Action determination device, action determination method, and action determination program

Abstract

To provide an action determination system, an action determination method, and an action determination program, capable of properly detecting an abnormal behavior by a pedestrian.SOLUTION: An action determination system 1001 according to the present disclosure includes an acceleration sensor 1, a walk determination unit 22, a threshold value calculation unit 24, a behavior determination unit 25, and an output unit 26. The acceleration sensor 1 detects acceleration of a user. The walk determination unit 22 determines whether or not the user is walking on the basis of the acceleration detected by the acceleration sensor 1 and a predetermined first threshold value. The threshold value calculation unit 24 calculates a second threshold value on the basis of acceleration detected in a period when the walk determination unit 22 has determined that the user is walking, among the accelerations detected by the acceleration sensor 1. The behavior determination unit 25 determines whether or not the user makes an abnormal behavior, on the basis of the acceleration detected by the acceleration sensor 1 and the second threshold value calculated by the threshold value calculation unit 24. The output unit 26 outputs a determination result by the behavior determination unit 25.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a behavior determination system, a behavior determination method, and a behavior determination program. [Background technology]

[0002] From the viewpoint of accident prevention and the like, techniques have been disclosed in recent years for detecting abnormal behavior of pedestrians on the road, such as a sudden run, change of direction, or sudden stop.

[0003] For example, Patent Document 1 discloses a detection system that determines whether a person is walking with a predetermined stride length by using walking conditions including at least one of the maximum acceleration value calculated from an acceleration signal obtained from a body movement sensor, the person's stride length, and the person's walking cycle. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2019-198532 A Summary of the Invention [Problem to be solved by the invention]

[0005] When the detection system described in Patent Document 1 detects abnormal behavior of a pedestrian based on stride length, walking cycle, etc., it takes a time equivalent to at least one step of the pedestrian's walking from the time the pedestrian starts to behave abnormally to the time the abnormal behavior is actually detected.

[0006] On the other hand, when the detection system described in Patent Document 1 detects abnormal behavior of a pedestrian based on the maximum value of the pedestrian's acceleration, it is difficult to determine an appropriate threshold because the magnitude of acceleration during normal walking differs for each user. As a result, the detection system described in Patent Document 1 sometimes fails to detect abnormal behavior of some pedestrians.

[0007] The present disclosure has been made to solve such problems, and has an object to provide a behavior determination system, a behavior determination method, and a behavior determination program that are capable of appropriately detecting abnormal behavior of a pedestrian. [Means for solving the problem]

[0008] A behavior determination system according to an embodiment of the present disclosure includes: An acceleration sensor that detects the acceleration of a user; a walking determination unit that determines whether or not the user is walking based on the acceleration detected by the acceleration sensor and a predetermined first threshold value; a threshold calculation unit that calculates a second threshold based on the acceleration detected during a period in which the walking determination unit determines that the user is walking, among the accelerations detected by the acceleration sensor; and a behavior determination unit that determines whether or not the user is performing abnormal behavior based on the acceleration detected by the acceleration sensor and the second threshold calculated by the threshold calculation unit; An output unit that outputs a determination result by the behavior determination unit. It is a behavior determination system.

[0009] A behavior determination method according to an embodiment of the present disclosure includes: Detect the user's acceleration, determining whether the user is walking based on the detected acceleration and a first predetermined threshold value; Calculating a second threshold value based on the acceleration detected during a period in which it is determined that the user is walking, among the detected accelerations; determining whether the user is performing an abnormal behavior based on the detected acceleration and the calculated second threshold value; Output the judgment result. This is a behavior determination method.

[0010] A behavior determination program according to an embodiment of the present disclosure includes: Detect the user's acceleration, determining whether the user is walking based on the detected acceleration and a first predetermined threshold value; Calculating a second threshold value based on the acceleration detected during a period in which it is determined that the user is walking, among the detected accelerations; determining whether the user is performing an abnormal behavior based on the detected acceleration and the calculated second threshold value; Outputting the determination result, causing a computer to execute processing, It is a behavioral judgment program. Effect of the Invention

[0011] The present disclosure makes it possible to provide a behavior determination system, a behavior determination method, and a behavior determination program capable of appropriately detecting abnormal behavior of a pedestrian. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration of a behavior determination system according to a first embodiment. [Diagram 2] 4 is a flowchart showing the operation of the behavior determination system according to the first embodiment. [Diagram 3] FIG. 11 is a block diagram showing a configuration of a behavior determination device according to a second embodiment. [Figure 4] FIG. 11 is a block diagram showing the configuration of a calculation unit according to a second embodiment. [Diagram 5] 10 is a flowchart showing the operation of a behavior determining device according to the second embodiment. [Figure 6] 10 is a flowchart showing the operation of a behavior determining device according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] (First embodiment) <Configuration of behavior judgment system> Hereinafter, a first embodiment of the present disclosure will be described in detail with reference to the drawings. First, the configuration of the behavior determination system according to the present embodiment will be described in detail. Fig. 1 is a block diagram for explaining the configuration of the behavior determination system according to the first embodiment.

[0014] The behavior determination system 1001 according to the present embodiment is a system for detecting abnormal behavior of a user. The behavior determination system 1001 according to the present disclosure includes an acceleration sensor 1, a walking determination unit 22, a threshold calculation unit 24, a behavior determination unit 25, and an output unit 26.

[0015] The acceleration sensor 1 detects the acceleration of a pedestrian. The acceleration sensor 1 supplies the detected acceleration to the walking determination unit 22, the threshold calculation unit 24, and the behavior determination unit 25.

[0016] The walking determination unit 22 determines whether or not the user is walking based on the acceleration detected by the acceleration sensor and a first threshold value that is determined in advance. The walking determination unit 22 outputs a determination result to the threshold value calculation unit 24.

[0017] The first threshold here is a threshold determined for a quantity that characterizes the detected acceleration, such as the magnitude of acceleration or the period of fluctuation, and is a threshold determined experimentally or empirically. Since the first threshold is a preset value, it is preferable to determine it for a quantity with small individual differences between users, and it may be determined for the period of fluctuation of acceleration, for example. However, the period of fluctuation of acceleration here refers to the period of fluctuation of acceleration caused by walking, which is a type of repetitive motion.

[0018] The threshold calculation unit 24 acquires the acceleration of the user from the acceleration sensor 1, and acquires a determination result as to whether or not the user is walking from the walking determination unit 22. Then, the threshold calculation unit 24 calculates a second threshold based on the acceleration detected during a period in which the walking determination unit determines that the user is walking, among the accelerations detected by the acceleration sensor. The threshold calculation unit 24 supplies the calculated second threshold to the behavior determination unit 25.

[0019] The second threshold here is a threshold set for a quantity that characterizes the detected acceleration, such as the magnitude of acceleration or a fluctuation period. However, unlike the first threshold, the second threshold is a threshold calculated for each user, and therefore does not need to be set for a quantity with small individual differences between users. In addition, since the second threshold is used to detect abnormal behavior of a user, it is preferable that the second threshold is set for a quantity that instantly reflects the abnormal behavior of the user. For example, the second threshold may be set for the magnitude of acceleration.

[0020] The behavior determination unit 25 acquires the acceleration of the user from the acceleration sensor 1, and acquires the second threshold from the threshold calculation unit 24. The behavior determination unit 25 determines whether or not the user is performing abnormal behavior based on the magnitude of the acceleration detected by the acceleration sensor and the second threshold calculated by the threshold calculation unit. The behavior determination unit 25 supplies the determination result to the output unit 26.

[0021] However, the abnormal behavior referred to here is, for example, a behavior accompanied by a change in acceleration such as a sudden run, a change of direction, or a sudden stop. Since abnormal behavior includes behavior that may cause an accident and behavior to avoid an accident, the occurrence of an accident can be suppressed by detecting abnormal behavior and taking appropriate measures.

[0022] The output unit 26 obtains the determination result as to whether or not the user is performing abnormal behavior from the behavior determination unit 25. Then, the output unit 26 outputs the determination result by the behavior determination unit 25. The output unit 26 may output the determination result to a device that issues a warning to the user, for example. The output unit 26 may also output the determination result to an external server for analyzing abnormal behavior of multiple users, for example.

[0023] <Operation of the behavior judgment system> Next, the operation of the behavior determination system, i.e., the behavior determination method according to the first embodiment, will be described in detail. Fig. 2 is a flowchart for explaining the operation of the behavior determination system according to the first embodiment. In the following description, Fig. 1 will be referred to as appropriate.

[0024] First, the acceleration sensor 1 detects the acceleration of the user (step S101). Next, the walking determination unit 22 determines whether or not the user is walking based on the acceleration detected by the acceleration sensor 1 and the first threshold value (step S102). Next, the threshold calculation unit 24 calculates a second threshold based on the acceleration detected during the period in which it is determined that the user is walking (step S103).

[0025] Next, the behavior determination unit 25 determines whether or not the user is performing abnormal behavior based on the magnitude of the detected acceleration and a second threshold value (step S104). Finally, the output unit 26 outputs the determination result as to whether or not the user is performing abnormal behavior (step S105), and the behavior determination system 1001 ends the series of operations.

[0026] If it is determined in step S104 that the user is performing abnormal behavior, the behavior determination system 1001 may execute step S105. If it is determined in step S104 that the user is not performing abnormal behavior, the behavior determination system 1001 may end the series of operations without executing step S105.

[0027] Here, the behavior determination system 1001 may execute steps S101 to S105 continuously in parallel. In other words, the behavior determination system 1001 may continuously update the second threshold based on the acceleration detected by the acceleration sensor, and determine whether or not the user is performing abnormal behavior. Alternatively, the behavior determination system 1001 may execute steps S101, S104, and S105 in parallel and continuously, and execute steps S102 and S103 at predetermined intervals. In other words, the behavior determination system 1001 may update the second threshold at predetermined intervals.

[0028] As described above, the behavior determination system 1001 according to this embodiment detects the acceleration of a user who is walking, and calculates a threshold value based on the acceleration. With such a configuration, the behavior determination system 1001 can set an appropriate threshold for each user, and as a result, can appropriately detect abnormal behavior of a pedestrian.

[0029] Second embodiment <Configuration of the behavior determination device> Hereinafter, the second embodiment of the present disclosure will be described in detail with reference to the drawings. First, the configuration of the behavior determination system according to the present embodiment will be described in detail. The behavior determination system according to the present embodiment is configured as a behavior determination device. Fig. 3 is a block diagram for explaining the configuration of the behavior determination device according to the second embodiment.

[0030] A behavior determination device 1002 according to the present disclosure is carried by a user while walking and detects the acceleration of the user while walking. The behavior determination device 1002 then determines whether the pedestrian is performing abnormal behavior, such as suddenly running, changing direction, or suddenly stopping, based on the detected acceleration. The behavior determination device 1002 according to this embodiment includes an acceleration sensor 1, a calculation unit 2, a warning unit 3, a position information acquisition unit 4, a storage unit 5, and a transmission unit 6.

[0031] The acceleration sensor 1 is a three-axis acceleration sensor that detects the acceleration applied to itself and supplies the detected acceleration to the calculation unit 2. The acceleration sensor 1 may be, for example, a piezo-resistance type acceleration sensor, but may be an acceleration sensor using any type as long as it can appropriately measure the acceleration of a user while walking.

[0032] Because the behavior determination device 1002 is carried by a user, the acceleration detected by the acceleration sensor 1 includes an acceleration component resulting from the user's walking and an acceleration component resulting from the earth's gravitational force. In the following description, the acceleration detected by the acceleration sensor 1 is referred to as the detected acceleration, the acceleration component resulting from the user's walking is referred to as the body movement acceleration component, and the acceleration component resulting from the earth's gravitational force is referred to as the gravitational acceleration component.

[0033] The detected acceleration is a three-dimensional vector quantity having a direction and a magnitude. Therefore, the acceleration sensor 1 may detect the magnitude of the acceleration in three mutually orthogonal directions, and detect the detected acceleration as a vector in a three-dimensional space having these three values ​​as x-axis components, y-axis components, and z-axis components, respectively.

[0034] Moreover, the acceleration sensor 1 detects the acceleration at a predetermined sampling frequency and continuously supplies the detected acceleration to the calculation unit 2. Note that the sampling frequency refers to the number of times the acceleration sensor 1 detects the acceleration per second.

[0035] The acceleration of a walking user fluctuates roughly periodically, with one period being the time it takes for the user to take one step. Therefore, it is preferable that the sampling frequency of the acceleration sensor 1 is sufficiently large compared to the number of steps per second taken by the user.

[0036] The calculation unit 2 acquires the detected acceleration from the acceleration sensor 1. Then, based on the acceleration, the calculation unit 2 determines whether or not the user is performing abnormal behavior, and outputs the determination result. The calculation unit 2 executes various processes described below by reading and executing software (computer programs) from the storage unit 5. The calculation unit 2 may be, for example, a microprocessor, an MPU (Micro Processing Unit), or a CPU (Central Processing Unit). The calculation unit 2 may include multiple processors.

[0037] 4 is a block diagram for explaining the configuration of the calculation unit 2. The calculation unit 2 includes an acceleration processing unit 21, a walking determination unit 22, a blocking unit 23, a threshold calculation unit 24, a behavior determination unit 25, and an output unit .

[0038] The acceleration processing unit 21 acquires the detected acceleration from the acceleration sensor 1. The acceleration processing unit 21 processes the acquired detected acceleration and calculates a value indicating the magnitude of the acceleration of the user. In particular, in this embodiment, a value indicating the magnitude of the body movement acceleration component is calculated as the value indicating the magnitude of the acceleration of the user. More specifically, the acceleration processor 21 performs earth gravitational force removal processing, non-negative calculation processing, and smoothing processing on the detected acceleration, and calculates a value indicating the magnitude of the body movement acceleration component.

[0039] The earth's gravity removal process is a process for removing the gravitational acceleration component from the acceleration detected by the acceleration sensor 1. The acceleration processing unit 21 first extracts the gravitational acceleration component included in the acceleration detected by the acceleration sensor 1. For example, the acceleration processing unit 21 may extract the gravitational acceleration component by processing the detected acceleration with a low-pass filter.

[0040] As described above, the detected acceleration includes a body motion acceleration component and a gravitational acceleration component. Here, the body motion acceleration component periodically varies in accordance with the walking motion of the user, whereas the gravitational acceleration component is constant. Therefore, when the acceleration sensor processes the detected acceleration with a low-pass filter, the periodically varying body motion acceleration component is removed, and the gravitational acceleration component can be extracted.

[0041] The acceleration processing unit 21 may extract the gravitational acceleration component by calculating an average value of a predetermined number of samples of the detected acceleration and processing the calculated average value with a low-pass filter. Furthermore, the acceleration processing unit 21 may process a predetermined number of samples of detected accelerations using a low-pass filter and calculate the average value thereof to extract the gravitational acceleration component. With such a configuration, the body motion acceleration component is more effectively removed from the detected acceleration, and as a result, the accuracy of extracting the gravitational acceleration component is improved.

[0042] Next, the acceleration processor 21 calculates a body motion acceleration component based on the detected acceleration and the gravitational acceleration component. The acceleration processor 21 may calculate the body motion acceleration component by, for example, calculating the difference between the detected acceleration and the gravitational acceleration component.

[0043] Furthermore, the acceleration processing unit 21 may take an inner product of the detected acceleration and a gravitational acceleration component converted into a unit vector, and calculate the value of the inner product as the magnitude of the body motion acceleration component. The magnitude of the body motion acceleration component calculated in this manner corresponds, strictly speaking, to the magnitude of the acceleration component in the direction parallel to the earth's gravitational force among the acceleration components resulting from the user's walking. When the user behaves abnormally, the magnitude of the acceleration component in the direction parallel to the earth's gravitational force among the acceleration components resulting from the user's walking also takes a value different from that when walking. Therefore, the magnitude of the body motion acceleration component calculated in this manner can also be used as a judgment value for judging whether the user is behaving abnormally. By using the body movement acceleration component calculated in this manner, it is possible to reduce noise components in the acceleration that occur when, for example, the behavior determining device 1002 collides with the body of the user.

[0044] When the earth's gravity removal process is performed by the method described above, the acceleration processing unit 21 can calculate the body movement acceleration component regardless of the orientation of the behavior determination device 1002 carried by the user. In other words, by performing the earth's gravity removal process by the method described above, the user does not need to pay attention to the orientation of the behavior determination device 1002.

[0045] Next, the acceleration processor 21 executes a non-negative calculation process. The non-negative calculation process is a process for calculating a value indicating the magnitude of a body movement acceleration component expressed as a vector quantity. The acceleration processing unit 21 may use the squared value of the magnitude of the body motion acceleration component as the value indicating the magnitude of the body motion acceleration component. In this case, the value indicating the magnitude of the body motion acceleration component can be calculated by taking the sum of the squared values ​​of the x-axis component, y-axis component, and z-axis component of the body motion acceleration component. The acceleration processor 21 may use the magnitude of the body motion acceleration component as the value indicating the magnitude of the body motion acceleration component. In this case, the value indicating the magnitude of the body motion acceleration component can be calculated by taking the square root of the sum of the values ​​obtained by squaring the values ​​of the x-axis component, the y-axis component, and the z-axis component of the body motion acceleration component.

[0046] The acceleration processor 21 may process the body motion acceleration component calculated by the earth's gravitational force removal process using a high-pass filter before executing the non-negative calculation process. In this case, the acceleration processor 21 can calculate a value indicating the magnitude of the amount of fluctuation of the body motion acceleration component. When a user behaves abnormally, the amount of variation in the body movement acceleration component also increases. Therefore, the abnormal behavior of the user can be detected even by using a value indicating the magnitude of the amount of variation in the body movement acceleration component as a threshold value.

[0047] Finally, the acceleration processing unit 21 executes a smoothing process. In the smoothing process, the acceleration processing unit 21 processes the value indicating the magnitude of the calculated body motion acceleration component using a low-pass filter. By executing the smoothing process, the acceleration processing unit 21 can reduce noise components included in the value indicating the magnitude of the calculated body motion acceleration component.

[0048] The acceleration processing unit 21 supplies a value indicating the magnitude of the calculated body movement acceleration component to the walking determination unit 22 and the behavior determination unit 25. In addition, the acceleration processing unit 21 outputs the processed acceleration value to the threshold calculation unit 24 via the cut-off unit 23. In the following description, the value indicating the magnitude of the calculated body movement acceleration component will simply be referred to as a value indicating the magnitude of the user's acceleration.

[0049] The walking determination unit 22 acquires a value indicating the magnitude of the acceleration of the user from the acceleration processing unit 21. Then, based on the value indicating the magnitude of the acceleration of the user and a first threshold value determined in advance, the walking determination unit 22 determines whether or not the user is walking. The walking determination unit 22 outputs the determination result to the blocking unit 23.

[0050] Here, the first threshold value may be, for example, a value indicating the magnitude of acceleration. In this case, the walking determination unit 22 may determine that the user is walking when the value indicating the magnitude of the user's acceleration is smaller than the first threshold value. The first threshold may be, for example, a value indicating time. In this case, the walking determination unit 22 may determine that the user is walking when the fluctuation period of the value indicating the magnitude of the user's acceleration is greater than the first threshold. However, the variation period of the value indicating the magnitude of the user's acceleration varies less among individuals than the value indicating the magnitude of the user's acceleration. Therefore, when a value indicating time is used as the first threshold value and the variation period of the value indicating the magnitude of the user's acceleration is used as the determination value, it is possible to determine whether the user is walking while suppressing the influence of differences in individual physical abilities.

[0051] The cut-off unit 23 acquires a determination result regarding whether or not the user is walking from the walking determination unit 22. In addition, the cut-off unit 23 acquires a value indicating the magnitude of the acceleration of the user from the acceleration processing unit 21, and supplies the value to the threshold calculation unit 24. Based on the determination result acquired from the walking determination unit 22, the cut-off unit 23 decides whether or not to supply a value indicating the magnitude of the acceleration of the user to the threshold calculation unit 24.

[0052] More specifically, the blocking unit 23 supplies a value indicating the magnitude of the acceleration of the user, which is calculated from the detected acceleration detected in a period in which it is determined that the user is walking, to the threshold calculation unit 24. On the other hand, the blocking unit 23 blocks the value indicating the magnitude of the acceleration of the user, which is calculated from the detected acceleration detected in a period in which it is determined that the user is not walking, without supplying it to the threshold calculation unit 24. In other words, the cutoff unit 23 supplies to the threshold calculation unit 24 only a value indicating the magnitude of the acceleration of the user during the period in which it is determined that the user is walking.

[0053] The threshold calculation unit 24 acquires a value indicating the magnitude of the acceleration of the user during the period in which it is determined that the user is walking from the blocking unit 23. Then, the threshold calculation unit 24 calculates a second threshold based on the acquired value indicating the magnitude of the acceleration of the user. The threshold calculation unit 24 outputs the calculated second threshold to the behavior determination unit 25.

[0054] As described above, the value indicating the magnitude of the user's acceleration is calculated based on the acceleration detected by the acceleration sensor 1. Therefore, it can be said that the threshold calculation unit 24 calculates the second threshold based on the acceleration detected during the period in which the walking determination unit 22 determines that the user is walking, among the accelerations detected by the acceleration sensor 1.

[0055] For example, the threshold calculation unit 24 may calculate the second threshold at the timing when a predetermined number of values ​​indicating the magnitude of the acceleration of the user are acquired from the blocking unit 23. In this case, the threshold calculation unit 24 may calculate the average value and standard deviation of the values ​​indicating the magnitude of the acceleration of the user. Then, the threshold calculation unit 24 may multiply the calculated standard deviation by a predetermined coefficient and add the result to the calculated average value to calculate the second threshold. According to this configuration, the second threshold is a value calculated based on the acceleration of the user while walking, and is therefore a value appropriately adjusted to the physical ability of the individual user. Therefore, when the second threshold calculated in this manner is used to detect abnormal behavior of the user, the abnormal behavior of the user can be detected more appropriately than when a default value is used as the threshold.

[0056] Furthermore, the threshold calculation unit 24 may extract larger values ​​from the values ​​indicating the magnitude of acceleration of a predetermined number of users at a predetermined rate. In this case, the threshold calculation unit 24 calculates the average value and standard deviation of the extracted values. Then, the threshold calculation unit 24 may multiply the calculated standard deviation by a predetermined coefficient and add the result to the calculated average value to calculate the second threshold value. When abnormal behavior determination is performed, the second threshold is set as an upper limit of the value indicating the magnitude of acceleration in normal walking. Therefore, when the second threshold is calculated as described above, the second threshold can be calculated more appropriately.

[0057] The threshold calculation unit 24 may calculate a plurality of thresholds having different values ​​as the second threshold. Then, the behavior determination unit 25 described later may evaluate the magnitude of the acceleration of the user based on the plurality of thresholds, and the warning unit 3 described later may change the warning mode for the user based on the evaluation result. With this configuration, when the behavior determination device 1002 detects abnormal behavior of a user, it can appropriately evaluate the risk of the abnormal behavior and issue an appropriate warning to the user in accordance with the risk of the abnormal behavior.

[0058] The threshold calculation unit 24 may calculate the second threshold continuously, or may calculate the second threshold every predetermined period. In other words, the threshold calculation unit 24 may update the second threshold continuously, or may update the second threshold every predetermined period. Furthermore, the threshold calculation unit 24 may update the second threshold when an instruction to update the second threshold is received from the user.

[0059] The behavior determination unit 25 obtains a value indicating the magnitude of the user's acceleration from the acceleration processing unit 21, and obtains a second threshold from the threshold calculation unit 24. Then, the behavior determination unit 25 determines whether or not the user is performing abnormal behavior based on the value indicating the magnitude of the user's acceleration and the second threshold. The behavior determination unit 25 outputs the determination result to the output unit 26. As described above, the value indicating the magnitude of the user's acceleration is a value calculated based on the acceleration detected by the acceleration sensor 1. Therefore, it can be said that the behavior determination unit 25 determines whether or not the user is performing abnormal behavior based on the acceleration detected by the acceleration sensor 1 and the second threshold value calculated by the threshold value calculation unit 24.

[0060] More specifically, when a value indicating the magnitude of the user's acceleration is greater than the second threshold calculated by the threshold calculation section 24, the behavior determination section 25 determines that the user is performing abnormal behavior. In this way, when the behavior determination unit 25 uses a value indicating the magnitude of the user's acceleration as the determination value, it is possible to reduce the time it takes from when the user performs abnormal behavior to when the abnormal behavior is detected, compared to when the behavior determination unit 25 uses the fluctuation period of the value indicating the magnitude of the user's acceleration as the determination value.

[0061] When the behavior determination unit 25 acquires a plurality of thresholds as the second threshold from the threshold calculation unit 24, the behavior determination unit 25 may evaluate the magnitude of the acceleration detected by the acceleration sensor 1 based on the plurality of thresholds. In other words, the behavior determination unit 25 may evaluate a value indicating the magnitude of the acceleration of the user based on the plurality of thresholds.

[0062] The behavior determination unit 25 may evaluate the value indicating the magnitude of the user's acceleration in stages or by dividing it into levels based on the acquired multiple thresholds. For example, the behavior determination unit 25 may evaluate that the risk level of the value indicating the magnitude of the user's acceleration is at level n or the nth stage when the value indicating the magnitude of the user's acceleration is equal to or greater than the n-th smallest threshold and smaller than the n+1-th smallest threshold, where n is an integer.

[0063] The output unit 26 outputs the determination result by the behavior determination unit 25. More specifically, the output unit 26 obtains the determination result as to whether or not the user is performing abnormal behavior from the behavior determination unit 25, and outputs the determination result to some or all of the warning unit 3, the storage unit 5, and the transmission unit 6. Note that the output unit 26 may notify some or all of the warning unit 3, the storage unit 5, and the transmission unit 6 that the user has performed abnormal behavior only when the behavior determination unit 25 determines that the user is performing abnormal behavior.

[0064] Returning to the explanation of Figure 3. The warning unit 3 obtains a determination result as to whether or not the user is behaving abnormally from the calculation unit 2. Then, when the behavior determination unit 25 determines that the user is behaving abnormally, the warning unit 3 notifies the user of a warning.

[0065] The warning unit 3 may be configured as, for example, an audio speaker, a flashing light, or a combination of these. When the warning unit 3 includes an audio speaker and the behavior determination unit 25 determines that the user is performing abnormal behavior, the warning unit 3 may play a warning sound such as a siren to the user. In this case, the warning unit 3 may play a warning voice message to the user. Note that data recording the warning voice message may be stored in the storage unit 5, for example.

[0066] When the threshold calculation unit 24 calculates multiple thresholds with different values ​​as the second threshold and the behavior determination unit 25 evaluates the magnitude of the acceleration detected by the acceleration sensor 1 based on the multiple thresholds, the warning unit 3 may change the warning manner for the user based on the evaluation result of the behavior determination unit 25.

[0067] For example, the warning unit 3 may change the warning voice message to be notified to the user depending on the evaluation result of the behavior determination unit 25. For example, when the magnitude of the user's acceleration is small, i.e., when the risk of abnormal behavior is low, the warning unit 3 may notify the user of a message such as "Please walk carefully" to issue a warning to the extent of urging caution. Also, when the magnitude of the user's acceleration is large, i.e., when the risk of abnormal behavior is high, the warning unit 3 may notify the user of a message such as "Please stop" to issue a warning to urge the user to stop.

[0068] The position information acquisition unit 4 uses an artificial satellite such as a Global Navigation Satellite System (GNSS) to acquire position information of the behavior determination device 1002. The position information acquisition unit 4 outputs the acquired position information to the storage unit 5 and the transmission unit 6.

[0069] The storage unit 5 is a storage device realized by using a RAM (Random Access Memory), a ROM (Read Only Memory), etc. The storage unit 5 stores programs for controlling the operation of the behavior determination device 1002. The calculation unit 2 reads out these programs from the storage unit 5 and executes them to execute the various processes described above.

[0070] The program includes a set of instructions (or software code) for making the computer perform one or more functions described in the embodiment when the program is read into the computer. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, the computer-readable medium or the tangible storage medium includes RAM, ROM, flash memory, SSD or other memory technology, CD-ROM, DVD (digital versatile disc), Blu-ray (registered trademark) disk or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. The program may be transmitted on a transitory computer-readable medium or a communication medium. By way of example and not limitation, the transitory computer-readable medium or the communication medium includes electrical, optical, acoustic, or other forms of propagating signals.

[0071] Furthermore, when the behavior determination unit 25 determines that the user is performing abnormal behavior, the storage unit 5 acquires the determination result from the calculation unit 2 and stores it. Furthermore, when the behavior determination unit 25 determines that the user has performed abnormal behavior, the storage unit 5 acquires the location information of the location where the determination was made from the location information acquisition unit 4. Then, the storage unit 5 associates the determination result with the location information and stores them. In addition, it is preferable that the stored determination result and location information are configured to be read out from the storage unit 5 by the user as appropriate. According to this configuration, it is possible to grasp the location where the user has engaged in abnormal behavior, which in turn makes it easier to grasp dangerous locations where the user is likely to engage in abnormal behavior, thereby making it possible to prevent accidents from occurring.

[0072] When the behavior determination unit 25 determines that the user is not walking normally, the transmission unit 6 acquires the determination result from the calculation unit 2 and transmits it to the external server device. Furthermore, when the behavior determination unit 25 determines that the user has performed abnormal behavior, the transmission unit 6 acquires the location information of the location where the determination was made from the location information acquisition unit 4. Then, the transmission unit 6 links the determination result with the location information and transmits them to an external server device. With this configuration, the location where the user engaged in abnormal behavior can be shared with others. Furthermore, by aggregating location information of locations where abnormal behavior occurred from multiple users, it becomes easier to grasp dangerous locations where people are likely to engage in abnormal behavior, and it becomes easier to implement accident prevention measures such as improving road conditions at dangerous locations.

[0073] <Operation of the behavior judgment system> Next, the operation of the behavior determination system, i.e., the behavior determination method according to the second embodiment, will be described in detail. Figures 5 and 6 are flowcharts for explaining the operation of the behavior determination device 1002 according to the second embodiment. In the following description, Figures 3 and 4 will be referred to as appropriate.

[0074] First, a detailed description will be given of Fig. 5. Fig. 5 is a flowchart showing an operation of the behavior determination device 1002 to calculate or update the second threshold. In the operation of calculating or updating the second threshold, first, the acceleration sensor 1 detects acceleration (step S201).

[0075] Next, the acceleration processing unit 21 calculates a value indicating the magnitude of the acceleration of the user (step S202). More specifically, the acceleration detected in step S201 is subjected to an earth gravitational force removal process, a non-negative calculation process, and a smoothing process to calculate a value indicating the magnitude of the acceleration of the user.

[0076] Next, the walking determination unit 22 determines whether or not the fluctuation period of the value indicating the magnitude of the acceleration of the user is greater than a first threshold value (step S203). When the fluctuation period of the value indicating the magnitude of the user's acceleration is smaller than the first threshold value (NO in step S203), the walking determination unit 22 determines that the user is not walking, and the behavior determination device 1002 executes step S201 again. When the fluctuation period of the value indicating the magnitude of the acceleration of the user is greater than the first threshold value (YES in step S203), the walking determination unit 22 determines that the user is walking, and executes step S204.

[0077] Next, the acceleration processing unit 21 supplies a value indicating the magnitude of the acceleration of the user to the threshold calculation unit 24 via the blocking unit 23 (step S204). More specifically, the acceleration processing unit 21 supplies to the threshold calculation unit 24 a value indicating the magnitude of the acceleration detected during the period in which it is determined that the user is walking. The behavior determining device 1002 may execute the operations of steps S201 to S204 multiple times before executing step S205. For example, the behavior determining device 1002 may repeatedly execute the operations of steps S201 to S204 until a predetermined number of values ​​indicating the magnitude of acceleration are supplied to the threshold calculation unit 24. Finally, the threshold calculation unit 24 calculates a second threshold (step S205), and the behavior determining device 1002 ends the series of operations.

[0078] The behavior determining device 1002 may perform the operations of steps S201 to S205 continuously or at predetermined intervals. Moreover, the behavior determining device 1002 may perform the operations of steps S201 to S205 when a user requests updating of the second threshold. The behavior determining device 1002 may execute the operations of steps S201 to S205 in parallel with the operations shown in the flowchart of FIG.

[0079] Next, Fig. 6 will be described in detail. Fig. 6 is a flowchart showing the operation of the behavior determining device 1002 to detect abnormal behavior of a user. The operation of detecting abnormal behavior of a user begins with the acceleration sensor 1 detecting acceleration (step S301).

[0080] Next, the acceleration processing unit 21 calculates a value indicating the magnitude of the acceleration of the user (step S302). More specifically, the acceleration detected in step S301 is subjected to an earth gravitational force removal process, a non-negative calculation process, and a smoothing process to calculate a value indicating the magnitude of the acceleration of the user. However, the operations of steps S301 and S302 may be replaced by the operations of steps S201 and S202.

[0081] Next, the behavior determination unit 25 determines whether or not the value indicating the magnitude of the acceleration of the user is greater than a second threshold value (step S303). If the value indicating the magnitude of the user's acceleration is smaller than the second threshold (NO in step S303), the behavior determination unit 25 determines that the user is not performing abnormal behavior, and the behavior determination device 1002 executes step S301 again. That is, the behavior determination device 1002 repeatedly executes steps S301 to S303 until the value indicating the magnitude of the user's acceleration becomes larger than the second threshold. If the value indicating the magnitude of the user's acceleration is greater than the second threshold value (YES in step S303), the behavior determination unit 25 determines that the user is performing abnormal behavior, and executes the operations in and after step S304.

[0082] Next, the warning unit 3 warns the user (step S304). Then, the location information acquisition unit 4 acquires location information (step S305), and the storage unit 5 stores the determination result and the location information (step S306). Finally, the transmission unit 6 transmits the determination result and the location information to the external server device (step S306), and the behavior determination device 1002 ends the series of operations. Note that steps S304 to S307 may be executed in a different order or in parallel, but steps S306 and S307 are executed after step S304. The behavior determining device 1002 repeatedly executes the operations of steps S301 to S307.

[0083] As described above, the behavior determination device 1002 according to this embodiment detects the acceleration of a user, and determines whether or not the user is walking using a first threshold. Then, the behavior determination device 1002 calculates a second threshold based on the acceleration detected during the period in which the user is determined to be walking, and determines whether or not the user is performing a moving behavior based on the second threshold. With this configuration, the behavior determination device 1002 can set an appropriate threshold for each user, even for the magnitude of acceleration, which varies greatly from user to user. As a result, the behavior determination device 1002 according to this embodiment can appropriately detect abnormal behavior of a pedestrian.

[0084] While the present invention has been described above with reference to the above-described embodiment, the present invention is not limited to the configuration of the above-described embodiment, and of course includes various modifications, alterations, and combinations that may be made by a person skilled in the art within the scope of the invention as defined in the claims of the present application.

[0085] A part or all of the above-described embodiments can be described as, but is not limited to, the following supplementary notes. (Appendix 1) An acceleration sensor that detects the acceleration of a user; a walking determination unit that determines whether or not the user is walking based on the acceleration detected by the acceleration sensor and a predetermined first threshold value; a threshold calculation unit that calculates a second threshold based on the acceleration detected during a period in which the walking determination unit determines that the user is walking, among the accelerations detected by the acceleration sensor; and a behavior determination unit that determines whether or not the user is performing abnormal behavior based on the acceleration detected by the acceleration sensor and the second threshold calculated by the threshold calculation unit; An output unit that outputs a determination result by the behavior determination unit. Behavior decision system. (Appendix 2) the walking determination unit determines that the user is walking when a period of change in a value indicating a magnitude of the acceleration of the user is greater than the first threshold value; The behavior determination unit determines that the user is performing abnormal behavior when a value indicating a magnitude of the acceleration of the user is greater than the second threshold calculated by the threshold calculation unit. 2. A behavior determination system according to claim 1. (Appendix 3) and a warning unit that notifies the user of a warning when the behavior determination unit determines that the user is performing an abnormal behavior. 3. The behavior determination system according to claim 2. (Appendix 4) the threshold calculation unit calculates a plurality of thresholds having different values ​​as the second threshold, the behavior determination unit evaluates a magnitude of the acceleration detected by the acceleration sensor based on a plurality of thresholds calculated by the threshold calculation unit; the warning unit changes a manner of warning given to the user based on a result of the evaluation by the behavior determination unit. 4. The behavior determination system according to claim 3. (Appendix 5) The method further includes a storage unit configured to store a result of the determination made by the behavior determination unit when the behavior determination unit determines that the user is performing an abnormal behavior. 3. The behavior determination system according to claim 2. (Appendix 6) When the behavior determination unit determines that the user is performing abnormal behavior, the storage unit further stores location information of a point where the determination result is output. 6. The behavior determination system according to claim 5. (Appendix 7) and a transmission unit configured to transmit a result of the determination to an external server device when the behavior determination unit determines that the user is not walking normally. 3. The behavior determination system according to claim 2. (Appendix 8) When the behavior determination unit determines that the user is not walking normally, the transmission unit further transmits position information of a point where the determination result is output. 8. The behavior determination system according to claim 7. (Appendix 9) An acceleration sensor that detects the acceleration of a user; a walking determination unit that determines whether or not the user is walking based on the acceleration detected by the acceleration sensor and a predetermined first threshold value; a threshold calculation unit that calculates a second threshold based on the acceleration detected during a period in which the walking determination unit determines that the user is walking, among the accelerations detected by the acceleration sensor; and a behavior determination unit that determines whether or not the user is performing abnormal behavior based on the acceleration detected by the acceleration sensor and the second threshold calculated by the threshold calculation unit; An output unit that outputs a determination result by the behavior determination unit. Behavior determination device. (Appendix 10) Detect the user's acceleration, determining whether the user is walking based on the detected acceleration and a first predetermined threshold value; Calculating a second threshold value based on the acceleration detected during a period in which it is determined that the user is walking, among the detected accelerations; determining whether the user is performing an abnormal behavior based on the detected acceleration and the calculated second threshold value; Output the judgment result. Behavior judgment method. (Appendix 11) Detect the user's acceleration, determining whether the user is walking based on the detected acceleration and a first predetermined threshold value; Calculating a second threshold value based on the acceleration detected during a period in which it is determined that the user is walking, among the detected accelerations; determining whether the user is performing an abnormal behavior based on the detected acceleration and the calculated second threshold value; Outputting the determination result, causing a computer to execute processing, Behavioral assessment program. [Explanation of symbols]

[0086] 1 Acceleration sensor 2 Arithmetic section 3 Warning section 4 Location information acquisition section 5 Storage section 6. Transmitter 21 Acceleration processing unit 22 Walking Judgment Section 23 Breaking section 24 Threshold calculation unit 25 Behavior Judgment Department 26 Output section 1001 Behavior Judgment System 1002 Behavior determination device

Claims

1. An acceleration sensor that detects the acceleration of a user; an acceleration processing unit that decomposes the acceleration detected by the acceleration sensor into a body motion acceleration component and a gravitational acceleration component, and extracts a magnitude of the body motion acceleration component as a value indicating a magnitude of acceleration resulting from the walking of the user; a walking determination unit that determines whether or not the user is walking based on a period of change in a value indicating a magnitude of acceleration resulting from walking of the user detected by the acceleration sensor and a predetermined first threshold value; a blocking unit that extracts only a value indicating a magnitude of the acceleration of the user detected by the acceleration sensor during a period in which the walking determination unit has determined that the user is walking, and blocks other values; a threshold calculation unit that calculates a second threshold based only on the value indicating the magnitude of the acceleration extracted by the blocking unit; a behavior determination unit that determines whether or not the user is performing abnormal behavior based on the acceleration detected by the acceleration sensor and the second threshold calculated by the threshold calculation unit; and a warning unit that notifies the user of a warning when the behavior determination unit determines that the user is performing an abnormal behavior. Behavior determination device.

2. The fluctuation period refers to a period of fluctuation in acceleration caused by walking, which is a type of repetitive motion. The behavior determination device according to claim 1 .

3. The acceleration processing unit performs an earth gravitational force removal process, a non-negative calculation process, and a smoothing process on the acceleration detected by the acceleration sensor, and calculates a value indicating the magnitude of a body motion acceleration component. The behavior determination device according to claim 1 .

4. the acceleration processing unit takes an inner product of the acceleration detected by the acceleration sensor and a unit vector of a gravitational acceleration component, and calculates the value of the inner product as a magnitude of a body motion acceleration component; The behavior determination device according to claim 1 .

5. The non-negative calculation process performed by the acceleration processing unit is a process of calculating a value indicating the magnitude of the body motion acceleration component expressed as a vector quantity, and the value indicating the magnitude of the body motion acceleration component is calculated as the square root of the sum of the value obtained by squaring the x-axis component, the value obtained by squaring the y-axis component, and the value obtained by squaring the z-axis component of the body motion acceleration component. The behavior determination device according to claim 3 .

6. the acceleration processing unit processes the body motion acceleration component calculated by the earth's gravitational force removal process using a high-pass filter, and then performs a non-negative calculation process; The behavior determination device according to claim 3 .

7. the behavior determination unit determines that the user is performing abnormal behavior when a value indicating a magnitude of the acceleration of the user is greater than the second threshold calculated by the threshold calculation unit. The behavior determination device according to claim 1 .

8. a location information acquisition unit that acquires location information of a point where a determination result is output when the behavior determination unit determines that the user is performing abnormal behavior; a storage unit that stores a result of the determination and the location information when the behavior determination unit determines that the user is performing an abnormal behavior; a transmission unit that transmits a result of the determination and the location information to an external server device when the behavior determination unit determines that the user is performing abnormal behavior. The behavior determination device according to claim 1 .

9. Detect the user's acceleration, The detected acceleration is decomposed into a body motion acceleration component and a gravitational acceleration component; extracting a magnitude of the body motion acceleration component as a value indicating a magnitude of acceleration resulting from the user's walking; determining whether the user is walking based on a fluctuation period of the value indicating the magnitude of the acceleration resulting from the walking of the user and a predetermined first threshold value; extracting only a value indicating a magnitude of the detected acceleration of the user during a period in which it is determined that the user is walking, and blocking others; calculating a second threshold value based only on the value indicating the magnitude of the acceleration extracted during the period in which it is determined that the user is walking; determining whether the user is performing an abnormal behavior based on the detected acceleration and the calculated second threshold value; When it is determined that the user is performing an abnormal behavior, a warning is issued to the user. Behavior judgment method.

10. causing a behavior determination device to execute the behavior determination method according to claim 9; Behavioral assessment program.

Images