JP2014148300A - Monitoring device, method, program, or system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a monitoring device, method, program, and system that appropriately monitor an area between platform doors and a vehicle and pose a little fear of erroneously deciding on presence or absence of an obstacle in the area.SOLUTION: A monitoring device includes: an image acquisition control unit; and state decision unit. The image acquisition control unit includes: a first image acquisition control block that acquires as a first image an image obtained during a first time interval that lasts until platform doors begin opening after a vehicle halts at a platform; and a second image acquisition control block that acquires as a second image an image obtained during a second time interval that lasts until the vehicle begins moving after the platform doors are closed. The state decision unit decides on presence or absence of an obstacle in a space between the platform doors and vehicle doors on the basis of a result of collation between the first image and second image that are acquired.

Description

  The present invention relates to an apparatus, method, program, or system for monitoring between a platform door and a vehicle.

  2. Description of the Related Art In recent years, platform doors have been installed on landing platforms for track vehicles (hereinafter referred to as platforms) from the standpoint of preventing passengers from falling onto tracks or preventing contact with traveling vehicles (platform doors). As an example, see Patent Document 1).

  In addition, proposals have been made to further improve the safety of passengers or the operation of vehicles by imaging the area between the platform door and the track or the vehicle.

  For example, Patent Document 2 discloses a platform door provided with an imaging unit, and a platform tip clearance adjustment device that fills the gap between the platform and the vehicle and prevents the passenger from falling on the track. When the image when there is no person or object between the platform door and the train is used as a reference image, and this reference image is compared with the actual captured image, Proposals have been made regarding accident prevention devices that do not release the platform tip clearance adjustment device.

JP 2000-016280 A JP 2003-146201 A

  However, in the apparatus described in Patent Document 2, since the reference image is fixed, an abnormal state can occur even when an object that has no safety problem is temporarily placed in the actual captured image. It may be determined that In addition, if the imaging condition of the reference image (for example, the light condition) is significantly different from the imaging condition at the time of actual image capturing, there is a risk of malfunction due to erroneous determination.

  The present invention has been made paying attention to the above-mentioned problems, and the object of the present invention is to appropriately monitor the area between the platform door and the vehicle, thereby making an error regarding the presence or absence of obstacles in the area. An object is to provide a monitoring apparatus, method, program, and monitoring system with a low risk of determination.

  Other objects and effects of the present invention will be easily understood by those skilled in the art by referring to the following description of the specification.

  The above technical problem can be solved by the monitoring apparatus, method, program, and monitoring system of the present invention having the following configuration.

  That is, the monitoring apparatus according to the present invention includes an image acquisition control unit that acquires an image captured by the imaging unit, and a state determination unit.

  The imaging unit is arranged to be able to image a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle stopped on the platform.

  The image acquisition control unit acquires an image obtained through the imaging unit as a first image during a first time interval until the platform door starts to open after the vehicle stops on the platform. A second image acquisition control unit and an image acquired through the imaging unit as a second image during a second time interval until the vehicle starts to move after the platform door is closed. An image acquisition control unit.

  The state determination unit determines the presence or absence of an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. To do.

  According to such a configuration, since a vehicle such as a train stops on the platform, the obstacle determination process is performed based on the image captured in the monitoring area in the time interval until the platform door starts to open. Judgment processing can be performed on the basis of the most recent image before the vehicle door and the platform door are opened. This allows temporary reflection of objects having no safety problem in the same area, and changes in imaging conditions (light Etc.) can be avoided, and monitoring with few erroneous determinations can be performed.

  In the monitoring apparatus according to the present invention, the first time period from when the vehicle stops on the platform until the platform door starts to open is detected by receiving the platform door opening start signal. It may be a thing.

  According to such a configuration, it is possible to reliably image the state immediately before the vehicle door and the platform door are opened.

  In the monitoring apparatus according to the present invention, the fact that the vehicle is in the second time interval from when the platform door is closed until the vehicle starts to move is detected by receiving the platform door closing signal. It may be.

  According to such a configuration, after the platform door is closed, the time interval until the vehicle starts to move is detected based on the platform door closing completion signal, so the reference image and the current image are reliably verified. It can be carried out.

  In the monitoring apparatus according to the present invention, the image acquisition control unit performs time series via the imaging unit during a third time interval from when the vehicle door starts to close until the platform door completely closes. A first continuous image acquisition unit that acquires a plurality of images obtained as a continuous image, wherein the state determination unit is configured to determine whether the platform door and the vehicle door are based on the acquired continuous image. It may be one that determines the presence or absence of a moving object in a space sandwiched between them.

  According to such a configuration, in the time interval from when the vehicle door starts to close until the platform door completely closes, the moving body is determined based on a plurality of time-series images captured in the time interval. Since the warning information is generated, it is possible to grasp the presence / absence of the moving object immediately before the obstacle determination, the movement history of the moving object that can become an obstacle later, and the like.

  At this time, the start time of the third time interval may be detected by receiving a closing signal for starting the platform door, and the end time may be detected by receiving a closing signal for closing the platform door. .

  According to such a configuration, it is detected based on the closing start signal and the closing completion signal related to the platform door that the vehicle door is in the third time interval from the start of closing until the platform door is completely closed. doing. Therefore, it is possible to reliably specify the time interval immediately before the platform door completes closing.

  In the monitoring device according to the present invention, the state determination unit uses, as a monitoring region, a predetermined region sandwiched between the platform door and a vehicle door of a vehicle that stops on the platform in the continuous image. The presence or absence of a moving object may be determined.

  According to such a configuration, the movement of a person or the like in the vehicle or the movement of a pedestrian or the like on the platform can be excluded from the detection target in advance, and the calculation processing load of moving body detection can be suppressed.

  In the monitoring device according to the present invention, the state determination unit excludes the presence or absence of a moving object while recognizing the reflection of a passenger on the wall surface of the platform door as a moving object in the monitoring area. It may be a thing to judge.

  According to such a configuration, it is possible to appropriately eliminate a false detection factor that should not be detected as a moving body.

  In the monitoring device according to the present invention, the predetermined area that can be imaged by the imaging unit further includes a track area when the vehicle is absent, and the image acquisition control unit includes: When the vehicle is absent, the vehicle further includes a second continuous image acquisition unit that acquires a plurality of images obtained in time series via the imaging unit as a continuous image, and the state determination unit includes the state determination unit when the vehicle is absent The presence / absence of a moving object in the predetermined area when the vehicle is absent may be determined based on the acquired continuous images.

  According to such a configuration, since the monitoring device is utilized not only when the vehicle is present but also when the vehicle is absent, the device can be used effectively, and access to the track or access to the platform door is possible. Etc. can be wary. This can further improve the safety of passengers and navigation.

  In the monitoring apparatus according to the present invention, the imaging unit includes two imaging units that capture images of the space sandwiched between the platform door and the vehicle door from different directions, and the two imaging units. One of these may alternatively be operable.

  According to such a configuration, a captured image under a more preferable imaging condition can be obtained, and erroneous determination can be further reduced.

  The present invention viewed from another aspect can also be understood as a monitoring system. That is, the monitoring system corresponds to a monitoring device provided in each door of the platform door, an imaging device connected to each monitoring device, a management device that manages those monitoring devices, and the management device. And a display device provided.

  In addition, the monitoring device includes an imaging device, an image acquisition control unit, a state determination unit, and a transmission unit.

  The imaging device is arranged to be able to image a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle stopped on the platform.

  The image acquisition control unit acquires, as a first image, an image obtained through the imaging device in a first time interval until the platform door starts to open after the vehicle stops on the platform. A second image acquisition control unit and a second time interval after the platform door is closed until the vehicle starts moving, and an image obtained through the imaging device is acquired as a second image. An image acquisition control unit.

  The state determination unit determines the presence or absence of an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. To do.

  When the transmission unit determines that there is an obstacle in the space based on the determination in the state determination unit, the transmission unit sends an abnormal state signal indicating that there is an obstacle in the space and the second image to the management device. Send to.

  The management device includes a receiving unit and a display control unit.

  The receiving unit receives the abnormal state signal and the second image received from the monitoring devices in association with the monitoring devices.

  The display control unit determines whether any one of the platform doors is in an abnormal state with respect to the display device based on the abnormal state signal received in association with each monitoring device and the second image. Display control is performed so as to display information indicating whether or not there is a second image corresponding to the door.

  According to such a configuration, it can be easily confirmed through the display unit, along with the image, which door has an abnormality.

  In the monitoring system according to the present invention, the monitoring device further includes an image enhancement information generation unit, and the image enhancement information generation unit is configured to check the comparison result between the first image and the second image. Based on this, image enhancement information for highlighting an area determined to be an obstacle is generated, the transmission unit of the monitoring device further transmits the image enhancement information, and the reception unit of the management device further includes The image enhancement information received from each monitoring device is received in association with each monitoring device, and the display control unit of the management device further receives the second image received in association with each monitoring device. And an image in which an area corresponding to an obstacle is emphasized may be displayed on the display device based on the image enhancement information.

  According to such a configuration, a portion corresponding to an obstacle can be easily grasped.

  In the monitoring system according to the present invention, the image acquisition control unit further includes, via the imaging device, a third time interval from when the vehicle door starts to close until the platform door completely closes. The image processing apparatus further includes a first continuous image acquisition unit that acquires a plurality of images obtained in a time series as a continuous image, and the state determination unit further includes the platform door and the based on the acquired continuous image. The presence or absence of a moving object in a space sandwiched between the vehicle door is determined, and the transmission unit further determines that there is a moving object in the space based on the determination in the state determination unit. And the second image is transmitted to the management device, and the reception unit further receives the warning state signal received from each of the monitoring devices and the previous image. The second image is received in association with each monitoring device, and the display control unit is further based on the warning state signal received in association with each monitoring device and the second image, Even if display control is performed on the display device to display information indicating which of the platform doors is in the alert state and a second image corresponding to the door. Good.

  According to such a configuration, in the time interval from when the vehicle door starts to close until the platform door completely closes, the moving body is determined based on a plurality of time-series images captured in the time interval. Since the warning information is generated, it is possible to grasp the presence / absence of the moving object immediately before the obstacle determination, the movement history of the moving object that can become an obstacle later, and the like. In addition, it is possible to easily confirm which door is in the alert state through the display unit together with the image.

  In the monitoring system according to the present invention, the monitoring system is further connected to the management device, and is connected to the integrated control device for managing the opening and closing of the entire platform door, and the integrated control device. An individual control device that individually controls the opening and closing of each door of the door, wherein the management device further includes an abnormality / warning information transmission unit, and the abnormality / warning information transmission unit is abnormal in the reception unit When a status signal is received, an abnormal status signal is transmitted to the integrated control device together with information on the monitoring device related to the abnormal status signal, and a warning status signal is received at the receiving unit. , To the integrated control device, a warning state signal is transmitted together with information on the monitoring device related to the warning state signal, and the integrated control device further includes an opening signal transmission unit, When the opening signal transmission unit receives the abnormal state signal, the opening signal transmission unit transmits the opening signal to the individual control device that controls the door monitored by the monitoring device related to the abnormal state signal. When the warning state signal is received, the door control signal is not transmitted to the individual control panel, and the individual control panel responds to the door opening signal from the integrated control panel. Control of opening of the door may be performed.

  According to such a configuration, the platform door in an abnormal state can be automatically opened again.

The present invention viewed from another aspect can also be grasped as a monitoring method. That is, this monitoring method is obtained via an imaging unit arranged so as to be capable of imaging a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle parked on the platform. A monitoring method based on image data obtained,
Acquiring the image obtained through the imaging unit as a first image in a first time interval after the vehicle stops on the platform until the platform door starts to open; and
Acquiring the image obtained through the imaging unit as a second image in a second time interval until the vehicle starts to move after the platform door is closed; and
Determining whether or not there is an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. It is characterized by that.

  According to such a configuration, since a vehicle such as a train stops on the platform, the obstacle determination process is performed based on the image captured in the monitoring area in the time interval until the platform door starts to open. Judgment processing can be performed on the basis of the most recent image before the vehicle door and the platform door are opened. This allows temporary reflection of objects having no safety problem in the same area, and changes in imaging conditions (light Etc.) can be avoided, and a monitoring method with few erroneous determinations can be provided.

The present invention viewed from another aspect can be grasped as a computer program. That is, this computer program is obtained via an imaging unit arranged so as to be capable of imaging a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle parked on the platform. A computer program for executing a monitoring method based on image data obtained by a computer,
Acquiring the image obtained through the imaging unit as a first image in a first time interval after the vehicle stops on the platform until the platform door starts to open; and
Acquiring the image obtained through the imaging unit as a second image in a second time interval until the vehicle starts to move after the platform door is closed; and
Determining whether or not there is an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. It is characterized by that.

  According to such a configuration, since a vehicle such as a train stops on the platform, the obstacle determination process is performed based on the image captured in the monitoring area in the time interval until the platform door starts to open. Judgment processing can be performed on the basis of the most recent image before the vehicle door and the platform door are opened. This allows temporary reflection of objects having no safety problem in the same area, and changes in imaging conditions (light Etc.) can be avoided, and a monitoring program with few erroneous determinations can be provided.

  According to the monitoring device, method, program, and system, the obstacle determination process is performed based on the image captured in the monitoring area in the time interval after the vehicle stops until the platform door starts to open. It is possible to provide a monitoring device, method, program, and monitoring system that are less likely to be erroneously determined regarding the presence or absence of an obstacle in the region between the platform door and the vehicle.

FIG. 1 is an explanatory diagram of the entire system. FIG. 2 is a configuration diagram of the monitoring unit. FIG. 3 is a general flowchart. FIG. 4 is a detailed flow of the reference image acquisition process. FIG. 5 is a detailed flow of the precaution processing. FIG. 6 is a detailed flow of the moving object detection process. FIG. 7 is an explanatory diagram of a detection target area at the time of preliminary warning. FIG. 8 is a detailed flow of the obstacle confirmation process. FIG. 9 is a flowchart of the obstacle determination process. FIG. 10 is an explanatory diagram of a determination example when there is an obstacle. FIG. 11 shows a display example on the display unit. FIG. 12 is a processing flow when the door is opened. FIG. 13 is a detailed flow of the monitoring process when the vehicle is absent. FIG. 14 is a detailed flow of the moving object detection process when the vehicle is absent. FIG. 15 is a display example on the display unit of the moving object detection process when the vehicle is absent. FIG. 16 shows another arrangement example of the imaging unit.

<1. First Embodiment>
Hereinafter, the first embodiment of the present invention will be described with reference to FIGS. 1 to 12.

<1.1 System configuration>
FIG. 1 shows an overview of a system according to the first embodiment.

  A vehicle 1 stopped on the platform is connected to a host device 2 relating to ATO (Automatic Train Operation) or the like. The host device 2 is connected to a general control panel 3 that mainly manages the opening and closing of the platform door. Further, the integrated control panel 3 is connected to a management unit 5 that manages a plurality of monitoring units 7 that monitor a predetermined area between the platform door and the vehicle 1. In addition, the management unit 5 is connected to a display unit 6 for displaying the abnormality / warning information and the monitoring image transmitted from each monitoring unit 7. Note that the various devices, control panels, and units described above may be dedicated machines that exhibit a specific function, a calculation unit such as a CPU, a storage unit such as a memory or a hard disk, or wired or connected to another device. It may be a general-purpose machine including any or all of a communication unit for performing wireless communication, an input interface such as a keyboard and buttons, and an output interface such as a monitor and a printer.

  Next, the flow of signals transmitted and received when the vehicle door and the platform door are opened and closed will be described with reference to FIG. 1 again. First, the flow of signals when the vehicle and the platform door are opened will be described.

  The vehicle 1 that has entered the platform and stopped transmits a vehicle door opening signal indicating that the vehicle 1 is to be opened to the host device 2 related to ATO or the like. The host device 2 that has received the opening signal transmits the opening signal of the vehicle door to the general control panel 3 that controls the opening and closing of the platform door.

  After receiving the door opening signal of the vehicle, the integrated control panel 3 transmits the opening signal of the platform door to the individual control panel 4 that individually controls each door of the platform door, and the open / close state of the platform door Is transmitted to the management unit 5. Thereafter, the individual control panel 4 receives the opening signal and opens each door of the platform door. The open / close state signal related to the open / close state of the platform doors transmitted from the general control panel 3 to the management unit 5 includes an opening start signal indicating that each door of the platform doors starts to open. In other words, the integrated control panel 3 transmits a signal indicating an opening start timing indicating the opening of the platform door to the management unit 5. The management unit 5 transmits a platform door open / closed state signal to each monitoring unit 7. The signal related to the open / close state of the platform door is used in the platform door monitoring process described later.

  Also, at the same time as or after the transmission of the door opening signal to the individual control board 4 by the above-mentioned general control board 3, the general control board 3 confirms reception of the vehicle door opening signal to the host device 2. A signal (reception ACK) is transmitted. The host device 2 receives the reception confirmation signal and transmits a reception confirmation signal (reception ACK) to the vehicle 1, and the vehicle 1 opens the vehicle door after receiving the reception confirmation signal. As described above, the opening process of the vehicle door and the platform door is performed.

  Next, the flow of signals when the vehicle and the platform door are closed will be described. The vehicle 1 performs a vehicle door closing process and transmits a vehicle door closing signal to the host device 2. The host device 2 receives the closing signal and transmits a vehicle door closing signal to the general control panel 3. The general control panel 3 receives a vehicle door closing signal and transmits a door closing signal to the individual control panel 4. In addition, each individual control panel 4 performs door closing processing and transmits an open / closed state signal of each door indicating the open / closed state of each door of the platform door to the general control panel 3. The door open / close state signal includes a door closing completion signal indicating that the door has been closed.

  The comprehensive control panel 3 transmits a door open / closed state signal to the management unit 5. This open / closed state signal includes a door closing start signal indicating that each door of the platform door starts to close, a door closing completion signal indicating that each door has been closed, and the like. That is, the general control panel 3 transmits a signal representing the closing start timing, closing completion timing, etc. of each door of the platform door to the management unit 5. The management unit 5 transmits a platform door open / closed state signal to each monitoring unit 7. The signal related to the open / close state of the platform door is used in the platform door monitoring process described later. As described above, the door closing process of the vehicle door and the platform door is performed.

  Here, the on-line signal of the vehicle will be described. When the vehicle enters or stops on the platform, the host device 2 detects a standing line state indicating that the vehicle exists. This may be detected, for example, based on the establishment of a wireless connection between the host device 2 and the vehicle 1, or based on other sensor information (infrared sensor information, imaging sensor information, etc.). May be detected. After detecting the standing state of the vehicle 1, the host device 2 transmits a standing state signal indicating that the vehicle is present to the general control panel 3. The integrated control panel 3 that has received the standing line status signal transmits the standing line status signal to the management unit 5. Furthermore, the management unit 5 transmits a standing line state signal to each monitoring unit 7 based on the standing line state signal. That is, at least the host device 2, the general control panel 3, the management unit 5, and the monitoring unit 7 have information regarding the on-line state on the platform of the vehicle 1.

  On the other hand, when the vehicle goes out from the platform or when the vehicle is absent from the platform, the host device 2 detects a standing line state indicating the absence of the vehicle. This may be detected based on, for example, the disconnection of the wireless connection between the host device 2 and the vehicle 1, or based on other sensor information or the like (infrared sensor information or imaging sensor information or the like). May be detected. After detecting the standing state of the vehicle 1, the host device 2 transmits a standing state signal indicating that the vehicle is absent to the general control panel 3. The integrated control panel 3 that has received the standing line status signal transmits the standing line status signal to the management unit 5. Furthermore, the management unit 5 transmits a standing line state signal to each monitoring unit 7 based on the standing line state signal. That is, at least the host device 2, the general control panel 3, the management unit 5, and the monitoring unit 7 have information regarding the on-line state on the platform of the vehicle 1.

  As will be described later, each monitoring unit 7 monitors the predetermined area by performing an imaging process or the like in accordance with the door open / closed state signal, and there are obstacles or moving objects in the predetermined area. Is determined to be in an abnormal state or a warning state, and the abnormality / warning information and a monitoring image at the time of the abnormality / warning state are transmitted to the management unit 5.

  When the management unit 5 receives the above-described abnormality / warning information, monitoring image, or the like, the management unit 5 performs display processing so that an image corresponding to the information is displayed on the display unit 6. Further, the abnormality / warning information is transmitted to the general control panel 3.

  When the general control panel 3 receives the abnormality / warning information, the general control panel 3 transmits the information to the higher-level device 2 and transmits a door open signal to the individual control panel 4 corresponding to the door indicating the abnormal state. That is, an open signal is transmitted to the door that is determined to be in an abnormal state by the monitoring unit 7. The individual control panel 4 that has received the opening signal performs door opening control.

<1.2 Configuration of the monitoring unit>
Next, an arrangement example of the monitoring unit 7 on the platform will be described with reference to FIG.

  FIG. 2A shows an arrangement example of the monitoring unit 7 on the platform. The platform door 11 is provided on the platform 9 and plays a role of preventing a passenger from falling onto the track 10 and preventing contact with a train. The platform door 11 has a plurality of doors, and each door is composed of two slidable door plates 12. The door plate 12 of each door is controlled by a corresponding individual control panel 4 (not shown), and is configured to be stored in the door pocket 8.

  In the figure, the monitoring unit 7 is disposed in a storage case 13 provided on a side surface of the platform door 11 on the track 10 side together with an imaging unit 14 configured by a camera or the like. The storage housing may be integrated with the platform door or may be a separate body. The imaging unit 14 is arranged so as to be capable of imaging at least a predetermined area (hereinafter referred to as a monitoring area) including a space sandwiched between the platform door 11 and a vehicle door of a vehicle stopped on the platform. Controlled by unit 7. The monitoring unit 7 also includes an image acquisition control unit (not shown) that acquires an image captured by the imaging unit, and a state (abnormal state) of the predetermined area based on the image acquired by the image acquisition control unit. Or an information processing unit such as a state determination unit (not shown) for determining a warning state.

  FIG. 2B is a view of the monitoring unit 7 as viewed from the front. As described above, in the same figure, the storage housing 13 constituting the monitoring unit 7 is integrally provided on the side surface of the platform door 11 on the track 10 side. A part of the storage casing 13 is provided with a transmission part or hole for imaging, and the imaging part 14 is configured to be able to image the outside of the storage casing 13 through the transmission part or hole.

<1.3 Monitoring operation>
Next, a monitoring operation for monitoring a predetermined area including a space sandwiched between the platform door 11 and the vehicle door of the vehicle stopped on the platform will be described in detail with reference to FIGS. 3 to 10.

  FIG. 3 is a general flowchart for explaining an overview of the monitoring operation performed by the monitoring unit 7.

  First, after the vehicle such as a train stops on the platform (confirmed by the vehicle presence state signal or the like), the monitoring unit 7 images the monitoring area using the imaging unit in the time interval until the platform door 11 starts to open. The image in a state where both the platform door and the vehicle door are closed is acquired as a reference image, that is, a reference image (S100).

  Next, the monitoring unit 7 acquires, as continuous images, a plurality of images obtained in time series via the imaging unit in a time interval from when the vehicle door starts to close until the platform door completely closes. When it is determined based on the continuous images whether a moving object such as a person or an object exists in the monitoring area in the time interval, and it is determined that a moving object such as a person or an object exists in the monitoring area Is transmitted to the management unit 5 (S200).

  Thereafter, after the platform door 11 is closed, the monitoring unit 7 acquires, as a current image, an image obtained by imaging the monitoring area using the imaging unit in a time interval until the vehicle starts to move. And when the said reference image and the said current image are compared and a predetermined difference exists between both, in the said difference part, it is an abnormal state in which the obstruction which is not preferable for vehicle operation or a passenger's safety exists. If the abnormal state is determined, abnormal information is transmitted to the management unit 5 (S300).

  That is, according to such a configuration, the obstacle determination process is performed based on the image captured in the monitoring area in the time interval until the platform door 11 starts to open after the vehicle such as the train stops on the platform. Therefore, the judgment process can be performed on the basis of the most recent image before the vehicle door and the platform door are opened. It is possible to avoid malfunctions due to changes (light conditions, etc.), etc., and to perform monitoring with few erroneous determinations.

  Further, in a time interval from when the vehicle door starts to close until the platform door completely closes, the mobile object is determined based on a plurality of time-series images captured in the time interval, and warning information is generated. Therefore, it is possible to grasp the presence / absence of the moving object immediately before the final obstacle determination, the movement history of the moving object that can become an obstacle later, and the like.

<1.3.1 Reference Image Acquisition Processing>
Details of the reference image acquisition process (S100) will be described with reference to FIG. In addition, this process (S100) is performed on the assumption that the stop of the vehicle has been confirmed by, for example, a vehicle presence state signal or the like.

  The monitoring unit 7 performs the process of reading the received data (S101) until a platform door opening start signal is received (S102 NO). Then, when the door opening start signal transmitted from the management unit 5 is received (S102 YES), the monitoring unit 7 transmits an imaging command signal that requests the imaging unit 14 such as a camera to perform imaging (S103). ). After that, when receiving the imaging command signal, the imaging unit 14 captures a still image of the monitoring area and transmits the captured image to the monitoring unit 7.

  After transmitting the imaging command signal (S103), the monitoring unit 7 repeats the readout processing of the captured image until the captured image is acquired (S105 NO) (S104). Thereafter, when a captured image is acquired (S105 YES), a process of storing the captured image as a reference image in an internal memory or the like is performed (S106).

  That is, in the reference image acquisition process according to the present embodiment, the platform door opening signal indicates that the vehicle is in a time interval until the platform door 11 starts to open after a vehicle such as a train stops on the platform. Detect based on. Therefore, it is possible to reliably image the state immediately before the vehicle door and the platform door are opened. It should be noted that this reference image acquisition process is desirably performed every time a vehicle enters the platform. Thereby, since the state determination can be performed based on the most recent state, it is possible to perform monitoring or the like with fewer erroneous determinations.

  In the above description, the time interval from when the vehicle stops on the platform until the platform door starts to open is detected by the platform door opening start signal. However, the present invention is not limited to this. Therefore, for example, by using the standing line status signal, the starting point of the standing line on the vehicle platform or the time point after a predetermined time has elapsed from the time point may be detected via a timer, or the opening signal of the vehicle door is directly detected. It is good also as such a structure. Further, it may be detected that the vehicle has stopped on the platform via the image acquired by the imaging unit 14. Furthermore, it is detected that the vehicle door starts to open through the image acquired by the imaging unit 14, and the platform door and the vehicle door before one to a plurality of time steps (for example, a frame) from the detection time point are detected. The image when both are closed may be used as the reference image. According to such a detection configuration using the image itself, the monitoring unit 7 starts opening the platform door after the vehicle stops on the platform without obtaining the open / closed state signal of the platform door or the presence signal of the vehicle. Images in the time interval up to can be detected.

<1.3.2 Precautionary processing>
Next, a detailed flow of the preliminary alert process (S200) will be described with reference to FIGS. When the preliminary warning process is started, the monitoring unit 7 performs a reception signal reading process (S202) until it receives a closing start signal related to the platform door (NO in S202). If a platform door closing start signal is received (YES in S202), a moving body detection process described later is started (S203). Thereafter, until the closing signal for the platform door is received (NO in S205), the received signal is read again (S204). When the closing completion signal regarding the platform door is received (S205 YES), a process of ending the moving body detection process is performed (S206).

  That is, in the preliminary warning processing according to the present embodiment, the closing start signal and the closing completion signal relating to the platform door indicate that the vehicle door is in the third time interval from the start of closing until the platform door is completely closed. And based on the detection. Therefore, it is possible to reliably specify the time interval immediately before the platform door completes closing.

  Next, details of the moving object detection process will be described with reference to FIG. When the moving body detection process is started, a moving image imaging command is issued to the imaging unit, and the monitoring unit 7 acquires at least two or more images from the moving images (S2031). Here, the two or more images are images that are referred to for difference detection for detecting a moving object. For example, when an image at the current time t is I (t), I (t) It may be a combination of I (t-1), a combination of I (t) and I (t + 1), or I (t) and I (t-2) or I (t ) And I (t + 2). Further, when the moving object detection process or the like is performed with reference to three or more images, for example, a combination of I (t−1), I (t), and I (t + 1) may be used, and I (t -2), I (t) and I (t + 2) may be combined. Needless to say, more images may be used.

  Next, the monitoring unit 7 performs a moving body detection process based on the two or more images (S2032). Various conventional methods can be used as the moving object detection processing method. For example, a technique for detecting a moving object based on an optical flow calculated using a gradient method, a block matching method, or the like, a technique for detecting a moving object based on a background difference, and the like are known.

  A detection target region in the case of performing detection processing of a moving body will be described with reference to FIG. In the same figure, the area | region between a vehicle door and a platform door is set as the detection target area | region 20 of a moving body (in the broken-line part of the figure). In this way, by setting a certain area as the moving object detection target area in the field of view of the imaging unit, it is possible to exclude the movement of a person in the vehicle or the movement of a pedestrian on the platform from the detection target in advance. In addition, it is possible to suppress the calculation processing load of the moving object detection.

  In addition, even within the detection target area 20, a false detection factor that should be detected as a moving object, although it is a difference in comparison between images, such as a shadow or a reflection on a home door surface such as a person in a vehicle, etc. Exists. In order to prevent such erroneous detection, it is preferable to use various known shadow exclusion filters, reflection exclusion filters, and the like.

  In the moving body detection process (S2032), when there is no moving body in the detection target area (NO in S2033), a normal notification process for notifying the management unit 5 that there is no moving body in the detection target area. This is performed (S2034). Specifically, the normal notification process is a process of transmitting a signal indicating that there is no moving body to the management unit 5, and together with the signal, the monitoring unit 7 that has performed the moving body detection process. An ID unique to the monitoring unit assigned to the image or an image determined to have no moving object may be transmitted. Note that when there is no moving object, the normal notification may be omitted.

  On the other hand, in the moving object detection process (S2032), if there is a moving object in the detection target area (S2033 YES), the management unit 5 has a moving object in the detection target area and is in a warning state. The warning notification process for notifying is performed (S2035). The alert notification process is specifically a process of transmitting a signal indicating that a moving body is present and is in a state to be alerted to the management unit 5, and detecting the moving object together with the signal. The ID unique to the monitoring unit assigned to the monitoring unit 7 that performed the processing, the image determined to have a moving body, and information for highlighting the moving body portion existing in the image (for example, movement Information representing the contour region of the part determined to be a body, rectangular region information for representing a rough moving body position including the contour region, color information for highlighting the moving body, etc.) may be transmitted.

  In the moving object detection process, the above process is continuously performed until the end of the moving object detection process (S206). Note that the image or the like at the time of the precaution processing may be stored in the internal memory of the monitoring unit or the internal memory of the management unit 5. According to such storage, it is possible to visually check a person who can be a hindrance later, the background of the object, and the like any number of times.

  In the above description, the time period from when the vehicle door starts to close until the platform door completely closes is detected based on the closing start signal and the closing completion signal related to the platform door. Not. Therefore, for example, it may be detected by directly using a vehicle door closing start signal and a door closing completion signal, or it may be configured to perform a preliminary warning process for a predetermined time required for the completion of the door closing from the time when the door closing start signal related to the platform door is received. Good. Further, the time point when the platform door or the vehicle door starts to close and the time point when the closing is completed may be detected based on the image acquired via the imaging unit 14.

<1.3.3 Obstacle confirmation process>
Next, details of the obstacle confirmation process (S300) using the reference image will be described with reference to FIGS.

  A detailed flow of the obstacle confirmation process will be described with reference to FIG. When the obstacle checking process is started, first, a reception signal reading process is performed until a platform door closing signal is received (NO in S302) (S301). When the closing signal is received (YES in S302), an obstacle determination process described later is performed (S303).

  As a result of the obstacle determination process (S303), when it is determined that an obstacle exists in the monitoring area (YES in S304), the monitoring unit 7 performs an abnormality notification process on the management unit 5 (S305). Specifically, the abnormality notification process is a process of transmitting a signal indicating that an obstacle exists in the monitoring area even after the closing of the door is completed and is in an abnormal state. An ID unique to the monitoring unit 7 given to the monitoring unit 7 that has performed the object determination process, an image determined to have an obstacle, and information for highlighting the obstacle present in the image (for example, , Information indicating the contour region of the part determined to be an obstacle, rectangular region information for representing a rough moving body position including the contour region, color information for highlighting the obstacle, etc.) Good.

  On the other hand, as a result of the obstacle determination process (S303), when it is determined that no obstacle exists in the monitoring area (S304 NO), the monitoring unit 7 performs a normal notification process to the management unit 5 ( S306). Specifically, the normal notification process is a process of transmitting a signal indicating that no obstacle exists in the monitoring area to the management unit 5, and the obstacle determination process is performed together with the signal. The ID unique to the monitoring unit 7 assigned to the monitoring unit 7 or an image determined to have no obstacle may be transmitted. When there is no obstacle, the normal notification may be omitted.

  That is, according to the obstacle confirmation processing according to the present embodiment, the time interval from when the platform door 11 is closed to when the vehicle starts to move is detected based on the platform door closing completion signal and is used as a reference each time. Since the image is acquired, the reference image and the current image can be appropriately verified, and only the true obstacle can be reliably detected.

  In the above example, the completion of closing is detected based on the closing signal, but the present invention is not limited to this. Therefore, for example, completion of closing may be detected based on an image acquired via the imaging unit 14 or may be detected based on information from other sensors (for example, an infrared sensor).

  Next, details of the obstacle determination process (S303) will be described with reference to FIG. When the obstacle determination process starts, the monitoring unit 7 issues an image capturing request to the image capturing unit 14 (S3031). In response to the imaging request, the imaging unit 14 images the monitoring area and transmits a captured image to the monitoring unit 7. The monitoring unit 7 performs readout processing of the captured image (S3032) until the captured image is acquired (NO in S3033). Thereafter, when a captured image is acquired from the imaging unit (YES in S3033), the image is stored as a current image in a storage unit such as an internal memory (S3034).

  Subsequently, the monitoring unit 7 performs a collation process between the above-described reference image and the current image (S3035). As a result, when there is no difference between the current image and the reference image or the difference is within a predetermined reference (YES in S3037), it is determined that there is no obstacle (S3037). On the other hand, when the difference between the current image and the reference image is greater than or equal to the predetermined reference (NO in S3036), it is determined that an obstacle exists (S3038). If it is determined that an obstacle is present, information for highlighting the obstacle in the image (for example, information indicating the contour area of the part determined to be an obstacle, a rough outline including the contour area) Rectangular area information for representing a moving body position, color information for highlighting obstacles, etc.).

  Here, the collation process between the reference image and the current image may be any known method for specifying the presence or absence of a difference between images or a difference portion, and may be performed by, for example, a background difference method. . That is, according to the background subtraction method, a portion that does not exist in the pre-acquired image, that is, an obstacle is detected from the difference between a reference image (which may be called a background image) that is a pre-acquired image and a current image that is an observation image. Can do.

  With reference to FIG. 10, a specific example when the obstacle matching process is performed using the background difference method will be described. FIGS. 10A, 10B, and 10C are images obtained by imaging the region between the vehicle door and the door plate via the imaging unit. FIG. 10A is an image captured in a time interval from when the vehicle stops on the platform until the platform door starts to open, that is, a reference image. FIG. 10B is an image captured in a time interval from when the platform door is closed until the vehicle starts to move, that is, a current image. In FIG. 10B, the umbrella is sandwiched between the vehicle doors.

  In the collation process, the reference image and the current image are collated by the background difference method (S3036). In this example, since the difference between the two images is the umbrella portion sandwiched between the vehicle doors, the difference amount for the portion is equal to or greater than a predetermined reference (NO in S3036), that is, there is an obstacle in the current image. It is determined to be a thing (S3038). When there is a difference due to the collation, as shown in FIG. 10C, image enhancement information that emphasizes the difference portion is generated for the obstacle portion. In the figure, rectangular frame display information indicating the difference part is generated in the peripheral part of the umbrella (broken line part in the figure). Note that the emphasis information is not limited to this example, and may be, for example, coloring of an obstacle-corresponding portion, emphasis of the outline of the obstacle, blinking, or display of characters near the obstacle.

  In the above description, the obstacle determination process is performed only once after the door is closed. However, the determination process may be performed a plurality of times until the vehicle starts. That is, after receiving the closing signal (S302 YES), the monitoring unit 7 captures a captured image (S3031), reads out (S3032), captures a captured image (S3033 YES), and matches a reference image (S3035). ) May be performed continuously or intermittently to determine obstacles (S3036 to S3038). And based on the said determination process result, you may perform a normal notification process (S306) or an abnormality notification process (S307) continuously or intermittently. At this time, the start timing of the vehicle may be detected using, for example, the above-described vehicle line status signal. Further, the start of the vehicle may be detected based on the image acquired via the imaging unit 14, or may be detected based on information from other sensors (for example, an infrared sensor).

  According to such a continuous or intermittent determination process, obstacles can be continuously detected after the door is closed and before the vehicle starts, so it is more reliable for passenger or operational safety. Can be secured.

<1.4 Display processing on display unit>
Next, display processing for displaying on the display unit 6 based on abnormality information or warning information issued from each monitoring unit 7 will be described.

  As described above, during the precaution processing, the monitoring unit 7 to the management unit 5 is determined to have the warning information, the ID unique to the monitoring unit, and the moving body depending on the presence or absence of the moving body in the monitoring area. , And image enhancement information (hereinafter, warning information, etc.) are transmitted.

  Further, during the obstacle confirmation processing, each monitoring unit 7 to the management unit 5 according to the presence or absence of an obstacle in the monitoring area, abnormality information, an ID unique to the monitoring unit, an image determined to have an obstacle, In addition, image enhancement information (hereinafter, abnormal information or the like) is transmitted.

  When the management unit 5 receives the warning information or the like, the management unit 5 performs display control on the display unit 6 based on the warning information or the like. That is, the management unit 5 causes the display unit 6 to display a moving image during a period in which the platform door in the alert state is closed. In addition, in the case of the said display, you may display the moving image regarding several doors simultaneously, and you may carry out the switch display process by a predetermined input or an automatic process. Further, the management unit 5 transmits the warning information to the general control panel. The general control panel 3 that has received the warning information or the like transmits the warning information or the like to the higher-level device 2.

  On the other hand, when the management unit 5 receives the abnormality information, the management unit 5 performs display control on the display unit 6 based on the abnormality information. FIG. 11 shows a display example of the display unit when an abnormal signal is issued from each monitoring unit due to the presence of an obstacle in the monitoring area. FIG. 11A shows an image (display example 1) captured by the platform door corresponding to the first door of the second car of the vehicle, and FIG. 11B shows the second of the third car of the vehicle. The image (display example 2) imaged with the platform door corresponding to a door is represented.

  FIG. 11A (Display Example 1) is an image captured between the platform door and the vehicle displayed on the display unit 6. In this example, an abnormal state is displayed at the bottom of the image. An abnormal display area for displaying a platform door is divided into three (33a to 33c). The abnormal display areas 33a to 33c are respectively provided with rectangular frame lines 32a to 33c, and numbers indicating the number of vehicles (numbers 1 to 3 in the figure are attached) in the frame lines. And the door numbers of the platform doors surrounded by a rectangular frame are arranged next to each other (in the figure, numbers 1 to 4 are assigned respectively).

  Also, in the figure, among the door numbers of the platform doors enclosed in a rectangular frame, a frame 34 representing the first of the second vehicle and a frame 35 representing the second of the third vehicle are highlighted. Has been. These highlights correspond to the monitoring unit or platform door that detected the abnormal condition.

  Furthermore, in the same figure, an image corresponding to the abnormality information is displayed on the screen center area 31 together with a display area 36 that clearly indicates the first door of the second car at the upper right of the screen. In the image corresponding to the abnormality information, the difference area is highlighted by a rectangular broken line 37.

  On the other hand, FIG. 11B (display example 2) is also an image captured between the platform door and the vehicle displayed on the display unit 6, and in this example, at the bottom of the image The abnormal display area for displaying the platform door in the abnormal state is provided in three (33a to 33c). The abnormal display areas 33a to 33c are respectively provided with rectangular frame lines 32a to 33c, and numbers indicating the number of vehicles (numbers 1 to 3 in the figure are attached) in the frame lines. And the door numbers of the platform doors surrounded by a rectangular frame are arranged next to each other (in the figure, numbers 1 to 4 are assigned respectively). Also, in the figure, among the door numbers of the platform doors enclosed in a rectangular frame, a frame 34 representing the first of the second vehicle and a frame 35 representing the second of the third vehicle are highlighted. Has been. These highlights correspond to the monitoring unit or platform door that detected the abnormal condition.

  Further, in the same figure, an image corresponding to the abnormality information is displayed on the screen center area 31 together with a display area 36 that clearly indicates the second door of the third car in the upper right of the screen. In the image corresponding to the abnormality information, the difference area is highlighted by a rectangular broken line 37.

  The display is configured such that a predetermined input can be manually input to the management unit 5 or an image corresponding to an abnormal state can be switched automatically. Thereby, it can be easily confirmed together with the image which door is abnormal.

<1.5 Door opening process>
Next, a process for reopening the platform door in an abnormal state such as a state where an obstacle is caught will be described.

  As described above, the integrated control panel 3 receives the abnormality information and the like from the management unit 5 together with the monitoring unit ID. Based on the abnormality information and the like, the integrated control panel 3 transmits a door opening command to the individual control panel 4 corresponding to the door in an abnormal state.

  FIG. 12 is a processing flow when the general control panel 3 performs the door opening process. The integrated control panel 3 performs a received signal reading process (S401) until the abnormality information is received (NO in S402). When the abnormality information is received (S402 YES), a door opening signal is transmitted to the individual control device 4 corresponding to the door corresponding to the monitoring unit ID that transmitted the abnormality information (S403). According to such a series of processes, the door opening signal can be transmitted from the general control panel to the door in an abnormal state.

  On the other hand, the individual control panel 4 that has received the door opening signal performs a door opening process of the platform door according to the door opening signal.

  Thereby, the platform door in an abnormal state can be automatically opened.

<2. Second Embodiment>
Next, a second embodiment will be described with reference to FIGS. The second embodiment performs the same operation as the first embodiment when the train is on the line (when the train is stopped on the platform, etc.), but when the vehicle is absent (the train is stopped on the platform, etc.) In the case of not going), it is intended to further improve the safety of passengers and operations.

<2.1 Monitoring process when vehicle is absent>
As described above, the monitoring unit 7 is transmitted with a vehicle status signal indicating whether or not the vehicle 1 is present from the host device 2 via the general control panel 3 and the management unit 5. In the present embodiment, the monitoring unit 7 uses the presence line state signal.

  FIG. 13 is a detailed flow illustrating the monitoring process performed in the monitoring unit 7. The monitoring unit 7 repeats the reception signal reading process (S501) until it receives a vehicle presence line signal (vehicle absence signal) representing the vehicle absence state (NO in S502). When the vehicle absence signal is received (S502 YES), the monitoring unit 7 starts a moving body detection process described later (S503). Thereafter, the reception signal reading process is repeated (S504) until a vehicle presence line signal (vehicle presence line signal) indicating the presence of the vehicle is received again (NO in S505). When the vehicle presence line signal is received (S505 YES), the moving body detection process is terminated (S506). That is, the monitoring device 7 performs the moving object detection process even when the vehicle is absent.

  Next, details of the moving object detection process when the vehicle is absent will be described with reference to FIG. When the moving body detection process in the absence of the vehicle is started, a moving image imaging command is issued to the imaging unit, and the monitoring unit 7 acquires at least two or more images from the moving images (S5031). Here, the two or more images are images that are referred to for difference detection for detecting a moving object. For example, when an image at a certain reference time t is I (t), I (t) And I (t-1), I (t) and I (t + 1), or I (t) and I (t-2) or I (t It may be a combination of t) and I (t + 2). Further, when the moving object detection process or the like is performed with reference to three or more images, for example, a combination of I (t−1), I (t), and I (t + 1) may be used, and I (t -2), I (t) and I (t + 2) may be combined. Needless to say, more images may be used.

  Next, the monitoring unit 7 performs a moving body detection process based on the two or more images (S5032). Various conventional methods can be used as the moving object detection processing method. For example, a technique for detecting a moving object based on an optical flow calculated using a gradient method, a block matching method, or the like, a technique for detecting a moving object based on a background difference, and the like are known. In the moving body detection process (S5032), when there is no moving body in the detection target area (NO in S5033), a normal notification process for notifying the management unit 5 that there is no moving body in the detection target area. This is performed (S5034). Specifically, the normal notification process is a process of transmitting a signal indicating that there is no moving body to the management unit 5, and together with the signal, the monitoring unit 7 that has performed the moving body detection process. An ID unique to the monitoring unit assigned to the image or an image determined to have no moving object may be transmitted. Note that when there is no moving object, the normal notification may be omitted.

  On the other hand, in the moving object detection process (S5032), if a moving object exists in the detection target area (S5033 YES), the management unit 5 has a moving object in the detection target area and is in a warning state. A warning notification process for notifying is performed (S5035). The alert notification process is specifically a process of transmitting a signal indicating that a moving body is present and is in a state to be alerted to the management unit 5, and detecting the moving object together with the signal. The ID unique to the monitoring unit assigned to the monitoring unit 7 that has performed the processing, the image determined to have a moving body, and information for highlighting the moving body portion present in the image (for example, the moving body) Information indicating a contour area of a portion determined to be, rectangular area information for representing a rough moving body position including the contour area, color information for highlighting the moving body, and the like may be transmitted. In the moving object detection process, the above process is continuously performed until the end of the moving object detection process (S506).

  When the management unit 5 receives the warning information or the like, the management unit 5 performs display control on the display unit 6 based on the warning information or the like. FIG. 15 is a display example on the display unit at the time of warning when the vehicle is absent. In this example, the area including at least the track is the monitoring area when the vehicle is absent. The head portion of a person who makes a face from the platform door is detected as a moving body, and is surrounded by a rectangle 40 for emphasis. Note that the highlighting method is not limited to the rectangular method, and other image enhancement methods such as contour enhancement and color addition may be used.

  According to the above configuration, since the monitoring unit is used not only when the vehicle is present but also when the vehicle is absent, the device can be used effectively, and a person who enters the track or approaches a platform door. Etc. can be wary. This can further improve the safety of passengers and navigation.

  In the above example, the absence of the vehicle is detected based on the on-line signal of the vehicle, but the present invention is not limited to this. Therefore, the absence of the vehicle may be detected based on the image acquired via the imaging unit 14, or the absence of the vehicle may be detected by another sensor (such as an infrared sensor).

<3. Third Embodiment>
Next, a third embodiment will be described with reference to FIG. The third embodiment is basically the same as the first embodiment, but the arrangement of the imaging units is different.

  FIG. 16 illustrates an arrangement example of the imaging units according to the present embodiment. In addition, in the same figure, the same code | symbol is attached | subjected about the structure similar to Embodiment 1. FIG. In the figure, the difference from the first embodiment is that storage housings (50a and 50b) are respectively provided on the left and right sides of the track side of the door, and one of the storage housings 50a is the first imaging unit 51a. And the monitoring unit 7 connected to the imaging unit 51a, the other storage case 50b includes a second imaging unit 51b connected to the monitoring unit 7, and further includes a first imaging unit. 51a and the second imaging unit 51b are arranged so as to face each other. In the same configuration, the monitoring unit 7 is configured to be able to alternatively acquire an image from either the first imaging unit 51a or the second imaging unit 51b.

  According to such a configuration, it is possible to obtain a captured image under a more preferable imaging condition (light condition or the like), and to reduce erroneous determination. For example, even when the light irradiation conditions are different between morning and evening, etc., by switching the imaging unit to be used between the first imaging unit and the second imaging unit at a predetermined time, etc. Appropriate imaging is possible.

  According to the present monitoring apparatus, method, program, and system, it is possible to provide a monitoring apparatus, method, program, and monitoring system that are less likely to cause an erroneous determination regarding the presence or absence of an obstacle in the area between the platform door and the vehicle. .

DESCRIPTION OF SYMBOLS 1 Vehicle 2 Host apparatus 3 General control board 4 Individual control board 5 Management unit 6 Display part 7 Monitoring unit 8 Door bag 9 Platform 10 Track 11 Platform door 12 Door board 13 Storage housing 14 Imaging part

Claims (15)

Including an image acquisition control unit that acquires an image captured by the imaging unit, and a state determination unit;
The imaging unit
A predetermined region including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle stopping on the platform is arranged so as to be capable of imaging;
The image acquisition control unit
A first image acquisition control unit that acquires an image obtained through the imaging unit as a first image in a first time interval after the vehicle stops on the platform and until the platform door starts to open; ,
A second image acquisition control unit configured to acquire an image obtained through the imaging unit as a second image in a second time interval until the vehicle starts to move after the platform door is closed; And the state determination unit includes:
Determining the presence or absence of an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image;
A monitoring device characterized by that.   2. The first time interval until the platform door starts to open after the vehicle stops on the platform is detected by receiving a platform door opening start signal. The monitoring device described.   2. The monitoring according to claim 1, wherein the second time period from when the platform door is closed to when the vehicle starts to move is detected by receiving a platform door closing completion signal. apparatus. The image acquisition control unit
A first sequence of acquiring a plurality of images obtained in time series via the imaging unit as a continuous image in a third time interval from when the vehicle door starts to close until the platform door is completely closed. An image acquisition unit,
The state determination unit
Based on the acquired continuous image, the presence or absence of a moving object in a space sandwiched between the platform door and the vehicle door is determined.
The monitoring apparatus according to claim 1.   The start time of the third time interval is detected by receiving a closing start signal of the platform door, and the end time is detected by receiving a closing completion signal of the platform door. The monitoring device described. The state determination unit
Among the continuous images, a predetermined area sandwiched between the platform door and a vehicle door of a vehicle that stops on the platform is used as a monitoring area to determine the presence or absence of a moving object.
The monitoring apparatus according to claim 4. The state determination unit
In the monitoring area, determining the presence or absence of a moving object while eliminating the recognition of the reflection of passengers on the wall surface of the platform door as a moving object,
The monitoring apparatus according to claim 6. The predetermined area that can be imaged by the imaging unit further includes a track area when the vehicle is absent, and the image acquisition control unit includes:
When the vehicle is absent, it further includes a second continuous image acquisition unit that acquires a plurality of images obtained in time series via the imaging unit as a continuous image,
The state determination unit
Based on a continuous image acquired when the vehicle is absent, the presence or absence of a moving object in the predetermined area when the vehicle is absent is determined.
The monitoring apparatus according to claim 1.   The imaging unit includes two imaging units that capture images of the space sandwiched between the platform door and the vehicle door from different directions, and one of the two imaging units is alternatively operable. The monitoring apparatus according to claim 1, wherein: A monitoring device provided on each door of the platform door, an imaging device connected to each monitoring device, a management device that manages the monitoring devices, and a display device provided corresponding to the management device And
The monitoring device includes an image acquisition control unit, a state determination unit, and a transmission unit,
The imaging device
A predetermined region including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle stopping on the platform is arranged so as to be capable of imaging;
The image acquisition control unit
A first image acquisition control unit that acquires an image obtained through the imaging device as a first image in a first time period after the vehicle stops on the platform and until the platform door starts to open; ,
A second image acquisition control unit that acquires an image obtained via the imaging device as a second image in a second time interval until the vehicle starts to move after the platform door is closed; And the state determination unit includes:
Based on the comparison result between the acquired first image and the second image, it is determined whether there is an obstacle in the space sandwiched between the platform door and the vehicle door,
The transmitter is
When it is determined by the determination in the state determination unit that an obstacle exists in the space, an abnormal state signal indicating that an obstacle exists in the space and the second image are transmitted to the management device,
The management device includes a receiving unit and a display control unit,
The receiver is
The abnormal state signal received from each monitoring device and the second image are received in association with each monitoring device,
The display control unit
Information indicating which of the platform doors is in an abnormal state with respect to the display device based on the abnormal state signal received in association with each of the monitoring devices and the second image; Display control to display the second image corresponding to the door,
A monitoring system characterized by that. The monitoring apparatus further includes an image enhancement information generation unit,
The image enhancement information generation unit
Based on the comparison result between the first image and the second image, generating image enhancement information for highlighting a region determined to be an obstacle,
The transmission unit of the monitoring device further includes:
Sending the image enhancement information;
The receiving unit of the management device further includes:
The image enhancement information received from each monitoring device is received in association with each monitoring device,
The display control unit of the management device further includes:
Based on the second image received in association with each monitoring device and the image enhancement information, an image in which an area corresponding to an obstacle is emphasized is displayed on the display device.
The monitoring system according to claim 10. The image acquisition control unit further includes:
A first sequence of acquiring a plurality of images obtained in time series via the imaging device as a continuous image in a third time interval from when the vehicle door starts to close until the platform door is completely closed. An image acquisition unit,
The state determination unit further includes:
Based on the acquired continuous image, it is determined whether there is a moving object in the space sandwiched between the platform door and the vehicle door,
The transmitter further includes:
When it is determined by the determination in the state determination unit that a moving object exists in the space, a warning state signal indicating that a moving object exists in the space and the second image are transmitted to the management device. ,
The receiver further includes
The warning state signal received from each monitoring device and the second image are received in association with each monitoring device,
The display control unit further includes:
Information indicating which of the platform doors is in a warning state with respect to the display device based on the warning state signal received in association with each of the monitoring devices and the second image; Display control to display the second image corresponding to the door,
The monitoring system according to claim 10. The monitoring system further includes:
An integrated control device connected to the management device for managing the opening and closing of the entire platform door;
An individual control device connected to the integrated control device and individually controlling the opening and closing of each door of the platform door;
The management device further includes an abnormality / warning information transmission unit,
The abnormality / warning information transmission unit
When an abnormal state signal is received by the receiving unit, the abnormal state signal is transmitted to the integrated control device together with information on a monitoring device related to the abnormal state signal, and a warning state signal is transmitted by the receiving unit. Is transmitted to the integrated control device together with information on the monitoring device related to the warning status signal,
The comprehensive control device further includes an opening signal transmission unit,
The door opening signal transmitter is
When the abnormal state signal is received, the door opening signal is transmitted to the individual control device that controls the door monitored by the monitoring device related to the abnormal state signal, while the warning state signal is received. Does not send an opening signal to the individual control panel,
The individual control panel is:
In response to the opening signal from the integrated control panel, the door opening of each platform door is controlled.
The monitoring system according to claim 12. A monitoring method based on image data obtained via an imaging unit arranged so as to be capable of imaging a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle parked on the platform Because
Acquiring the image obtained through the imaging unit as a first image in a first time interval after the vehicle stops on the platform until the platform door starts to open; and
Acquiring the image obtained through the imaging unit as a second image in a second time interval until the vehicle starts to move after the platform door is closed; and
Determining whether or not there is an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. A monitoring method characterized by that. A monitoring method based on image data obtained via an imaging unit arranged so as to be capable of imaging a predetermined area including a space sandwiched between a platform door provided on the platform and a vehicle door of a vehicle parked on the platform A computer program for executing
Acquiring the image obtained through the imaging unit as a first image in a first time interval after the vehicle stops on the platform until the platform door starts to open; and
Acquiring the image obtained through the imaging unit as a second image in a second time interval until the vehicle starts to move after the platform door is closed; and
Determining whether or not there is an obstacle in the space sandwiched between the platform door and the vehicle door based on the collation result between the acquired first image and the second image. A computer program characterized by that.

Images