JP2011108137A - System for position detection and behavior recognition of detection object within building - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system capable of performing position detection and behavior recognition of a detection target (person, object) moving inside a building, and performing information with respect to the detection object, as needed. <P>SOLUTION: Within a floor 1 of the building having a grid ceiling 5 in which ceiling boards are attached in a rectangular pattern, a detection means that includes a wide area sensor 15 having the whole floor 1 as a detection range, a gate sensor 16 arranged at a position for passing between partitioned zones, and a spot sensor 17, arranged in each zone so that the inside of the zone can be detected in detail is provided on each ceiling board of the grid ceiling 5. The ceiling boards provided with the detection means are arranged at predetermined positions, according to the layout of the grid ceiling 5. Position detection or behavior recognition of the detection target person or detection target object can be performed with stepwise accuracy by the detection means 15, 16, 17 arranged at each position. Furthermore, necessary information can be also given to the detection object person. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a position detection and action recognition system for detection objects (people, objects) moving in a building, and in particular, in a building where a grid ceiling is employed, various detection means are set on the ceiling plate of the grid ceiling. The present invention relates to a position detection and behavior recognition system for a detection target in a building that can detect a position of the detection target and recognize a behavior and can perform information assist for the detection target as needed.

  In recent years, various position detection systems called location systems for specifying the position of a moving detection target (person, object) have been developed and proposed. When the movement range of the detection target is outdoors, many location systems using GPS have been put into practical use. On the other hand, when it is limited to buildings, there are various factors such as the detection target range, the number of detection targets, the sensitivity of the detection equipment, and so on. (Patent Documents 1 to 3).

  In Patent Document 1, as a human position specifying means for specifying a human position for moving an object, an object with an RFID tag, a reader / writer for reading the position of the object, and an object position based on the read data are disclosed. An object position calculation device and a plurality of imaging devices are disclosed. For example, an imaging device that directly recognizes the state of a person is installed at a position higher than the average height of the person near the ceiling in the room. The image data captured by the imaging device is transmitted to the object position calculation device, and the object position calculation device, for example, determines a portion estimated as the same object (human) from a plurality of image data in which the same region is imaged. It has a function of estimating the physical position of a human being, using the principle of triangulation, by specifying by image clustering processing or the like and performing stereo processing between the same objects.

  Patent Document 2 discloses a position management system in which an ultrasonic tag and an RFID tag are selectively used depending on the accuracy. The purpose is to improve detection accuracy by complementing each other's detection ability by using the plurality of detection tags. At this time, in the position management system that detects information on these tags, detection devices for different types of tags are alternately arranged in a grid at intervals of 50 cm on the ceiling of the detection space. As described above, by arranging the grids at equal intervals on the ceiling surface, it is possible to easily perform storage / collation at the arrangement positions of the two types of detection devices.

  Patent Document 3 discloses a detection device that focuses on the recognition of behavior to be detected. For example, in a limited space such as a toilet, bathroom, bedroom, etc., by identifying the posture of the target person in the place, a critical state where the elderly etc. become unconscious while falling An indoor monitoring system that can be discovered early is disclosed. In Patent Document 3, for example, a bright spot image acquired every predetermined time by an FG visual sensor installed on a ceiling surface of a private toilet is compared with a reference plane bright spot image stored in advance, and a bright spot is detected. Extract the moved area. Then, the feature amount of the moving bright spot region is calculated, and it is determined whether the posture of the person in the private room toilet is “standing”, “sitting”, or “falling”.

JP 2008-275324 A JP 2007-17414 A JP 2001-28086 A

  In the invention of Patent Document 1, a sensor for detecting a position in a space is installed for each purpose. For this reason, there is no correlation between the individual sensors, and as a result, a system that combines the individual sensors to detect the situation and make a comprehensive judgment is not available. In addition, the size and connection method of each sensor vary depending on the purpose, and are not standardized. For this reason, it is not possible to change the detection range freely according to the modification of the target floor.

  Further, in the invention of Patent Document 2, the detection device is characterized in that it is provided in a grid shape and alternately at a predetermined interval on the ceiling of the detection space, but is directly attached and fixed to the existing ceiling member. In order to respond to changes in the position detection range and detection items, when updating the function of the sensor or changing the mounting position, it is necessary to perform interior work and electrical work, which is troublesome.

  In the invention of Patent Document 3, a monitoring range is set in a limited target space such as a toilet, and a three-dimensional shape such as a person moves within the target space. Since it is necessary to prepare a reference surface bright spot image in this target space in advance, it is difficult to take a wide target space from the ability of image processing, and enormous calculation processing is also required for collation with the bright spot moving area. Necessary.

  Therefore, the object of the present invention is to solve the problems of the conventional technology described above, and by using a grid-like grid ceiling, each detection range is specified from the detection of the position of the detection target according to the application. Detecting the position of the target to be detected in a step-by-step manner, including recognizing individual actions, and rearranging the ceiling panels of the grid ceiling in response to layout changes in the building, etc. The purpose is to provide a system that enables action recognition.

  In order to achieve the above-described object, the present invention provides a detection object person or a detection object with a specific ID on a floor to be detected in a building where a grid ceiling is constructed. In the position detection and action recognition system of the detection target, a plurality of types of detection means are provided on the ceiling plate of the grid ceiling, the ceiling plate is arranged at a predetermined position of the grid ceiling, and each of the detections arranged at each position According to the present invention, the position detection and the action recognition of the detection target person or the detection target object are performed with stepwise accuracy.

  The detection means includes a wide area sensor provided on the ceiling plate so that the entire floor is set as a detection range, and a gate sensor provided on a ceiling plate disposed at a location passing between zones defining the floor. A spot sensor provided on the ceiling plate in the zone so that detailed detection in each zone is possible, and from the information for each detection range of each sensor, the position detection of the detection target person or the detection target, the behavior recognition It is preferable to carry out with stepped accuracy.

  The wide area sensor detects the position of a specific ID existing in the floor, the gate sensor performs a specific ID movement and position detection for each zone, and the spot sensor specifies a specific ID in the zone. It is preferable to perform detailed position detection and behavior recognition of the ID.

  It is preferable to provide support information for the action of the specific ID according to the position information and action recognition information of the specific ID obtained by the detection means.

  It is preferable that the support information is provided to the specific ID or information is provided to another specific ID for the action support of the specific ID according to the action form of the specific ID. .

  In accordance with the layout or zone change in the floor, etc., each detection means provided on the ceiling plate of the grid ceiling is expanded and moved for each ceiling plate according to the new system layout, after the change It is preferable to enable detection corresponding to the layout or zone.

  According to the present invention, for each ceiling plate of the grid ceiling of the floor of the detection target range, by installing a detection device according to the application, from the detection of the position of the detection target to the action recognition of the identified individual step by step Can grasp. In addition, by replacing and changing the position of the ceiling plate of the grid ceiling where the detection equipment is installed according to the change in the facility layout in the building, etc., it is possible to easily change the new system and adapt to the new layout, There is an effect that it is possible to detect the position of the detection target in the building that is integrated and rich in variability and to recognize the behavior.

The plane layout figure which showed the example of arrangement | positioning of various detection means and the zone layout of a floor permeate | transmitting the grid ceiling of the floor to which the position detection of the detection target in the building of this invention and the action recognition system are applied. The schematic explanatory drawing which showed the example of bed arrangement | positioning in the room | chamber shown in FIG. 1 which showed typically the grid ceiling, arrangement | positioning of various detection means, and the relationship of the detection subject in each aspect in a zone with the perspective view. The partial side view which expanded and showed the example of a partial cross section of a grid ceiling, arrangement | positioning of various detection means provided in the grid ceiling, and a wiring example. Explanatory drawing which showed the example of a detection of the gate sensor provided in the grid ceiling, and a detection target person. Explanatory drawing which showed typically the spot sensor provided in the grid ceiling, and its detection area example. The block diagram which showed one Example of the system configuration | structure of the position detection of the detection target in the building of this invention, and an action recognition system. The flowchart which showed the example of the procedure of position detection of a detection subject person in a floor, and action recognition.

  Hereinafter, the following examples will be described with reference to the accompanying drawings as forms for implementing the position detection and behavior recognition system of a detection target in a building of the present invention.

  FIG. 1 is a plan view showing an embodiment in which the position detection and behavior recognition system of a detection target in a building of the present invention is applied. This figure is a plan view of a specific floor (hereinafter simply referred to as a floor) of a hospital as an application example. The figure shows an example of a floor in which three rooms 7A, each independently recognized as a zone, a central corridor 7B, an examination room 7C, a waiting room 7D, a treatment room 7E, and a nurse station 7F are arranged. ing. Moreover, each object person detection means 10 installed in the predetermined position of the grid ceiling 5 (broken line display) of this floor is typically shown. In the figure, a wide area sensor 15 as a target person detection means 10 in the floor, a gate sensor that makes it possible to grasp a detection target for each zone, and detailed position detection or behavior recognition of the detection target person in the target zone. A spot sensor that can be used is arranged.

Hereinafter, the arrangement and configuration of each target person detection unit 10 illustrated in FIG. 1 will be described.
[Wide area sensor]
In the present embodiment, the wide area sensor 15 uses position information detection means configured by a wireless LAN. This position information detection means is a transmission means such as an ID tag attached to a detection target (person, thing), an access point 15a provided on the grid ceiling 5 and capable of covering a predetermined area, and transmitted from the access point 15a. And a position information server 21 for processing the detected position detection information. The ID tag is a tag (active tag) having a transmission function attached to a device for identifying a position and a target person (active tag), and transmits radio waves to a predetermined access point 15a. In the case of a hospital facility such as this embodiment, ID tags are provided separately for medical personnel such as doctors and nurses and patients (inpatients and outpatients), and each person possesses or is attached to a device. Is done. It is preferable that medical staffs have a PDA (Personal Digital Assistant: portable information terminal) as information assisting means to be described later. On the other hand, it is preferable for patients to have or wear compact IC cards, button tags, wristband tags, and the like. In addition, a transmitter such as a tag for attaching a specific ID is also attached to a medical device, a measuring device, a medical device cart, or the like that requires position specification or confirmation of usage.

  The access point 15a functions as an antenna that receives radio waves from an ID tag or a PDA wireless LAN card and transmits it to the location information server 21, and is installed on the grid ceiling 5 so as to cover an area where signals can be received. The The location information server 21 is a platform for location management and information management such as calculation of the location of each ID tag and PDA, accumulation of location information data, location information display function, information assist processing to be described later, and the like, provided in the nurse station 7F. The data processing unit 20 is installed as one of control devices.

[Gate sensor]
As shown in FIGS. 1, 2, and 4, the gate sensor 16 is a detection target (person, equipment) (hereinafter simply referred to as a detection target person) such as a hospital room 7 </ b> A, a central corridor 7 </ b> B, and between adjacent rooms. .) Is provided with an ultrasonic reader 16 (hereinafter referred to as a reader 16) on each of the grid ceilings 5 positioned before and after the passage. As shown in FIG. 4, the reader 16 is provided on the grid ceiling 5 at a position so as to straddle a passage portion (gate) between rooms. Then, the ultrasonic wave transmitted from the ID tag 8 possessed by the detection target person P passing through the gate is passed by the detection target person from the reception order and the reception wave number of the ultrasonic waves received by the two readers 16 as they pass. Identify the direction. Since the reader 16 cannot obtain the relationship between the ultrasonic wave and the ID tag 8, the ultrasonic wave transmitted from the ID tag 8 is received by the transmission command by the weak radio wave to the ID tag 8 existing in the target zone. To. At this time, it is preferable to set basic information such as a target ID tag 8 for issuing a command, a transmission interval, and a transmission command order in advance for the transmission command. In addition, the transmission instruction | command to the several ID tag 8 is performed based on the predetermined transmission order. Thereby, it is determined from which ID tag 8 the ultrasonic wave received by the reader 16 is transmitted. In addition to the ultrasonic wave, weak radio waves, infrared rays, wireless LAN, and the like can be adopted as the transmitted wave.

[Spot sensor]
As shown in FIG. 1, FIG. 2, and FIG. Installed. In the present embodiment, a dome-like case 13 as shown in FIG. 5D is attached as the spot sensor 17, and a human sensor 17A and a three-dimensional distance detection means 17B constituting the spot sensor 17 are provided therein. Is housed. As the human sensor 17A, an infrared sensor capable of recognizing moving objects in all sensing areas in the room 1 (FIGS. 5A and 5B: areas 2A, 3A, and 4A) is used. Since the infrared sensor detects and reacts to the weak static electricity on the human body, there is a person in the room 1 even when the person (subject) is sleeping on the bed 2 and hardly moves. Can be detected. Further, as the three-dimensional distance detection means 17B, a CCD camera having a multi-focus function capable of multi-point focusing in the imaging region is used.

  As the detection means 17B, as long as the distance can be detected with an accuracy error of about several centimeters, various three-dimensional distance sensors composed of passive measurement techniques such as a multi-point focusing CCD camera and a stereo camera. Can be used. In addition, as an active measurement method, a detection beam (ultrasound, infrared, etc.) is emitted from the ceiling toward the bed and floor to detect the presence and position (height) of the person to be detected and block the beam. It is also possible to compare the height and the reference height surface for comparison and determine the situation of the person to be detected.

  FIG. 2 shows, from the top of the grid ceiling 5, the grid ceiling for explaining the arrangement state of the hospital room 7 </ b> A and the central hallway 7 </ b> B on the floor 1 to which the position detection and behavior recognition system of the building of the present invention is applied. 5 is a schematic explanatory view showing 5 through. FIG. In the same figure, due to the situation of the person to be detected, the situation of a plurality of persons to be detected (the situation on the bed 2, the situation at the bedside, the situation immediately after entering the private room) is shown at the same time. . For example, the bed 2 shows a state where the detection target person P1 is sleeping. Moreover, the detection target person P2 falls to the floor of the bedside 3 and the state which cannot get up is shown. Furthermore, a state where a detection target person (or a visitor) P3 stands near the entrance (gate) of the hospital room 7A is shown.

  Furthermore, in the same figure, the human sensor 11 and the three-dimensional distance detection means 12 (FIG. 5 (d)) as the subject situation detection means 10 provided in the hospital room 7A are substantially at the center of the ceiling surface 5. The installed state is shown. The detection operation range (sensing region) of these devices is schematically illustrated as a substantially conical image.

  Here, a configuration of the grid ceiling 5 and a configuration example of various sensors set on the grid ceiling 5 will be described with reference to FIG. FIG. 3 shows a partial cross-sectional view of the grid ceiling 5 of the present embodiment. The grid ceiling 5 has a structure in which a ceiling plate end portion 5e (edge) is supported on a support flange 50a portion of a T-shaped grid bar (crossbar) 50 having a known ready-made size. The grid size (grid size) of the grid ceiling 5 is a 600 mm × 600 mm module that has been widely used in recent years. In FIG. 3, the access point 15 a of the wide area sensor 15 on the ceiling plate supported by the grid bar 50, the hub 52 connected from the access point 15 a by the LAN cable 51, and the LAN cable trunk line between the grid bars 50. 53 is shown. As shown in FIG. 2, the wide area sensor 15 and the gate sensor 16 described above can be freely added or relocated within the constructed LAN by cascade connection via the LAN cable 51 and the hub 52 serving as a connection point. Is possible. The grid ceiling module includes, for example, a 640 mm × 640 mm module, and the present invention is not limited to the above dimensions as long as the standard is unified.

  Since the grid-like grid bars that support the ceiling plate of the grid ceiling 5 have the same dimensions, even when a specific ceiling plate to which each sensor is attached is replaced with a new grid position, By rewiring the connection point, it can be used immediately as each detection means.

  FIG. 4 is a schematic explanatory view schematically showing the operation state of the gate sensor 16 described above. As shown in the figure, the gate sensor 16 is installed on the ceiling plate 5b of the grid ceiling 5 along the gate passing direction. These gate sensors 16 receive the transmission ultrasonic waves from the ID tag 8 worn by the detection target person P, detect the passing direction of the detection target person P from the reception time lag, and increase or decrease the number of persons in each zone. The position information server 21 (FIG. 1) has a function of counting. As described above, since the ID tag information does not correspond to the generated ultrasonic wave, the position information server 21 checks each ultrasonic wave and the specific ID.

  Next, the operation of the spot sensor 17 described above will be described with reference to FIGS. FIGS. 5A and 5B are plan layout diagrams showing an example of the arrangement of the beds 2 in the room 1 shown in FIG. FIG. 3C is a front view schematically showing a detection operation range (sensing region) of the spot sensor 17 installed on the grid ceiling 5 in the room 1. As shown in FIGS. 5 (a) and 5 (b), the bed 2 is arranged near the wall at the center or corner of the hospital room 7A. Around the bed 2, a sensing area having a U-shape with a width of about 1 m recognized as the bedside area 3A is defined.

  As described above, the inside of the hospital room 7A is the bed area 2A having the plane size of the bed 2 (for example, width 0.9 m × length 2.2 m), the bedside area 3A so as to surround the periphery, and the entire floor 4 outside the bed 4A. Are divided into three sensing areas of the floor area 4A. Then, in each sensing area, as will be described later, the position (height) data of the subject obtained by the spot sensor 17 is compared with the reference height in the area, and the ratio occupied is calculated. In addition, it is possible to determine the behavior of the target person (for example, a situation that takes into consideration the wake-up state, the fall / fall, or the possibility of the presence of a visitor in the hospital room 7A).

Here, based on the action recognition information of the detection target person obtained by the spot sensor 17, information assist that can be performed in the next stage will be described.
When the spot sensor 17 obtains the action recognition of the person to be detected or detailed position information, it is possible to take a support action for improving the action or situation of the person to be detected. This is called “information assist” in this specification. An example of this information assist will be briefly described below according to the position or behavior of the detected or recognized target person.
If the person to be detected is a medical worker such as a doctor or nurse, the position detection information of the target person (operating room, treatment room, hospital room) and special equipment used for the medical action at the same time are used at the same time. Information (determined from the position information of the device) is obtained, the position information server verifies the combination with a control unit (not shown), and as an information assist, the proper usage method of the device used, if necessary Information to the doctor or the like is provided from the control unit to the PDA or PHS possessed by the doctor or the like as support information in an appropriate data format of image information, text information, and voice information. On the other hand, in the case of information from the ID tag worn by the patient, for example, movement outside the ward due to dredging behavior is confirmed, or in the case of a patient on the bed, the wake-up state and the occurrence of a fall / fall accident can be remotely I can grasp. In that case, an emergency call is made to the PDA or PHS owned by the nurse in charge, the nearest nurse or helper. Thereby, a nurse etc. can arrive at an accident position quickly and can perform required treatment.

  FIG. 6 is a schematic configuration of the present invention, each of the above-described detection means, and a data processing means capable of grasping the position information processing and action recognition information of the detection target person and performing processing for providing necessary assist information. It is the block diagram which showed the relationship with an information assistance means. With reference to the figure and FIG. 7, the procedure of the position detection of the detection target in the building of this invention, the position detection in an action recognition system, and an action recognition method is demonstrated.

As described above, the ceiling of the corresponding floor of the building is configured by the grid ceiling 5 as shown in FIG. 1, for example, and the wide area sensor 15 and the gate sensor are arranged at predetermined positions on the grid ceiling 5 according to the layout of the floor. 16 and a spot sensor 17 are arranged. Position detection and action recognition of the person to be detected on this floor (for example, the floor of the layout of FIG. 1) will be described with reference to FIGS.
Initially, specific IDs such as patients who are admitted to the ward on this floor, ward nurses, doctors in charge, helpers, and the like are initially registered. A visitor or the like receives a visitor ID tag at the hospital reception, is detected by the wide area sensor 15 upon confirmation of arrival at the floor, is authenticated as a new ID in the location information server, and is confirmed with the location confirmation on this floor. Is additionally registered in the location information server as an ID for this purpose (step 100, hereinafter referred to as S100).

Usually, the position of the person (thing) to be detected in the floor is repeatedly confirmed at a predetermined timing by the wide area sensor 15 (S110). Further, it is possible to grasp the number of people for each zone and detect the position of the person to be detected as necessary, and to confirm the proper personnel arrangement and the work flow of the person to be detected (S120). In that case, each zone (room) is divided, and at the gate through which the detection target person passes, the movement direction and the number of movements of the detection target person are confirmed by the gate sensor 16, and the number of persons in each zone at a predetermined timing is confirmed ID position detection is performed (S130).
Further, it is possible to confirm whether or not action recognition with a specific ID is performed (S140). By the selection, the detailed position detection or action recognition of the specific ID is performed by the spot sensor 17 (S150, 160). As a result of this behavior recognition, it is determined whether support information (information assist) needs to be provided for the specific ID (S170). In this determination, a predetermined determination criterion is set in advance, and the determination is made immediately in the relationship between the recognized action and the specific ID, and information assist is performed by an appropriate timing, procedure, and means (S180). Examples of the contents and means of the provided information are as described above.

  In the above explanation, the method of detecting the position of the person to be detected at each facility on the hospital ward floor and the method of behavior recognition were explained as an example, but in the nursing facility, general office office, the floor of an experimental facility such as a laboratory, etc. It goes without saying that can also be applied. That is, the present invention is not limited to the above-described embodiments, and various modifications within the scope of the claims are possible, and technical means appropriately modified within the scope of the claims are included. Embodiments obtained by combining are also included in the technical scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Floor 2 Bed 3 Bedside 4 Floor 5 Grid ceiling 10 Target person detection means 15 Wide area sensor 16 Gate sensor 17 Spot sensor 17A Human sensor 17B Three-dimensional distance detection means (detection camera)
20 Data processing means 30 Information assist means P, P1, P2, P3 Person to be detected

Claims (9)

  In the position detection of the detection target person or the detection target attached with the specific ID in the floor to be detected in the building where the grid ceiling is constructed, the position detection of the detection target in the building for detecting the behavior, and the action recognition system, A kind of detection means is provided on the ceiling plate of the grid ceiling, the ceiling plate is arranged at a predetermined position of the grid ceiling, and the detection means or the detection object is detected by the detection means arranged at each position. A position detection and action recognition system for a detection target in a building characterized by performing position detection and action recognition with stepwise accuracy.   The detection means includes a wide area sensor provided on the ceiling plate so that the entire floor is set as a detection range, and a gate sensor provided on a ceiling plate disposed at a location passing between zones defining the floor. A spot sensor provided on the ceiling plate in the zone so that detailed detection in each zone is possible, and from the information for each detection range of each sensor, the position detection of the detection target person or the detection target, the behavior recognition The position detection and behavior recognition system for a detection target in a building according to claim 1, which is performed with stepwise accuracy.   The position detection and behavior recognition system for a detection target in a building according to claim 2, wherein the wide area sensor detects the position of a specific ID existing in the floor.   The position detection and action recognition system according to claim 2, wherein the gate sensor performs movement and position detection of a specific ID for each zone.   The said spot sensor performs the detailed position detection of specific ID in the said zone, and action recognition, The position detection of the detection target in a building of Claim 2, The action recognition system of Claim 2 characterized by the above-mentioned.   The support information is provided for the action of the specific ID according to the position information and action recognition information of the specific ID obtained by the detecting means. Position detection and behavior recognition system in the building described in 2.   The position detection / behavior recognition system according to claim 6, wherein the support information is provided to the specific ID for action support of the specific ID.   The position detection / behavior recognition system according to claim 6, wherein the support information is provided to another specific ID for the action support of the specific ID.   In accordance with the layout or zone change in the floor, etc., each detection means provided on the ceiling plate of the grid ceiling is expanded and moved for each ceiling plate according to the new system layout, after the change 2. The position detection and behavior recognition system of a detection target in a building according to claim 1, wherein detection corresponding to a layout or a zone of the building can be performed.

Images