JP2010205191A - Safety management device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a safety management device for appropriately managing user safety on a station platform by appropriately determining whether or not the behavior of a user may be peculiar on the station platform. <P>SOLUTION: The safety management device 1 processes pickup images of a camera 2 picking up the station platform and detects the users arriving at the station platform. The safety management device 1 further registers a second leaving train since the user's arriving at the station platform as a determined train of the user, using operation management data of trains for each detected user. The safety management device 1 determines and outputs that the behavior of the user may be peculiar when it detects the user whose associated determined train has left on the station platform. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a safety management device that manages user safety in a station platform.

  Conventionally, in order to ensure the safety of a user who uses an escalator, there is one that monitors a user's behavior and outputs a message to that effect when a user who takes a specific behavior is detected (Patent Document 1). 2). These devices set the escalator entrance to the target area, and determine that the user's behavior is unique when the user stays in the target area for a predetermined time. Moreover, the user in the target area is also monitored by processing the captured image of the camera that is capturing the target area.

JP 2000-53361 A JP 2000-115750 A

  However, as in Patent Documents 1 and 2, etc., in the configuration in which whether or not the user's behavior is peculiar is determined by the residence time, it is appropriately determined whether or not the user's behavior at the station platform is peculiar. Can not do it. Specifically, the user's residence time at the station platform is affected by the timing at which the user arrives at the station platform and the train arrival and departure times. Also, after arriving at the station platform because the train is congested or arriving at the destination early, do not take the first train that departs from the station platform. There are relatively many users who get on the next train that departs from the home. For this reason, it cannot be appropriately determined whether or not the user's behavior is peculiar based on the residence time of the user at the station platform.

  An object of the present invention is to provide a safety management device capable of appropriately managing the safety of a user at a station platform by appropriately determining whether or not the behavior of the user at the station platform is unique.

  In order to achieve the above object, the safety management device of the present invention is configured as follows.

  A person detection means processes the captured image of the imaging device that images the station platform, and detects the person being imaged. For each person detected by the person detection means, the determination train correspondence means associates a plurality of trains that depart from the station platform as determination trains after the timing when the person is detected. The person detection means may be configured to always detect a person, or may be configured to detect a person at a predetermined timing. For example, detection of a person at the station platform at a certain time before the arrival time of the train at the station platform (for example, 30 seconds before) or after a certain time (for example, 1 minute) has elapsed since the departure of the train at the station platform. May be performed.

  In addition, in the station service system, the operation status of the train is managed by the operation management device, so by using the operation management data managed by this operation management device, it is possible to identify a judgment train to be associated with each detected person. .

  Moreover, a determination means determines that this person's action is peculiar, when a person detection means detects the person from whom the corresponding determination train departed from the said station platform. That is, a user who has been staying at the station platform even though a plurality of trains have departed from the station platform after arriving at the station platform is determined to have a unique behavior. When the determining means determines that the action of a certain person is unique, the output means outputs that effect.

  In this way, for each user, whether the behavior of the user is unique depending on the number of trains that depart from the station platform while the user is staying at the station platform. Determine if. Therefore, it is possible to appropriately determine whether or not the behavior of the user at the station platform is unique, and the safety management of the user at the station platform can be appropriately performed.

  For example, if the user arrives at the station platform and the second train is the judgment train, the train arrives at the station platform because the train is congested or arrives at the destination early. After that, the user does not get on the first train that departs from the station platform, and does not determine that the behavior of a user car that has passed this train and got on the next train is unique. Moreover, since the user who passes two trains is rare, the user who passed two trains can determine that the action is peculiar.

  In addition, even if the trains depart from the same station platform, there are trains with different types such as express and ordinary, and destinations (trains with different stopping stations). It is also possible to determine a judgment train to be associated with a person. For example, the determination train to be associated with each user is not the second train that departs after the user is detected, but the train that departs first after the timing at which the person was detected The train and the second train having the same stop station may be used as the determination train.

  According to the present invention, it is possible to appropriately determine whether or not the behavior of the user at the station platform is unique, and the safety management of the user at the station platform can be performed appropriately.

It is a figure which shows the structure of the outline of the monitoring system to which a safety management apparatus is applied. It is a block diagram which shows the structure of the principal part of a safety management apparatus. Judgment table It is a flowchart which shows operation | movement of a safety management apparatus.

  Hereinafter, the safety management apparatus which is embodiment of this invention is demonstrated.

  FIG. 1 is a diagram showing a schematic configuration of a monitoring system to which a safety management device according to an embodiment of the present invention is applied. This monitoring system is a system that detects a user whose behavior at a station platform is unique and notifies a staff member of the station to that effect in order to manage the safety of the users at the station platform. This monitoring system includes a safety management device 1, a camera 2, an operation management device 3, and a notification device 4.

  The safety management device 1 detects a user whose behavior at the station platform is unique. What is necessary is just to arrange | position the camera 2 so that the captured image of the substantially whole station platform may be obtained. In other words, the camera 2 may be one camera 2 as long as a captured image of substantially the entire station platform is obtained, and the station platform is divided into a plurality of areas, and a plurality of cameras 2 are arranged for each divided area. May be. The camera 2 inputs the captured image to the safety management device 1. The camera 2 is, for example, a video camera that outputs a captured image of several tens of frames per second. The operation management device 3 is a device that manages the operation status including train arrival and departure times for each station platform. Moreover, this operation management apparatus 3 has memorize | stored the operation management data which shows the operation condition of a train. In addition, when the train delay occurs, the operation management data is updated according to the generated delay. The safety management device 1 and the operation management device 3 are connected via a data communication line so that data communication is possible. The safety management device 1 acquires operation management data through data communication with the operation management device 3. The notification device 4 includes a display panel and a speaker, and performs various types of notifications with images and sounds for station staff. The notification device 4 also performs a notification instructed from the safety management device 1.

  FIG. 2 is a block diagram illustrating a configuration of a main part of the safety management device. The safety management device 1 includes a control unit 11, an image processing unit 12, a storage unit 13, a communication unit 14, and an output unit 15. The control part 11 controls the operation | movement of each part of the safety management apparatus 1 main body, and performs the detection process etc. of the user whose action in a station platform is peculiar. The image processing unit 12 processes the captured image input from the camera 2 and individually detects a user located at the station platform. The image processing unit 12 may be provided for each camera 2 or may be provided for each group to which a plurality of cameras 2 belong. The grouping of the cameras 2 may be performed in units of station platforms, for example.

  The storage unit 13 stores an operation program for operating the main body and also stores a determination table shown in FIG. This determination table is for each user who detects that he / she is located at a station platform, a determination train for determining whether or not the user's feature amount and the user's behavior are unique, and the user It is the table which matched the detection time which detected this. This determination table is created for each station platform. In FIG. 3, the determination table in the station platform A is shown. The judgment train will be described later.

  The communication unit 14 performs data communication with the operation management device 3. The safety management device 1 acquires operation management data including train departure and arrival times for each station platform by data communication with the operation management device 3. The output unit 15 is connected to the notification device 4 and outputs to the notification device 4 an instruction for screen display and voice notification related to a notification of a user whose behavior at the station platform is unique.

  Next, the operation of the safety management device 1 will be described.

  Generally, a user who gets on a train gets on the first train that departs from the station platform after arriving at the station platform. In addition, after arriving at the station platform because the train is congested or arriving at the destination early, the train that departs from the station platform is not used. There are relatively many users who get on the next train that departs from the home. However, there are few users who travel on two or more trains. In other words, the behavior of users who travel on two or more trains is unique. When the safety management device 1 detects a user who has passed two or more trains after arriving at the station platform, the safety management device 1 outputs the user as a user whose behavior is unique. Here, for each user, after detecting that the user is located at the station platform, the second train departing from the station platform is set as a determination train, and is registered in the determination table described above.

  In addition, here, a user who has passed two or more trains is determined as a user whose behavior is unique. However, the number of trains that have been missed is not limited to two or more, but three or more. It may be. In this case, the determination train may be the third train or the fourth train.

  FIG. 4 is a flowchart showing the operation of the safety management device. The safety management device 1 performs the process shown in FIG. 4 for each station platform. Here, one station platform A will be described as an example.

  The safety management device 1 captures an image captured by the camera 2 that images the station home A when the first time before the arrival time of the train at the station home A (for example, 30 seconds before) (s1, s2). The timing for capturing the captured image of the camera 2 can be determined by the train operation status acquired from the operation management device 3. The safety management device 1 processes the captured image of the camera 2 captured in s2 in the image processing unit 12, and extracts a feature amount for each user who is captured (s3). In s3, it is only necessary to extract the feature amount of the type of feature that the individual user can specify with a certain degree of accuracy. For example, it may be a feature amount of the user's face, or a feature amount such as clothes or a system.

  The safety management device 1 determines whether or not each user whose feature value is extracted in s3 is a user already registered in the determination table (s4 and s5). For this determination, for each user, the feature amount extracted in s3 this time is collated with each feature amount registered in the determination table, and some of them have a similarity exceeding a predetermined threshold value. For example, it is determined that the user is registered. In other words, if none of the feature quantities registered in the determination table exceeds a predetermined threshold, it is determined that the user is not registered (a newly detected user). judge.

  If the safety management device 1 determines that the user is not registered in the determination table in s4, the safety management device 1 registers the feature amount, the determination train, and the detection time in association with each other (s6). The feature amount is the feature amount extracted in s3. The determination train is a second train that departs from the station platform A thereafter. The determination train can be specified by the train operation management data acquired from the operation management device 3. The detection time is the time at this time, not the time when the user actually arrived at the station platform A.

  Next, when the second time (for example, about 1 minute) elapses after the departure of the train arriving at the station platform A, the safety management device 1 takes in the captured image of the camera 2 that captures the station platform A again. (S7, s8). This fixed time is determined in consideration of the time when the user who gets off the train arriving at the station platform A moves to the ticket gate and disappears from the station platform A. By appropriately setting this fixed time, it is possible to prevent a user who gets off the train from being a target of this processing, and to prevent the processing load of the safety management device 1 main body from being increased unnecessarily.

  The safety management device 1 processes the captured image of the camera 2 captured in s8 in the image processing unit 12, and extracts a feature amount for each user who is captured (s9). s9 is the same process as s3. The safety management device 1 determines whether each user whose feature value is extracted in s9 is a user already registered in the determination table (s10, s11). The process concerning s10 is the same as s4 mentioned above. If the safety management device 1 determines that the user is not registered in s10, the safety management device 1 registers the feature value, the determination train, and the detection time in association with each other (s12). Here, many of the users registered in the determination table are users who arrive at the station platform A immediately after the previous train departs. The process related to s12 is the same process as s6 described above.

  Further, when the safety management device 1 determines that the user is registered in s10, whether or not the determination train registered in the determination table for the vehicle used is a train that has already started from the station platform A. Determine (s13). That is, in s13, it is determined whether this user is a user who has passed two trains. If the safety management device 1 determines in s13 that the determination train has already departed, the safety management device 1 determines that the user's behavior in the station platform A is unique, and outputs that fact from the output unit 15 (s14). ). In s14, for example, the output unit 15 outputs a message output instruction to the notification device 4 such as “A user with a strange behavior is in the station platform A. Please respond to the station staff nearby.” The notification device 4 outputs this message by display or voice, and notifies a station staff or the like that a user whose behavior is unique is at the station platform A. At this time, the captured image of the camera 2 capturing a user with a specific behavior may be input to a monitor installed in the station office or the like so that it can be confirmed in the station office.

  As described above, the safety management device 1 determines whether or not the behavior of the user at the station platform is peculiar, based on the number of trains that have been overtaken (here, two). Management can be performed appropriately according to the train operation status.

  Also, the processing of s3 to s8 shown in FIG. 4 is deleted, and the processing of s9 to s14 is performed using the captured image of the camera 2 before the arrival time at which the train arrives at the station platform A. May be. In this case, the detection of the user whose action is unique is not immediately after the departure of the determination train but immediately before the arrival of the next train of the determination train. That is, detection of a user whose action is unique is delayed in time from the above example, but the processing load of the safety management device 1 can be reduced.

  Also, the processing of s1 to s6 shown in FIG. 4 may be deleted, and the processing of s9 to s14 may be performed using the captured image of the camera 2 a predetermined time before the departure time from the station platform A. Good. In this case, it cannot be determined whether the timing when the user arrived at the station platform was before the departure of the train that departed immediately before or after the departure. For this reason, after arriving at the station platform A, there are users who use the second train as the determination train, and there are users who use the third train as the determination train. Therefore, although the detection of the user whose behavior is unique may be delayed in time from the example shown in FIG. 4, the processing load of the safety management device 1 can be reduced.

  In the above explanation, when two trains are passed, it is determined that the behavior is peculiar. However, from the same station platform, different types of trains such as express and normal, X station, Y station Trains with different destinations may depart. Considering this, the judgment train that determines that the behavior is peculiar is not a train that departs after the user has arrived at the home, but is simply the second train that departs. It is good also as the 2nd train with the same stop station from a station. For example, at a station platform where express and ordinary trains depart alternately, after the user arrives at station platform A, the first (first) express train, the second ordinary train, and the third express train depart Later, if this user is located at the station platform A, the behavior may be determined to be unique. In this case, even if this user is detected at the station platform A until the second ordinary train departs and the third express train departs, it is not determined that the behavior is unique. As described above, after the user arrives at the station platform A, the second train (the third express train in the above example) having the same stop station from the own station may be used. Similarly, when the first train (the first train) that departs after arrival of the user at the station platform A is a regular train, the third train that departs may be used as the determination train (second train). The train that departs from is an express train.)

  In addition, in the station platform where the train whose destination is X station and the train where Y station departs alternately, the first (first) train (the destination is X station) after the user arrives at station platform A Train) After the second train (the train whose destination is Y station) and the third train (the train whose destination is X station) depart, if this user is located at station platform A The configuration may be such that the behavior is determined to be unique. In this case, this user is detected at station platform A until the second train (the train whose destination is Y station) departs and the third train (the train whose destination is X station) departs. Even so, the action is not determined to be unique.

  Further, in the above description, the case where the determination train of the user is determined by distinguishing the type such as express or normal and the destination is exemplified, but the determination train may be determined in consideration of both of them. Good.

  The user information registered in the determination table may be deleted when it becomes unnecessary. For example, if a user is registered in the determination table and is not detected twice in succession by the detection of the user at the station platform A, the user is assumed to have moved on the train. The information relating to the user may be deleted from the determination table. Alternatively, it may be deleted from the determination table when the elapsed time from the detection time exceeds a predetermined time such as one hour. In this case, the elapsed time until deletion is set to a longer time than the determination train for the user departs.

  Moreover, it is good also as a structure which updates a judgment train also about the user who has already decided the judgment train and has registered into the judgment table, when a train delay, a train suspension, etc. generate | occur | produce and an operation condition changes. . In this way, it is possible to cope with train delays, suspension of service, and the like.

  In addition, for each user detected at the station platform, the captured image of the camera 2 is processed and the movement of the user is monitored while tracking the movement of the user at the station platform. If you are at home, you may determine that the behavior is unique.

1-Safety management device 2-Camera 3-Operation management device 4-Notification device 11-Control unit 12-Image processing unit 13-Storage unit 14-Communication unit 15-Output unit

Claims (4)

Human detection means for processing a captured image of an imaging device that images a station platform and detecting a captured person;
For each person detected by the person detection means, a determination train correspondence means for associating a plurality of trains departing from the station platform as a determination train after the timing of detecting the person,
Determination means for determining that the person's behavior is unique when the person detection means detects a person whose associated determination train departs from the station platform;
An output unit that outputs an output to the effect when it is determined that the behavior of a certain person is unique.   2. The safety according to claim 1, wherein the person detection unit is a unit that processes a captured image captured by the imaging device and detects a captured person before a predetermined time before a train arrives at the station platform. Management device.   The determination train correspondence means is a means for associating a second train having the same stop station as a determination train with a train that departs earlier among trains that start after the timing at which the person is detected. Item 3. The safety management device according to item 1 or 2.   The safety according to claim 1, wherein the person detection unit is a unit that processes a captured image input to the captured image input unit and detects a feature amount of the face of the person being captured. Management device.

Images