WO2010029918A1

WO2010029918A1 - Behavior analysis system and computer program

Info

Publication number: WO2010029918A1
Application number: PCT/JP2009/065668
Authority: WO
Inventors: 光雄松本; 博文関根
Original assignee: 株式会社マクロウェア
Priority date: 2008-09-09
Filing date: 2009-09-08
Publication date: 2010-03-18
Also published as: JP5572093B2; JPWO2010029918A1

Abstract

A behavior analysis system for efficiently collecting data required for behavior analysis while psychological and physical burdens on the subject are reduced to a minimum, and making an analysis without exerting a burden on the observer, and quickly and flexibly coping with a change about the area in which the subject moves. The behavior analysis system includes a predetermined transmitter/receiver worn by the subject, position determining devices capable of determining that the subject is present near by means of radio communication with the transmitter/receiver, a storage device storing map data about a predetermined area and position data about the position determining devices the positions of which are shown on the map, a limb sensor worn by the subject for continuously detecting the movement of a leg or an arm of the subject, and a movement detection data processing device for calculating movement estimation data used for estimating where the subject moves within the area with the lapse of time using data about the movement of the leg or the arm detected by the limb sensor, the data about the subject position determination by the position determining devices, the map data, and the position data.

Description

Behavior analysis system and computer program

The present invention relates to a technique for estimating a behavior and analyzing a behavior of a worker or a working vehicle working in a factory building or a store, or a movement of a customer in a store.

For example, a method using ultrasonic waves has been proposed as a technique for tracking the working behavior of workers or working vehicles working in a factory building or store. Further, there is a technology that applies RFID (Radio Frequency Identification) used for management in electronic money and physical distribution to specify identification positions of workers, work wheels, and the like in a factory. It is used in the security field, such as entering and leaving a factory or office.
For example, it is a technique described in Patent Document 1.

This technology is suitable for simple analysis such as passing through fixed points. In other words, it is a technique aimed at detecting the presence of an object or person, and its route cannot be estimated.

JP 2002-128277 A

A global positioning system (GPS) is widely used for position estimation or path estimation. Recently, it is also installed in mobile phones, and it can be applied to humans because it is inexpensive and lightweight.
However, since it is necessary to receive a signal from a satellite, it is difficult to use indoors.

There is also a technology called Inertial Navigation System (INS). This technology enables high-accuracy position identification and behavior estimation, such as airplane autopilot. However, the device becomes large if sufficient accuracy is to be obtained. For this reason, sufficient accuracy cannot be obtained with a size that allows the device to be worn by a person who is not to be measured. Furthermore, there is a principle problem that errors accumulate and increase with the moving distance.

From the above, the estimation of the movement route and the behavior analysis of the worker for the purpose of improving the work efficiency have been mainly performed by the accompanying of the observer or the video image. When accompanied by an observer, detailed information can be collected in real time. In addition, the method using video shooting allows repeated observation and comparison with other results.

It is expected that accompanying the observer currently employed will affect the behavior of the subject.
Furthermore, these methods not only impose a heavy burden on the observer, but also cannot avoid subjective intervention.
Recording by video shooting can eliminate the burden on the observer and prevent subjective intervention, but there is a problem that it takes a lot of time to analyze the video that has been shot.

The problem to be solved by the present invention is to efficiently collect data necessary for behavioral analysis while minimizing the psychological and physical burden on the subject and analyze it without giving a burden to the observer. It is possible to provide a technique capable of responding quickly and flexibly to changes relating to the range in which a subject acts.

The object of the invention described in claims 1 to 4 is to efficiently collect data necessary for behavior analysis while minimizing the psychological and physical burden on the subject and to burden the observer. It is an object of the present invention to provide a behavior analysis system that can perform analysis without giving any change, and can respond quickly and flexibly to changes in the range in which a subject acts.
The object of the invention described in claims 5 to 8 is to efficiently collect data necessary for behavior analysis while minimizing psychological and physical burden on the subject, and to burden the observer. It is an object of the present invention to provide an action analysis program that can be analyzed without giving any change, and that can quickly and flexibly cope with changes related to the range in which the subject behaves.

(First invention)
A first invention in the present application relates to a behavior analysis system for a subject that calculates how a subject moves within a predetermined region.
That is, a predetermined transmitter / receiver worn by the subject, a plurality of position specifying devices capable of specifying that the subject exists in the vicinity by wireless communication with the transmitter / receiver, map data regarding the predetermined region, and A storage device that pre-stores position data of the position specifying device in map data; a limb sensor that is worn on a subject and continuously detects the movement of a leg or arm; and a leg and / or that the limb sensor detects Motion for calculating motion estimation data for estimating where the subject moved over time in the region using motion data relating to body movement, subject specifying data by the location specifying device, the map data, and the location data A detection data processing device.

(Glossary)
The “subject” is, for example, a worker who assembles a product in a factory, a worker who performs loading / unloading or inspection of goods in a distribution warehouse, a shopper who comes to shopping, and the like. For example, when the subject works on a forklift, the subject and the forklift are integrated, and the “transceiver” may be attached to either the subject or the forklift.

“Transmitter / receiver” refers to a wireless communication device that has a function of receiving a signal transmitted from a position specifying device and returning a signal to the position specifying device. Those which do not require a power source are preferable because they are small and light. This is because the burden on the wearing subject is reduced. Note that the “signal” is a signal using radio waves, sound waves, ultrasonic waves, light, infrared rays, or the like, and may be a single signal or a combination of signals.
It is desirable that the “transceiver” is built in or integrated with the “limb sensor”. For example, a mobile phone incorporating a short-range wireless communication module as a “transceiver” and also incorporating a “limb sensor” is preferable as a “transceiver” according to the behavior analysis system of the present application.

The “limb sensor” refers to the movement of the subject's leg or arm and / or the body movement by an acceleration sensor capable of measuring acceleration, a gyro sensor capable of measuring angular velocity, or a combination of these sensors. Detect. The subject's attachment site is the “limb body” and either or both of acceleration and three-dimensional movement are detected. The movement of the subject's “walk” (how fast is one step at a time) , Etc.). Conventionally, in order to measure the vicinity of the center of gravity of the subject, a technique of attaching to the waist is generally used, and a gyro sensor capable of measuring angular velocity is generally not included.
If the goal is to measure only the action trajectory, the limb body sensor is attached to the subject at the leg, such as the thigh and ankle, where the measurement error is higher than the arm, such as the upper arm or wrist. Hard to come out. This is because the arm part may move other than walking, and the acceleration and angular velocity (three-dimensional movement) measured in that case are irrelevant to the walking action.

Now, for example, when the subject is a forklift crew, the gyro sensor may be unnecessary. This is because it is not necessary to detect the movement of the vertical movement when the purpose is to grasp the movement of the forklift. However, it is better to equip the gyro sensor when the purpose is to grasp the movement of the subject including getting on and off the forklift.
The acceleration and gyro signal measured by the “limb body sensor” are recorded in association with the passage of time. Therefore, a timer for recording the passage of time from the start of measurement, or a clock for adding time data to the measured acceleration or gyro signal may be incorporated. The storage device for recording may be stored in a detachable storage device provided inside the limb sensor, or when a predetermined amount of data is stored, it is transmitted to the means for arithmetic processing via the communication means. Also good.
When the worker who walks without getting on the vehicle is the subject, either the acceleration sensor or the gyro sensor may be sufficient. However, even if it is a walking worker, if both the acceleration sensor and the gyro sensor are incorporated, and both data are acquired and processed, errors can be reduced or corrected in the detection of “steps” described later. It can be used and useful.

The “position specifying device” is based on bidirectional communication with the “transceiver” described above. If a transponder is used for the “transceiver”, the “transceiver” does not require a power source required for communication. This is effective for grasping the movement of the subject in a building where GPS (Global Positioning System) cannot be used.
The position specifying device is not limited to be installed only in a place where GPS cannot be used. For example, it is naturally possible to install a position identifying device outside a factory (outdoors where GPS functions) for a subject who moves in and out of the factory.

(Function)
Map data relating to a predetermined area and position data of the position specifying device in the map data are stored in advance in a storage device.
The subject wears a transceiver and a limb sensor and enters a predetermined area. When the position specifying device provided in the predetermined area detects the transmitter / receiver attached to the subject, the position specifying device and the transmitter / receiver communicate bidirectionally. The transceiver or the location specifying device records the bidirectional communication in association with the passage of time.
The subject is equipped with a limb sensor that continuously detects the movement of the leg or arm, and acquires movement data regarding the movement of the leg and / or the body. Using the motion data, subject identification data by the location identification device, the map data, and the location data, motion estimation data is estimated by the motion detection data processing device to estimate where the subject has moved over time in the region. calculate.

Since there is no observer to observe the subject, the psychological burden on the subject can be suppressed, and the burden on the observer is not given. In addition, data necessary for behavior analysis can be efficiently collected and analyzed. Moreover, even if there is a change regarding the range in which the subject behaves, it is possible to respond quickly and flexibly by updating the setting of the position specifying device and the map data regarding the behavior range.

(Second invention)
The second invention in the present application also relates to a behavior analysis system for a subject that calculates how the subject moves within a predetermined region.
That is, using the movement data relating to leg and / or body movement detected by the limb sensor, the identification data of the subject by the position identification device, the map data, and the position data, the total distance that the subject has moved within the area is calculated. The present invention relates to a behavior analysis system including a distance calculation device that calculates total distance data to be estimated. The transmitter / receiver, the position specifying device, the storage device, and the limb body sensor worn by the subject are the same as in the configuration of claim 1.

(Glossary)
The calculation algorithm for estimating the total distance is calculated from the total distance along the path between the position detection devices (MS) detected by the limb sensor and the position estimated from the relative stride. For example, when the path is such that the subject cannot move in a straight line within the predetermined area, it is desirable to be able to calculate the total distance along the path.

(Function)
According to the subject behavior analysis system in the second aspect of the present invention, motion data relating to leg and / or body movement detected by a limb sensor, subject identification data by the location identification device, map data, and location data are used. The total distance data for estimating the total distance that the subject moved within the region can be calculated.

(Third invention)
The third invention is a subject behavior analysis system having both the motion detection data processing device according to the first invention and the distance calculation device according to the second invention. The apparatus, the storage device, and the limb body sensor are the same as those in the first and second aspects.

(Function)
According to the behavior analysis system according to the third aspect of the invention, using the movement data relating to the movement of the leg and / or the body detected by the limb sensor, the identification data of the subject by the position identification device, the map data, and the position data, The motion detection data processing device can calculate motion estimation data for estimating where the subject has moved over time in the region and total distance data for estimating the total distance the subject has moved in the region.

(Variations of the first and third inventions)
The first invention and the third invention can also provide the following variations.
That is, the limb body sensor includes a gyro sensor, and the motion detection data processing device calculates using the maximum value of the angular velocity detected by the gyro sensor or a value in the vicinity including the maximum value, and uses the relative stride. It is also possible to provide a behavior analysis system that calculates a subject's movement.

(Function)
By calculating the relative stride and providing an estimation algorithm using the relative stride, it is possible to more accurately estimate where the subject has moved over time.

(Fourth invention)
A fourth invention according to the present application relates to a behavior analysis program for a subject in which a computer calculates how the subject moves within a predetermined region.
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. Data input procedure for inputting motion data from a limb sensor that continuously detects motion, map data relating to the predetermined area, and data in which the position specifying device in the map data is stored in a storage device in advance A storage procedure; and a motion detection data processing procedure for calculating motion estimation data for estimating where the subject has moved over time in the region using the motion data, the map data, and the position data. A computer program for execution.

(Fifth invention)
A fifth invention according to the present application relates to a behavior analysis program for a subject that uses a computer to calculate how the subject moves within a predetermined region.
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. Data input procedure for inputting motion data from a limb sensor that continuously detects motion, map data relating to the predetermined area, and data in which the position specifying device in the map data is stored in a storage device in advance A computer for causing a computer to execute a storage procedure and a distance calculation procedure for calculating total distance data for estimating a total distance that the subject has moved in the region using the motion data, the map data, and the position data It is a program.

(Sixth invention)
A sixth invention according to the present application relates to a behavior analysis program for a subject that uses a computer to calculate how the subject moves within a predetermined region.
The program is a behavior analysis program for a subject that uses a computer to calculate how the subject moves within a predetermined area.
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. Data input procedure for inputting motion data from a limb sensor that continuously detects motion, map data relating to the predetermined area, and data in which the position specifying device in the map data is stored in a storage device in advance Motion detection data processing for calculating motion estimation data for estimating where the subject moved over time in the region using the storage procedure, the motion data, the subject specific data, the map data, and the position data Using the procedure, motion data, the map data, and the position data, the subject moves in the area. A computer program for causing a computer to execute a distance calculation procedure for calculating total distance data for estimating a total distance.

(Variations of the fourth and sixth inventions)
The invention which limited the computer program concerning the 4th or 6th invention can also be provided.
That is, the invention according to the variation is that the motion detection data processing procedure is calculated using a maximum value of an angular velocity detected by a gyro sensor or a value in the vicinity including a maximum value, and from the position specifying device and the transceiver The computer program according to claim 5, wherein the calculation is performed using the calculated position data and the relative stride.

The computer program according to the fourth to sixth inventions and the computer program according to the above-described variations can be stored and distributed on a recording medium. It is also possible to download via a communication line such as the Internet. Examples of the recording medium include a semiconductor memory, a hard disk, a CD-R, an MO (magneto-optical disk), and a DVD-R.

According to the first to third inventions, data necessary for behavior analysis is efficiently collected while minimizing the psychological and physical burden on the subject without giving the observer a burden. It was possible to provide a behavior analysis system that can be analyzed and that can quickly and flexibly cope with changes related to the range in which the subject behaves.
According to the fourth to sixth inventions, it is possible to efficiently collect data necessary for behavioral analysis while minimizing the psychological and physical burden on the subject and without burdening the observer. It was possible to provide a behavior analysis program that can be analyzed and that can quickly and flexibly cope with changes related to the range in which the subject behaves.

It is a block diagram which shows the whole this invention. It is a conceptual diagram which shows the expression method of a channel | path and a specific position signal detection apparatus. It is a graph which shows the time passage of a position specific signal. It is a graph for the detection of "walk" by showing a gyro signal and time passage. It is a graph which shows the relationship between the instantaneous value from an acceleration sensor, and time. It is the graph which computed the relative distance from acceleration. It is a figure for demonstrating the procedure of a position estimation when a test subject walks the passage. It is a figure for demonstrating the procedure which estimates the position based on a movement distance. It is a figure which shows the variation regarding the principle of a position specification apparatus. It is the block diagram which caught this invention from the surface of the input / output of data. It is the block diagram which caught this invention from the surface of the input / output of data. It is a conceptual diagram which shows the channel | path where a test subject walks, a position specific apparatus installation position, and a landing position. It is a graph for speed detection of "walk" by showing a gyro signal and time passage. FIG. 14 is a partially enlarged view of 4 to 10 seconds in FIG. It is a figure which shows the position where the test subject walked and how the test subject's left and right legs advanced. It is a figure which shows a gyro signal when a test subject stops in the middle of a passage with time.

The configuration of this system is shown in FIG.
It is prepared as map data in a predetermined area such as a factory building or a store, and is stored in advance in a predetermined storage means (not shown).
In the predetermined area related to the prepared map data, a position specifying device (milestone, abbreviated as “MS” or “# ˜” in the figure) is provided at a predetermined point such as a passage.

The subject wears motion detection means (sensor device) including a position specifying signal detection device and a walking detection device.
Here, the position specifying signal detection device is a device that receives a position specifying signal by transmitting and receiving signals (single or multiple use of radio waves, sound waves, ultrasonic waves, light, and infrared rays) with the position specifying device.
The walking detection device is a device that detects the walking state of the subject by incorporating a gyro sensor, an acceleration sensor, and the like. In this embodiment, it is fixed to the subject's ankle or upper knee by a device in which the position specifying signal detection device and the walking detection device are integrated.
Data acquired by the position specifying signal detection device and the gait detection device is digitized and stored in a storage device built in the sensor device. When the measurement is completed, the information is input from the storage device to the behavior estimation unit. The storage device is generally a recording medium that is detachable from the sensor device.

When a moving object such as a shopping cart or a forklift is to be measured instead of the subject, a “movement detection device” is adopted instead of the “walking detection device”. Since this movement detection device does not require detection of the “walk” provided in the walk detection device, a mechanism (such as a sensor) and software processing for that purpose can be omitted.

If the sensor device and the predetermined computer have a communication function, data detected by the sensor device may be appropriately transmitted to the computer by communication. In that case, processing close to real time is possible.

FIG. 2 shows a state where a position specifying device (milestone) is installed at a predetermined location in a predetermined area related to the prepared map data, modeled together with a passage.
In the predetermined area, there are a path with ID = i, a path with ID = j, and a path with ID = k. From the vicinity of the center of the path i, the path j forms a T-shape with the path i. positioned. Further, the passage k is located from the vicinity of the center of the passage j so as to form a T-shape with the passage j.
The length of the passage i is L1, the length of the passage j is L2, and the length of the passage k is L3.

The position specifying device is indicated by #m and #n. That is, #m is located at one end in the passage i, and the coordinates are (x7, y7). #N is located on the side of the passage j at the intersection of the passage i and the passage j, and the coordinates are (x8, y8).
The position of #m is the starting point of the passage i and is represented by (x0, y0). The end point of the passage i is (x1, y1).
The intersection of the passage i and the passage j, that is, the starting point of the passage j is (x2, y2). The end point of the passage j is (x3, y3).
The intersection of the passage j and the passage k, that is, the starting point of the passage k is (x4, y4). The end point of the passage k is (x5, y5).

(Motion detection means)
As shown in FIG. 1, the motion detection means includes a position specifying signal detection device, a walking detection device, a movement detection device, and a storage device.
The position specifying signal detecting device includes a receiving unit that receives a signal (single or multiple use of radio waves, sound waves, ultrasonic waves, light, and infrared rays) transmitted from the position specifying device (milestone), and the position specifying device. An ID detection unit that detects solid identification data (ID).
The ID detected by the ID detection unit is stored in association with the reception time data. It may be stored in a storage device in the motion detection means, or may be sequentially transmitted to the information input means described above.

The walking detection device in the motion detection means is used by being mounted on the subject's ankle, near the knee, or somewhere on the leg, and is equipped with an acceleration sensor and a gyro sensor. The acceleration sensor detects “the subject's horizontal movement (front / back / left / right) and up / down” and the gyro sensor detects “the subject's front / back / left / right swing and leg twist”.

The position specifying data and the motion data stored in the storage device in the motion detection means are input to the behavior estimation means via the information input means. The behavior estimation means includes a motion data processing unit and an estimation unit.
The motion data processing unit includes a specific position detection unit, a “walk” detection unit, and a movement distance detection unit.

(Specific position detector)
The specific position detection unit extracts the solid identification data (ID) from the position specific signal detected by the position specific signal detection device in the motion detection means. Based on the individual identification data, the position where the subject is is obtained. An example of the time change of the position specifying signal is shown in FIG. The horizontal axis indicates the elapsed time (seconds) from the reference time when the measurement is started, and the vertical axis indicates the solid identification data.

(Walk detection unit)
When the subject walks, his / her leg is swung back and forth, and the swing before and after that is detected as a gyro signal. FIG. 4 shows the gyro signal detected by the above gyro sensor. The horizontal axis indicates the elapsed time (seconds) from the reference time when the measurement is started, and the vertical axis indicates the instantaneous value of the gyro signal.

FIG. 4 shows a state in which the subject stops once after having advanced six steps, and then stops at the seventh step after having advanced six steps again. When FIG. 4 is observed against the state, it can be seen that a signal having periodicity is output. Moreover, it can be considered that the local maximum in the time axis direction of the instantaneous value coincides with the “one step” of the subject.
Therefore, in order to detect the “walk”, preprocessing such as noise removal for extracting the maximum point is executed, and it is possible to accurately grasp how many steps the subject has moved.

FIG. 5 shows an acceleration sensor signal (vertical axis is an instantaneous value) acquired under the same conditions as in FIG. Similar to the gyro sensor signal, regularity is seen in the signal, so that it is possible to extract the “walk” using the acceleration signal.

In addition, when it is not necessary or possible to detect “walk” as in a shopping cart or a work vehicle (for example, a forklift), the movement distance is estimated from the movement data in the movement distance detection unit. For example, an integral value is obtained using a signal from an acceleration sensor installed on the moving body, and a change over time of the moving distance is calculated.

FIG. 6 shows an example in which the movement distance is calculated from the acceleration signal. The horizontal axis in FIG. 6 represents the elapsed time from the start of measurement, and the vertical axis represents the relative distance.
The movement starts at time t = 6 (s), and then moves at a substantially constant speed until time t = 11 (s). It stops between 11 and 14 (s), then moves again at t = 14 (s), and stops at 24 (s). Focusing on the vertical axis, a distance of 2.4 is moved from t = 6 to 11, and a distance of 4.8 (7.2-2.4) is moved from t = 14 to 24 (s). , 6 to 11 (s) and 14 to 24 (s) are calculated as the ratio 2.4 / 4.8 of the moving distance.
In addition, it is a measurement error that the relative distance indicates minus.

(Estimator)
As shown in FIG. 1, the estimation unit includes a walking estimation unit and a movement estimation unit. The walking estimation unit and the movement estimation unit use position specifying data and motion data input to the behavior estimation unit via the information input unit.

(Walking estimation unit)
With reference to FIG. 7, the procedure of position estimation when the subject walks along the passage will be described.
FIG. 7A shows a passage as a region where a subject is inspected, and an arrangement position of the position specifying device in the passage. The passage is L-shaped on the back, and the position specifying device (# 1) is arranged in the middle of the x-axis direction and the position specifying device (# 2) is arranged in the middle of the y-axis direction in FIG. Yes.

FIG. 7B shows the position of the subject position P at time T, using the position specifying devices (# 1, # 2) and time as axes. It is a figure for calculating | requiring the ratio of -T2 |.
│T-T1│: │T-T2│ = m: n
Can be calculated. According to FIG. 7B, it can be seen that the subject was at the position P1 at time T1 and P2 at time T2.

FIG. 7C shows the detection result of the detected “walk” using the motion detection data measured when the subject walks along this passage. A “walk” at times T1 to T2 is detected.
Assuming that the time when the second step by the subject occurs is T, from the correspondence between FIG. 7A and FIG. 7B, the position P (T) of this worker at time T is expressed as follows. be able to.
P (T) = (nP1 + mP2) / (m + n)
Although this calculation is a calculation for obtaining a so-called internal division, it may be calculated by calculating the distance between P1 and P2 along a path that is not necessarily a straight line (forms a reverse L shape in FIG. 7).

(Movement estimation part)
A procedure for estimating the position based on the movement distance will be described with reference to FIG. 8 (also FIG. 7 if necessary).
FIG. 8 (a) shows “walk” with the position specifying signal and time as axes, and obtains the position of the subject's position P at time T, which is the same as FIG. 7 (b). Is.
As shown in FIG. 7A, it can be seen that the subject was at the position of the position specifying device ID = # 1, that is, P1, at time T1. On the other hand, since the ratio of the moving distance between the time | t−t1 | and | t−t2 | is L1 / L2 according to FIG. 8B, the distance L ( T) can be expressed as:
L (T) = (P1L2 + P2L1) / (L1 + L2)
This calculation is also a calculation for obtaining a so-called internal division, and is a calculation of the distance along the passage.

(Details of positioning device)
FIG. 9 illustrates details of the principle and variations of the position specifying device. In FIG. 9A, two types of radio waves are transmitted from one position specifying device, one reaches only a narrow range, and the other reaches the other side of the passage. If the transmitter / receiver worn by the subject receives both of the two types, it has passed near the position specifying device in the passage. If only one radio wave can be received, it means that it has passed far away from the position specifying device in the passage.

10 and 11 are diagrams illustrating the above-described configuration from the viewpoint of data input / output. FIG. 11 is a variation of FIG. 10, in which a sensor attached to a subject does not collect and transmit all data, but the location specifying device directly transmits data such as subject passage obtained by the location specifying device. It is said.

(Estimation of stride)
The means for calculating using FIGS. 4, 5, and 7 is based on the premise that the subject has moved from P1 to P2 at a constant speed. Therefore, it does not reflect that the subject changed walking speed or stopped between P1 and P2.
Therefore, in order to grasp the detailed movement of the subject, a means for estimating the stride will be described with reference to FIGS.

FIG. 12 shows a state in which a subject walks through an L-shaped passage and walks from the installation position P1 of the position specifying device to the next installation position P2. In FIG. 12, “L” and “R” indicate positions where the subject's feet have landed (left foot = L, right foot = R).
FIG. 13 shows the gyro signal of the gyro sensor mounted on the subject's left foot, with the X axis being time (seconds) and the Y axis being angular velocity ω (t). The subject at this time has confirmed that he / she walked for about 12 seconds without getting faster or slower.

According to FIG. 13, a periodic signal having a clear local maximum value appears. On the other hand, the gyro signal output from the gyro sensor attached to the left foot of the subject indicates the “walking” of the subject because it detects the “speed of swinging the foot” of the pedestrian.
In consideration of the measurement result of the angular velocity, a value obtained by integrating this signal with respect to time may correspond to the magnitude of “swing” of the subject's foot, that is, “step length”.

FIG. 14 is a partially enlarged view of FIG. As a result of taking a video of the movement of the subject and comparing it with FIG. 14, the subject kicks up the foot at the time tk1 in FIG. 14 from the correspondence between the movement of the leg of the subject and the signal, and the subject moves the foot at the time tk1. It is known that it has landed.
The length Ik corresponding to the stride (k = number obtained by assigning the maximum points in the gyro signal in order of time) is calculated by integrating the gyro signal between time tk1 and time tk2, as shown in the following (Equation 1). To do.

The calculation process according to the above (Equation 1) is executed for the vicinity of all local maximum points k (1 to 11), and the value obtained by normalizing the calculated stride with the value of the first step is redefined as a new Ik. Table 1 shows. Hereinafter, this Ik is referred to as “relative stride”.

Analyzing the above (Table 1), the subject walks with approximately the same stride from the first step to the tenth step, and only the last step (the eleventh step) is about ½ (half a step). It is calculated that the person walked in step). This analysis result was consistent even when compared with the above-mentioned moving image data of the subject.

Next, a procedure for estimating where the subject was at a certain time using the relative stride (Ik) will be described.
13 and (Table 1), the subject walked from P1 to P2 in the inverted L-shaped passage shown in FIG. 12 with a relative stride I1 to I11.
If the coordinates of P1 and P2 in FIG. 12 represent the coordinates of the installation position of the position specifying device, the foot position Pn of the nth step can be estimated to be a value represented by the following (Equation 2). . (The maximum value of n in this case is “11”)

The calculation according to (Expression 2) is a calculation for obtaining a so-called internal division, but is not limited to a straight line in the walking direction of the subject, and may be an operation for performing internal division along the passage.

As shown in FIG. 15, a case where the subject stops in the middle of P1 to P2 (P3) in the inverted L-shaped passage will be described. The gyro signal of the gyro sensor attached to the subject's left foot at this time is shown in FIG. P3 has no position detection device.

According to FIG. 16, it can be read that the subject stopped at the seventh step after about 5 seconds from the start of walking, stopped for about 6 seconds, and then stopped after walking for another four steps. This is known to be consistent with the moving image data obtained by photographing the subject.
Thus, it can be grasped that the subject stopped because the value of the gyro signal becomes zero in a predetermined time zone. For this reason, it is effective to grasp the relative stride (Ik) according to (Equation 1) described above and estimate the position of the subject according to (Equation 2).

If a large number of position detection devices are installed, it can be estimated that the vehicle has stopped. For example, in the case shown in FIGS. 15 and 16 described above, if there is a position detection device other than P1 and P2 at a place where the vehicle has stopped, the movement of the subject can be obtained without using the means for estimating the stride. Can be estimated.
However, since it is not realistic to install a large number of position detection devices, means for estimating the stride is effective.

(Subject's behavior detection according to the type of subject)
In the above-described embodiment, the position specifying device is installed at a predetermined point such as a passage, and the walking detection device is fixed to the subject's ankle. However, there are various variations when it is desired to grasp the overall behavior of the subject.
For example, if you want to understand the behavior of the subject walking to point A and point B in the aisle, carrying parts to point C, and assembling those parts manually at point C. Is as follows.

First, a walking detection device (α) including a position detection signal detection device is attached to the upper arm or wrist of the subject, and separately from the walking detection device, the walking detection device (β) is also applied to the ankle or upper knee of the subject. Installing. The position specifying device is installed at points A and B in the passage as in the above-described embodiment.
When the subject is walking toward point A or point B in the passage or walking toward point C, both the walking detection devices (α) and (β) detect the angular velocity. When the subject is assembling at point C, the foot of the subject should have stopped, so the gait detector (β) does not detect the angular velocity, but the gait detector (α ) Detects the angular velocity.
In other words, in a state where the angular velocity is detected only from the walking detection device (α), it can be estimated that the subject is assembling work at the point C, and the walking velocity detection devices (α) and (β) have the angular velocity. In the detected state, it can be estimated that he is walking to carry the parts before assembly. The walking place can be estimated from the data of the position specifying device and the position specifying signal detection device.

If the position specifying device is installed at the point C in addition to the points A and B in the passage, the walking detection device worn by the subject is a walking including the position specifying signal detection device worn on the upper arm or the wrist. Only the detection device (α) is sufficient.
That is, the walking detection device (α) detects the angular velocity even when the subject is walking to carry the parts before assembly or when the parts are assembled at the point C. However, if the position specifying device and the position specifying signal detection device can grasp that the subject is at the point C, it can be estimated that the subject is assembling the parts. If the subject is not at the point C, it can be estimated that the subject is walking toward the point A, B, or C, and information from the position specifying device and the position specifying signal detecting device is obtained. It is possible to estimate which point the person is walking toward.

The present invention relates to the manufacturing industry of measuring devices such as sensors, the service industry that performs data collection and analysis using the measuring instrument, the software development industry that creates data collection and analysis programs using the measuring instrument, and the personnel in factories, etc. It can be used in the consulting industry that deals with the efficient placement of services and the service industry that collects and analyzes marketing data.

Claims

A subject behavior analysis system that calculates how a subject moves within a predetermined area,
A predetermined transceiver mounted on the subject;
A plurality of position specifying devices capable of specifying that the subject exists in the vicinity by wireless communication with the transceiver; and
A storage device that stores in advance map data related to the predetermined area and position data of the position specifying device in the map data;
A limb sensor that is attached to the subject and continuously detects the movement of the leg or arm;
Using the movement data relating to the movement of the leg or arm detected by the limb sensor, the identification data of the subject by the position identification device, the map data, and the position data, it is estimated where the subject moved with time in the area. A motion detection data processing device for calculating motion estimation data to be
Behavior analysis system with
A subject behavior analysis system that calculates how a subject moves within a predetermined area,
A predetermined transceiver mounted on the subject;
A plurality of position specifying devices capable of specifying that the subject exists in the vicinity by wireless communication with the transceiver; and
A storage device that stores in advance map data related to the predetermined area and position data of the position specifying device in the map data;
A limb sensor that is worn on the subject and continuously detects the movement of the leg or arm;
Using the movement data relating to the movement of the leg or arm detected by the limb sensor, the identification data of the subject by the position identification device, the map data and the position data, the total distance for which the subject has moved within the area is estimated. A distance calculation device for calculating distance data;
Behavior analysis system with
A subject behavior analysis system that calculates how a subject moves within a predetermined area,
A predetermined transceiver mounted on the subject;
A plurality of position specifying devices capable of specifying that the subject exists in the vicinity by wireless communication with the transceiver; and
A storage device that stores in advance map data related to the predetermined area and position data of the position specifying device in the map data;
A limb sensor that is worn on the subject and continuously detects the movement of the leg or arm;
Using the movement data relating to the movement of the leg or arm detected by the limb sensor, the identification data of the subject by the position identification device, the map data, and the position data, it is estimated where the subject moved with time in the area. A motion detection data processing device for calculating motion estimation data to be
A distance calculation device that calculates total distance data for estimating a total distance that the subject has moved in the region, using movement data relating to movement of a leg or arm detected by the limb body sensor;
Behavior analysis means for analyzing the behavior of the subject using the motion estimation data and the total distance data;
Behavior analysis system with
The limb sensor includes a gyro sensor,
The motion detection data processing device calculates a maximum value of an angular velocity detected by a gyro sensor or a value in the vicinity including a maximum value, and calculates the position data calculated from the position specifying device and the transceiver and the relative The behavior analysis system according to any one of claims 1 and 3, wherein the behavior is calculated using a stride.
A subject behavior analysis program for calculating by a computer how a subject moves within a predetermined area,
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. A data input procedure for inputting motion data from a limb sensor that continuously detects motion;
A data storage procedure in which map data relating to the predetermined area and position data of the position specifying device in the map data are stored in a storage device in advance;
Using the motion data, the map data, and the position data, a motion detection data processing procedure for calculating motion estimation data for estimating where the subject moved with time in the region;
A computer program for causing a computer to execute.
A subject behavior analysis program for calculating by a computer how a subject moves within a predetermined area,
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. A data input procedure for inputting motion data from a limb sensor that continuously detects motion;
A data storage procedure in which map data relating to the predetermined area and position data of the position specifying device in the map data are stored in a storage device in advance;
Using the movement data, the map data, and the position data, a distance calculation procedure for calculating total distance data for estimating the total distance that the subject has moved in the region;
A computer program for causing a computer to execute.
A subject behavior analysis program for calculating by a computer how a subject moves within a predetermined area,
The program includes a plurality of programs in the predetermined area, position data from a position specifying device that can specify that a subject wearing a predetermined transmitter / receiver is in the vicinity, and a leg or arm attached to the subject. A data input procedure for inputting motion data from a limb sensor that continuously detects motion;
A data storage procedure in which map data relating to the predetermined area and position data of the position specifying device in the map data are stored in a storage device in advance;
Using the motion data, the subject specific data, the map data, and the position data, a motion detection data processing procedure for calculating motion estimation data for estimating where the subject moved over time in the region;
Using the movement data, the map data, and the position data, a distance calculation procedure for calculating total distance data for estimating the total distance that the subject has moved in the region;
A computer program for causing a computer to execute.
The motion detection data processing procedure is calculated by using a maximum value of angular velocity detected by the gyro sensor or a value in the vicinity including the maximum value, and the position data calculated from the position specifying device and the transceiver and the relative data 8. The computer program according to claim 5, wherein the calculation is performed using a stride.