JP2006223651A - System, apparatus and program for feature point position acquisition - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an art for surely detecting an intended movement at an appropriate timing using a small-size pressure sensor. <P>SOLUTION: This feature point position acquisition apparatus acquires feature point positions of pressure which a user applies to a bed in every predetermined movement when the user performs an movement on the bed disposed with a large-size pressure sensor. This feature point position acquisition apparatus is provided with a measurement data storage means storing pressure values for every measurement points measured by the large-size pressure sensor when the user performs the movements and times corresponding thereto; movement data storage means storing movement identification information for every predetermined movement among the movements of the user and the times performing the movements corresponding to the movement identification information; and a feature point position acquisition means acquiring measurement point positions where change in pressure value exceeds a predetermined threshold in every predetermined movement using the measurement data stored in the measurement data storage means and the movement data stored in the movement data storage means. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technique for detecting a predetermined operation of a bed user.

  A fall accident of a bed user such as a care recipient often occurs when the user tries to get out of the bed after getting up from the bed. Therefore, it has been a problem in the past to automatically detect that the user gets up from the bed at an appropriate timing and notify the caregiver.

Various methods have been proposed to solve such problems (Patent Documents 1 and 2, etc.). As one of them, a pressure sensor is installed on a bed, and a load is applied to the pressure sensor. There is a method for determining whether or not the user is going to get out of the bed based on whether or not there is (for example, Patent Document 3).
JP 2003-265546 A JP 2004-154242 A U.S. Pat. No. 4,448,043

  However, when a pressure sensor is used, an operation other than rising, such as turning over, may be determined as rising. In addition, since a large pressure sensor that covers the entire body is expensive, it is common to use an inexpensive small pressure sensor (for example, a width of about 10 cm and a length of about 30 cm). In the present situation, it is necessary to rely on experience to arrange the appropriate part of the bed so as to reduce erroneous determination.

  In addition, the way of getting up is different for each user, and the timing for notifying caregivers etc. is also different for each user, so even if the layout is appropriate for one user, it is appropriate for another user. It may not be. As described above, there is a problem that it is difficult to appropriately arrange a small pressure sensor for each user in order to detect a desired operation.

  The present invention has been made in view of the above points, and provides a technique for appropriately arranging a small pressure sensor on a bed so as to surely detect a desired operation at an appropriate timing. For the purpose.

  The above problem is that when a user performs an operation on a bed on which a large pressure sensor having a plurality of measurement points is arranged, the position of the characteristic point of the pressure exerted on the bed by the user is acquired for each predetermined operation. A characteristic point position acquisition device, wherein the user performs a measurement data storage means for storing a pressure value for each measurement point measured by the large pressure sensor and a corresponding time when the user performs an operation, and the user Operation data storage means for storing the action identification information for each predetermined action in the operation and the time when the action corresponding to the action identification information was performed, and the measurement data stored in the measurement data storage means And feature point position acquisition means for acquiring a measurement point position at which a change in the pressure value exceeds a predetermined threshold value for each predetermined operation using the operation data stored in the operation data storage means Having It can be solved by the feature point position acquisition device according to symptoms.

  In addition, the feature point position acquisition unit is configured to extract, from the measurement data stored in the measurement data storage unit, a time extraction unit that extracts a time at which a change in pressure value exceeds the threshold value for each measurement point; and the time extraction The action identification information of the action corresponding to a certain time extracted by the means is acquired from the action data storage means, and the measurement point position at which the change in the pressure value exceeds a predetermined threshold at the time of the certain time, Position acquisition means for acquiring as a feature point position of an action corresponding to the action identification information can be included.

  The feature point position acquisition device further receives feature point position storage means for storing the feature point position in association with the action identification information for each predetermined action, and receives the input of the action identification information. The feature point position corresponding to the information is acquired from the feature point position storage means, the center position of the feature point is calculated using each acquired feature point position, and the center position of the small pressure sensor is arranged as the center position. You may have with the gravity center position calculation means output as a position to perform.

  The invention also provides a feature point position of the pressure exerted on the bed by the user for each predetermined operation when the user of the bed on which the large pressure sensor having a plurality of measurement points is arranged operates on the bed. A program for causing a computer to realize a function as a feature point position acquisition device to be acquired, the storage device of the computer having a pressure for each measurement point measured by the large pressure sensor when the user performs an operation The value and the time corresponding to the value are stored as measurement data, and the operation identification information for each predetermined operation in the user's operation and the time when the operation corresponding to the operation identification information is performed are the operation data. The program stores the change in the pressure value for each predetermined operation using the measurement data and operation data stored in the storage device. Can also be configured as a program for causing to function as a feature point position acquisition means for acquiring a measurement point position beyond the threshold determined because.

  Further, according to the present invention, when a user of a bed in which a large pressure sensor having a plurality of measurement points performs an operation on the bed, the feature point position of the pressure exerted on the bed by the user is determined for each predetermined operation. The feature point position acquisition system to acquire the feature point position, wherein the feature point position acquisition system includes a bed on which the large pressure sensor is disposed, a camera for photographing an operation of the user on the bed, and the feature point position. For each predetermined operation, and the feature point position acquisition device is a pressure at each measurement point measured by the large pressure sensor when the user performs an operation. In parallel with receiving the value and the corresponding time from the large pressure sensor as measurement data, the image data of the operation is received together with the time from the camera, and these data are stored in the storage means. Operation data, which is created using image data and time, and includes operation identification information for each predetermined operation and a time when an operation corresponding to the operation identification information is performed is stored in a storage unit, and the measurement is performed. Using the data and the operation data, a feature point position acquisition system configured to acquire a measurement point position at which the change in the pressure value exceeds a predetermined threshold for each predetermined operation. You can also.

  According to the present invention, since the feature point position can be acquired for each predetermined operation of the user, the feature point position of the operation corresponding to the appropriate timing for notifying the caregiver or the like is covered. If a small pressure sensor is disposed in the position, it is possible to reliably detect the operation at an appropriate timing. In addition, by acquiring the feature point position of each user, it is possible to arrange a small pressure sensor according to each user.

  Further, by arranging the centroid position of the small pressure sensor at the centroid position of the feature point, the small pressure sensor can be arranged so as to cover many feature point positions.

  Embodiments of the present invention will be described below with reference to the drawings. First, an outline of the present embodiment will be described along the flowchart of FIG.

  In the present embodiment, data of pressure change during the rising operation is acquired using a large pressure sensor while photographing the operation when the bed user gets up with the camera (data acquisition step: step 1). Then, the operation and time are associated with each other based on the camera image (operation and time association step: step 2). Thereafter, characteristic position information having a strong correlation with each action is acquired from the pressure distribution at the time when each action is performed (feature point position acquisition step: step 3), and position information corresponding to the desired action is obtained. And determining the arrangement position of the small pressure sensor (small pressure sensor arrangement position determining step: step 4).

(System configuration)
A system configuration which is a premise of the above-described processing in each stage will be described with reference to FIG. As shown in FIG. 2, this system includes a bed 1 on which a large pressure sensor 11 is installed, a camera 3, and a data collection / processing device 5. The large pressure sensor 11 has a size of an adult single bed (a size that does not protrude when a person is sleeping or getting up), and is installed on the bed 1 (under the bed mat). . The large pressure sensor 11 has pressure measurement points that can measure the magnitude of pressure arranged in a mesh.

  The camera 3 is installed at a position where the motion of the user of the bed 1 can be photographed. The data collection / processing device 5 is connected to the large pressure sensor 11 and the camera 3 via a network 7 such as a LAN, receives pressure data from the large pressure sensor 11, and receives image data from the camera 3. .

  FIG. 3 shows a detailed configuration of the data collection / processing device 5.

  As shown in FIG. 3, the data collection / processing device 5 uses a feature point position acquisition unit 51 for acquiring position information of a characteristic pressure for each operation, and an extracted feature point position, Small pressure sensor optimum arrangement calculation unit 53 for calculating the optimum arrangement of the small pressure sensor, a storage device 55, a data input / output unit 57 for accessing each file in the storage device 55 to read / write data, and a camera 3 and a data communication unit 59 that communicates with devices connected via a network such as the large pressure sensor 11.

  The feature point position acquisition unit 51 includes a measurement data reading unit 511 that reads measurement data, a pressure change calculation unit 513 that calculates a change in pressure value, a pressure change comparison unit 515 that compares the pressure change with a threshold value, and an operation name search unit. 517 and a feature point information storage unit 519. The small pressure sensor optimal arrangement calculation unit 53 provides an interface for allowing the user to input an operation name, an operation name input unit 531 for receiving the input operation name, a feature point reading unit 533, and an optimal arrangement. And an arrangement calculation unit 535 that performs calculation for.

  As shown in FIG. 4, the storage device 55 stores a measurement point data file 551, an operation data file 553, and a feature point file 555 for each operation. Further, an image data file 557 for storing image data sent from the camera 3 is provided.

  In the measurement point data file 551, the measurement point position, measurement time, and pressure value in the large pressure sensor 11 are stored in association with each other. The operation data file 553 stores an operation name and time in association with each other. In the feature point file 555 for each operation, the operation name and the feature point coordinates are stored in association with each other.

  The data collection / processing device 5 can be realized by mounting a program according to the present invention for executing the processing described below on a computer having an interface device with the camera 3 and the large pressure sensor 11. It is. The program can be stored in a recording medium such as a CD-ROM.

  In the following, processing contents at each stage in the above system will be described.

(Data acquisition stage)
First, using the system shown in FIG. 2, the camera 3 is used to photograph various operations when the user of the bed 1 gets up. Data generation time is given to the image data obtained by photographing, and the image data to which the time data is given is transmitted from the camera 3 to the data collection / processing device 5. The data collection / processing device 5 stores the image data to which the time data is added in the storage device 55. The time data may be attached on the data collection / processing device 5 side instead of on the camera 3 side.

  FIG. 5 shows an example of a time-series change (time chart) of the user's rising operation. The elapsed time is shown on the lower side in FIG. In the example of FIG. 5, the comforter is first moved after the user wants to get up. Then, lying down, putting your hands and elbows on the futon, raise the upper body with the side rails. The data collection / processing device 5 records an image related to such an operation.

  In parallel with the above photographing by the camera 3, the pressure change is measured by the large pressure sensor 11, the time is given to the measurement data, and the pressure data and the corresponding time data are transmitted to the data collection / processing device 5. The The data collection / processing device 5 stores the pressure data and corresponding time data in the storage device 55 as the measurement point data file 551. The time data may be attached on the data collection / processing device 5 side instead of on the large pressure sensor 11 side.

  An example of data in the measurement point data file 551 is shown in FIG. As shown in FIG. 6, the measurement time and the pressure at that time are recorded for each measurement point coordinate in the large pressure sensor 11. That is, in the measurement point coordinates (x1, y1), the pressure value is 1 at the time (2004/11/26, 07: 00: 00: 0.00), and the time (2004/11/26, 07:00: (01.000), the pressure value is 2, and so on. FIG. 6 shows an example in which the pressure measurement sampling time is 1 second.

(Association stage of action and time)
After collecting the image data and the pressure measurement data, the operation data file 553 is created by associating each operation of the bed user with the time when the operation was performed using the image data. The association between each operation of the bed user and the time when the operation was performed may be performed manually while the operator is viewing the image, or may be automatically performed using an existing image processing method.

  In the case of manual operation, the operator looks at the image and inputs the time when a predetermined operation (for example, “landing”) is performed and a name for identifying the predetermined operation to the data collection / processing device 5 To do. The data collection / processing device 5 records the input time and name in the operation data file 553. FIG. 7 shows an example of the contents of the operation data file 553. As shown in FIG. 7, the operation name and the time when the operation of the operation name was performed are recorded.

(Feature point position acquisition stage)
After the operation data file 553 is created, the data collection / processing device 5 extracts position information indicating a characteristic pressure change in each operation. With reference to the flowchart of FIG. 8, the process which the data collection and processing apparatus 5 performs in the feature point position acquisition stage is demonstrated.

  First, the data collection / processing device 5 reads out the pressure values at all measurement points and the corresponding times from the measurement data file 551 in the storage device 55 and stores them in the memory (step 11). Next, using the data held in the memory, a change in pressure value at a certain measurement point is calculated over the entire time (step 12). The change in the pressure value can be obtained, for example, as the amount of change in the pressure value per unit time. The calculation result of step 12 is stored in the memory in association with the time.

  Subsequently, the data collection / processing device 5 compares the absolute value of the pressure change value with a predetermined value (threshold value α) at each time, so that the time when the pressure change exceeds the threshold value α. Is extracted (step 13). Then, the operation data file 553 is searched using the time when the pressure change exceeds the threshold value α, and the operation name corresponding to the time is extracted (step 14). Here, it is assumed that the operation name is extracted as a match if the time when the pressure change exceeds the threshold value α is within a predetermined error range with respect to the time in the operation data file 553.

  Then, the data collection / processing device 5 records the measurement point coordinates targeted in step 12 at the location of the extracted operation name in the feature point file 555 for each operation (step 15).

  Thereafter, it is checked whether or not the processing of step 12 to step 15 has been completed for all measurement points (step 16). If not, the processing of step 12 to step 16 is executed for the next measurement point. If completed, the process is terminated.

  As a result, for example, a feature point measurement file 555 for each operation as shown in FIG. 9 is obtained. In the example of FIG. 9, for example, regarding the operation of “wake up”, the pressure change is a threshold value at each measurement point of (x1, y1), (x2, y3), (x5, y1), (x7, y4). It shows that α was exceeded. That is, the points at these positions are characteristic points in the operation of “wake up”.

  In the above processing, the feature point position is extracted based on the magnitude of the pressure change. However, a point having a pressure value larger than a certain threshold value at a specific operation time may be used as the feature point.

(Small pressure sensor placement position determination stage)
By using the feature point position for each action obtained as described above, the feature point position in the desired action in the action name is extracted, and a small pressure sensor is arranged so as to cover the feature point position as much as possible. Thus, the desired operation can be efficiently detected using a small pressure sensor. For example, in the example of FIG. 9, when an alarm notification is required at the time of “wake up”, (x1, y1), (x2, y3), (x5, y1), (x7, y4) One or a plurality of small pressure sensors are arranged so as to cover these points as much as possible.

  As described above, the small pressure sensor may be manually arranged by visual observation or the like, or the arrangement position of the small pressure sensor may be automatically obtained as described below.

  An example of processing in which the data collection / processing device 5 automatically obtains the location of the small pressure sensor will be described with reference to the flowchart of FIG.

  First, an operator who desires an alarm notification for a certain operation inputs the name of the operation from the input screen displayed on the data collection / processing device 5 (step 21). The data collection / processing device 5 reads the feature point position corresponding to the input action name from the feature point file 555 for each action (step 22).

  Then, the data collection / processing device 5 obtains and outputs the barycentric positions of all the feature point positions using the coordinates of the read feature point positions (step 23). Thereafter, the operator aligns the position of the center of gravity with the position of the center of gravity of the small pressure sensor, and places the small pressure sensor on the bed (step 24).

  In the above example, the barycentric positions of all feature point positions are obtained. However, when using a plurality of small pressure sensors, the area including the feature point positions is divided into a plurality of areas, and the feature point positions are The center of gravity may be obtained.

  Further, the data collection / processing device 5 displays the feature points and the shape of the small pressure sensor on the screen so that the operator can cover the feature points as much as possible using the one or more positions of the center of gravity as a guide. You may arrange | position a small pressure sensor visually on a screen. The data collection / processing device 5 outputs the result, and the operator actually places the small pressure sensor on the bed based on the output. An example of the arrangement of the band-shaped small pressure sensor is shown in FIG.

  After the small pressure sensor is arranged to detect the desired operation, the small pressure sensor detects the pressure change and notifies the nurse call system etc. via the network in the actual operation detection stage. Will do.

  As described above, by using the method of the present embodiment, it is possible to acquire a feature point position corresponding to a desired operation. Therefore, by disposing the small pressure sensor so as to cover the feature point position, it is possible to reliably detect a desired operation such as a rising motion of a person. In other words, it is possible to reliably detect a desired operation such as a person's rising operation using only a small pressure sensor that can cover only a part of the body. Further, if the desired operation is an operation corresponding to an appropriate timing for notifying a caregiver or the like, the operation can be reliably detected at an appropriate timing. Further, by arranging the centroid position of the small pressure sensor at the centroid position of the feature point, the small pressure sensor can be arranged so as to cover many feature point positions.

  The present invention is not limited to the above-described embodiment, and various modifications and applications can be made within the scope of the claims.

It is a flowchart for demonstrating the outline | summary of embodiment of this invention. It is a system configuration figure in an embodiment of the invention. 3 is a functional configuration diagram of a data collection / processing device 5. FIG. 5 is a diagram showing files stored in a storage device 55. FIG. It is a figure which shows an example of the time-sequential change of a user's rising operation. It is a figure which shows an example of the content of the measurement point data file. It is a figure which shows an example of the content of the operation | movement data file 553. It is a flowchart which shows the process of the data collection and processing apparatus 5 in the feature point position acquisition stage. It is a figure which shows an example of the content of the feature point measurement file 555 for every operation | movement. It is a flowchart which shows an example of the process which calculates | requires the arrangement | positioning location of a small pressure sensor automatically. It is a figure which shows an example of arrangement | positioning of a strip | belt-shaped small pressure sensor.

Explanation of symbols

1 Bed 3 Camera 5 Data Collection / Processing Device 7 Network 11 Large Pressure Sensor 51 Feature Point Position Acquisition Unit 53 Small Pressure Sensor Optimal Arrangement Calculation Unit 55 Storage Device 57 Data Input / Output Unit 59 Data Communication Unit 511 Measurement Data Reading Unit 513 Pressure Change Calculation unit 515 Pressure change comparison unit 517 Operation name search unit 519 Feature point information storage unit 531 Operation name input unit 533 Feature point reading unit 535 Arrangement calculation unit 551 Measurement point data file 553 Operation data file 555 Feature point file 557 for each operation data file

Claims (8)

A feature point position acquisition device that acquires a feature point position of pressure exerted on a bed by a user when the user performs an action on a bed on which a large pressure sensor having a plurality of measurement points is arranged. Because
Measurement data storage means for storing a pressure value for each measurement point measured by the large pressure sensor when the user performs an operation and a corresponding time;
Action data storage means for storing action identification information for each of the predetermined actions in the user's actions and a time when an action corresponding to the action identification information is performed;
Using the measurement data stored in the measurement data storage unit and the operation data stored in the operation data storage unit, measurement in which the change in the pressure value exceeds a predetermined threshold value for each predetermined operation A feature point position acquisition device comprising: feature point position acquisition means for acquiring a point position. The feature point position acquisition means includes:
Time extraction means for extracting, from the measurement data stored in the measurement data storage means, the time at which the change in pressure value exceeds the threshold for each measurement point;
Measurement point position at which the change of the pressure value exceeds a predetermined threshold at the time of the certain time obtained from the operation data storage means, the action identification information of the action corresponding to the certain time extracted by the time extracting means The feature point position acquisition apparatus according to claim 1, further comprising: a position acquisition unit that acquires a feature point position of an action corresponding to the action identification information. Feature point position storage means for storing the feature point position in association with the action identification information for each predetermined action;
In response to the input of certain action identification information, the feature point position corresponding to the action identification information is acquired from the feature point position storage means, and the center of gravity position of the feature point is calculated using each acquired feature point position, The feature point position acquisition device according to claim 1, further comprising: a center-of-gravity position calculation unit that outputs the center-of-gravity position as a position where the position of the center of gravity of the small pressure sensor is arranged.   The feature point position acquisition device further includes measurement data receiving means for receiving measurement data measured by the large pressure sensor, and image data receiving means for receiving image data obtained by photographing the user's action with a camera. The feature point position acquisition apparatus according to any one of claims 1 to 3. When a user of a bed on which a large pressure sensor having a plurality of measurement points is placed operates on the bed, the feature point position at which the user obtains the feature point position of the pressure exerted on the bed for each predetermined operation A program for causing a computer to realize a function as an acquisition device,
The storage device of the computer stores the pressure value for each measurement point measured by the large pressure sensor and the corresponding time as measurement data when the user performs an operation. The operation identification information for each of the predetermined operations and the time when the operation corresponding to the operation identification information was performed are stored as operation data,
The program causes the computer to
Using the measurement data and operation data stored in the storage device, function as feature point position acquisition means for acquiring a measurement point position at which the change in the pressure value exceeds a predetermined threshold for each predetermined operation A program characterized by letting The feature point position acquisition means includes:
Time extraction means for extracting, from the measurement data, the time at which the change in pressure value exceeds the threshold for each measurement point;
The action identification information of the action corresponding to a certain time extracted by the time extracting means is acquired from the action data, and the measurement point position at which the change in the pressure value exceeds a predetermined threshold at the time of the certain time, The position acquisition means which acquires as a feature point position of the operation | movement corresponding to the said operation | movement identification information. When a user of a bed on which a large pressure sensor having a plurality of measurement points is placed operates on the bed, the feature point position at which the user obtains the feature point position of the pressure exerted on the bed for each predetermined operation An acquisition system,
The feature point position acquisition system includes a bed on which the large pressure sensor is disposed, a camera that captures an action on the user's bed, and a feature point for acquiring the feature point position for each predetermined operation. A position acquisition device,
The feature point position acquisition device includes:
In parallel with receiving from the large pressure sensor as measurement data the pressure value for each measurement point measured by the large pressure sensor and the corresponding time when the user performs the operation, The image data of the operation is received together with the time, and these data are stored in the storage means.
The operation data including the operation identification information for each predetermined operation, created using the image data and the time, and the operation time corresponding to the operation identification information is stored in the storage unit,
Using the measurement data and the operation data, a measurement point position acquisition system that acquires a measurement point position at which the change in the pressure value exceeds a predetermined threshold value for each predetermined operation. The feature point position acquisition device includes:
The feature point position is stored in storage means in association with the action identification information for each predetermined action,
In response to the input of certain motion identification information, the feature point position corresponding to the motion identification information is acquired from the storage means, and the center of gravity of the feature point is calculated using each of the acquired feature point positions. Is output as a position where the position of the center of gravity of the small pressure sensor is arranged.

Images