user activities are classified into five types, but in this embodiment, “housework activities” are further divided into two types, and finally “sitting” and “ The user's activities are classified into 6 types: standing position, exercise, walking, daily activity action, and desk work.

  First, a technique for classifying “housework activities” into “daily activity activities” and “desk work” will be described. Here, four types of differences between (1) sitting and standing, (2) walking and running, (3) walking and housework activities (with vacuum cleaner), and (4) sitting and deskwork will be described. FIG. 10 shows the measurement results of the acceleration sensor 20 and the angular velocity sensor 25 in time series for five types of states. 10 (a) is “sitting position”, FIG. 10 (b) is “standing position”, FIG. 10 (c) is “sitting position (resting on a chair)”, FIG. 10 (d) is “walking”, FIG. (E) shows data on the condition of “housework activity (vacuum cleaner)”. In addition, the sampling frequency in measurement was 32 Hz, and the behavior determination meter 110 was fixed to the left chest. The longitudinal acceleration is indicated by “acc-x”, the vertical acceleration is indicated by “acc-z”, and the horizontal acceleration is indicated by “acc-y”. Further, the angular velocity in the horizontal direction is indicated by “jya-x”, and the angular velocity in the vertical direction is indicated by “jya-z”.

(1) About the difference between sitting and standing In general, people in a sitting position tend to lean forward compared to a standing position, and the gravitational acceleration applied downward is in the vertical direction (acc-z) and forward. It is decomposed in the direction (acc-x) (see FIGS. 10A and 10B). Further, when leaning on a chair or the like in the sitting position, the downward gravitational acceleration is decomposed into a vertical direction (acc-z) and a backward direction (acc-x) when viewed in detail (see FIG. 10C). From the above, it has been confirmed that the sitting position and the standing position can be discriminated by the ratio of the longitudinal acceleration (acc-x) and the vertical acceleration (acc-z). Although there is a method of discriminating from the viewpoint of fluctuation using FFT analysis, there is room for improvement in discrimination accuracy as compared with a method using a ratio of longitudinal acceleration (acc-x) and vertical acceleration (acc-z). There is.

(2) Differences between walking and running When walking and running at the same speed, fluctuations in pitch and amplitude are observed. As shown in FIG. 11, when walking and running at the same speed and the same pitch, it was confirmed that the amplitude increased during the running. From this, walking and running can be distinguished by using the product of pitch and amplitude. Moreover, there is a possibility of improving the discrimination accuracy by multiplying the product of pitch and amplitude by the height which is the body information. In general, the acceleration value increases dramatically when moving from walking to running. FIG. 12 shows a relationship between estimated energy consumption and actually measured energy consumption during walking and running. Here, the estimated energy consumption during travel uses a formula for calculating the estimated energy consumption during walking. As described above, there is a problem in that the estimation formula greatly overestimates the actually measured energy consumption during traveling. Therefore, by determining whether walking or running and using separate estimation formulas or correction formulas for walking and running, the calculation accuracy of energy consumption can be improved.

(3) Difference between walking and housework activities As shown in FIGS. 10 (d) and 10 (e), the walking in FIG. 10 (d) shows periodic changes in acceleration and angular velocity, whereas FIG. The vacuum cleaner (housework activity) in (e) shows a non-periodic change in acceleration and a change in angular velocity. Here, the coefficient of variation in angular velocity was 0.044 when walking, and 0.149 when vacuumed (housework activity). Therefore, by using the coefficient of variation of acceleration and angular velocity, it is possible to accurately discriminate between walking and housework activities.

(4) Difference between sitting position and desk work FIG. 13 shows measurement results of desk work and sitting position (resting) in time series. As shown in the figure, a non-periodic acceleration change and an angular speed change are observed in the deskwork state, but a non-periodic change is not observed in the sitting position. Note that the change in acceleration is actually very small and is not more than the current threshold value set so as not to detect the shaking of the vehicle. By setting to, it is possible to distinguish between desk work and sitting (resting).

  Based on the above knowledge, the action determination meter 110 that classifies user activities into six types of “sitting position”, “standing position”, “exercise”, “walking”, “daily activity action”, and “desk work” and the processing of the classification determination are described below. Explained.

  FIG. 14 is a functional block diagram illustrating a schematic configuration of the behavior determination meter 110 according to the present embodiment. As described above, this behavior determination meter 110 is obtained by adding the biaxial angular velocity sensor 25 to the behavior determination meter 10 of the first embodiment. Since the configuration other than the additional configuration is the same, the description of the same configuration and its operation is omitted.

  The angular velocity sensor 25 measures the angular velocity of each movement in the horizontal direction and the vertical direction. More specifically, the angular velocity sensor 25 outputs to the main control unit 11 at a predetermined sampling period. And the calculating part 12 acquires the output of the angular velocity sensor 25 via the main control part 11, and calculates the variation coefficient of a predetermined period. Here, for convenience, the variation coefficient of the x component angular velocity will be referred to as x component variation coefficient CVx, and the variation coefficient of the y component angular velocity will be referred to as y component variation coefficient CVy.

  Next, the classification determination process with the above configuration will be described with reference to the flowchart of FIG. In the process of this flowchart, in the flowchart of FIG. 7, the output of the angular velocity sensor 25 is added as a data reading target in the process of S10, and the calculation of variation coefficients CVx and CVy is added in the process of S12. Further, the processing (S26 to S30) in the case where the triaxial synthetic acceleration Atotal is less than 200 counts (Y in S22) is replaced as described below. Therefore, the description of the same processing is omitted as appropriate.

  In the state where the behavior determination meter 10 is on, the main control unit 11 reads the measurement data of the acceleration sensor 20 (X-axis acceleration sensor 21, Y-axis acceleration sensor 22, Z-axis acceleration sensor 23) and measurement data of the angular velocity sensor 25. The data is stored in the storage unit 13 (S10a).

  Subsequently, the calculation unit 12 calculates the horizontal component acceleration Ah, the vertical component acceleration Av, the three-axis composite acceleration Atotal, the x component variation coefficient CVx, and the y component variation coefficient CVy (S12a). Then, following S12a, the processes of S14, S16, S18, S20, and S22 are performed. When the triaxial synthetic acceleration Total is 200 counts or more in the process of S22 (N of S22), the process of S24 is performed.

  Next, when the triaxial resultant acceleration Atotal is less than 200 counts (Y in S22), characteristic processing (S25a to S29a) is performed in the present embodiment. Specifically, the behavior determination unit 31 determines whether or not the horizontal / vertical component ratio Ah / Av is less than 5 and the x component variation coefficient CVx that is the angular velocity variation coefficient in the horizontal direction is less than 1. (S25a). When the horizontal / vertical component ratio Ah / Av is less than 5 and the x component variation coefficient CVx is less than 1 (Y in S25a), the behavior determination unit 31 is in the “walking” state of the user activity state. (S26a).

  When the horizontal / vertical component ratio Ah / Av is less than 5 and the condition that the x component variation coefficient CVx, which is the angular velocity variation coefficient in the vertical direction, is less than 1 (N in S25a), the action determination unit 31 It is determined whether the horizontal / vertical component ratio Ah / Av is less than 5 and the y component variation coefficient CVy, which is the angular velocity variation coefficient in the vertical direction, is 1 or more (S27a). When the horizontal / vertical component ratio Ah / Av is less than 5 and the y component variation coefficient CVy is 1 or more (Y in S27a), the behavior determination unit 31 is in the “desk work” state of the user activity. (S28a). When the horizontal / vertical component ratio Ah / Av is less than 5 and the y component variation coefficient CVy is less than 1 (N in S27a), the behavior determination unit 31 indicates that the user's activity state is “daily life”. It is determined that the state is "operation" (S29a).

  When a determination result is obtained in the processes of S18, S20, S24, S26a, S28a, and S29a, the determination result of the behavior determination unit 31 is stored and accumulated as determination history data in the storage unit 13 in association with the date and time information. (S32).

  In S25, if the horizontal / vertical component ratio Ah / Av is less than 5 and the y component variation coefficient CVy is less than 1 (Y in S25a), it is determined that the state is “walking” immediately. Instead, the state of “walking” or “running” may be classified based on the walking speed and the walking pitch. The walking speed and the walking pitch can be easily calculated from the measurement result of the acceleration sensor 20. In this case, the calculation unit 12 calculates the walking speed and the walking pitch in the process of S12.

  The present invention has been described based on the embodiments. This embodiment is an exemplification, and it will be understood by those skilled in the art that various modifications can be made to the combination of each component and each processing process, and such modifications are also within the scope of the present invention. . For example, although each component has been described as a functional block, naturally, the function of each component may be exhibited by executing a processing program for each function by a common CPU, memory, or the like, and is particularly limited. Those skilled in the art will understand that this is not the case.

10, 110 Action determination meter 11 Main control unit 12 Calculation unit 13 Storage unit 14 Operation unit 15 Display unit 16 Input / output interface 20 Acceleration sensor 21 X-axis acceleration sensor 22 Y-axis acceleration sensor 23 Z-axis acceleration sensor 25 Angular velocity sensor 30 Determination Unit 31 Behavior determination unit 32 Activity age calculation unit 33 Obesity activity index calculation unit 34 Activity amount calculation unit 40 Application server 41 User data management unit 42 User data input / output unit 43 Model storage unit 44 User data storage unit 50 Management computer 51 Main Control unit 52 Action determination management unit 53 Data storage unit 54 Model storage unit 55 Display unit 56 Input / output interface 70 Body composition meter 90 Network line

Claims (17)

Acceleration acquisition means for acquiring horizontal acceleration and vertical acceleration;
Action discriminating means for classifying a user's activity state based on each acceleration in the horizontal direction and the vertical direction;
An action determination device comprising:   The behavior determination apparatus according to claim 1, further comprising a three-axis acceleration sensor that measures the horizontal acceleration and the vertical acceleration as the acceleration acquisition unit.   The behavior determination apparatus according to claim 1, wherein the acceleration acquisition unit acquires a measurement result of an external acceleration sensor.   The said action discrimination means refers to the ratio of the magnitude of the acceleration in the horizontal direction and the vertical direction when classifying the activity state of the user. Action determination device.   5. The behavior determination device according to claim 1, further comprising an activity amount calculating unit that calculates user energy consumption based on the classified user activity state. 6.   The activity age calculation means which estimates the activity age of a user by comparing the composition ratio of the classified activity status with a first reference configuration pattern in which the age and the activity status are associated with each other. Item 6. The action determination device according to any one of Items 1 to 5.   It comprises an obesity activity index calculating means for calculating the obesity activity index of the user by comparing the composition ratio of the classified activity state with a second reference configuration pattern in which the body type and the activity state are associated with each other. The behavior determination apparatus according to claim 1.   The behavior determination apparatus according to claim 1, further comprising an interface that communicates with an external device. An acceleration acquisition step of acquiring horizontal acceleration and vertical acceleration;
An action determination step for classifying a user's activity state based on each acceleration in the horizontal direction and the vertical direction;
An action determination method characterized by comprising:   The behavior determination method according to claim 9, wherein the behavior determination step refers to a ratio of magnitudes of the horizontal acceleration and the vertical acceleration in order to classify a user's activity state. .   The behavior determination method according to claim 9, further comprising an activity amount calculation step of calculating energy consumption of the user based on the classified user activity state.   An activity age calculation step of estimating the activity age of the user by comparing the composition ratio of the classified activity status with a first reference configuration pattern in which the age and the activity status are associated with each other. Item 12. The behavior determination method according to any one of Items 9 to 11.   The method further comprises an obesity activity index calculating step of calculating an obesity activity index of the user by comparing the composition ratio of the classified activity state with a second reference configuration pattern in which the body type and the activity state are associated with each other. The behavior determination method according to claim 9. Comprising an angular velocity acquisition means for acquiring a horizontal angular velocity and a vertical angular velocity;
The behavior determination device according to claim 1, wherein the behavior determination unit reflects the angular velocity acquired by the angular velocity acquisition unit in a process of classifying a user's activity state.   15. The behavior determination apparatus according to claim 14, wherein the behavior determination unit reflects the variation coefficient of the angular velocity acquired by the angular velocity acquisition unit in a process of classifying a user's activity state. An angular velocity acquisition step of acquiring a horizontal angular velocity and a vertical angular velocity;
The behavior determination method according to any one of claims 9 to 13, wherein the behavior determination step reflects the angular velocity acquired in the angular velocity acquisition step in a process of classifying a user's activity state. The behavior determination method according to claim 16, wherein the behavior determination step reflects the variation coefficient of the angular velocity acquired in the angular velocity acquisition step in a process of classifying a user's activity state.

Images