JP2011046516A - Monitoring system in elevator car - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a monitoring system in an elevator car capable of accurately determining the violence of a passenger even if a distance up to the passenger in the car from a camera is changed. <P>SOLUTION: The monitoring system includes the camera 2 arranged in the elevator car 1 and photographing the passenger in the car 1, a mirror for reflecting the passenger from a side surface in the car from a side and arranged so as to reflect an image on the camera 2, an image processing device 3 for detecting the violence state of the passenger from a recording image of the camera, and a portable maintenance device 5 for dividedly registering an image of a distance measuring area by registering an area of reflecting the mirror of the recording image photographed by the camera 2 in the image processing device as the distance measuring area. The image processing device determines a distance between the camera 2 and the passenger from an image of the passenger detected from the recording image of the divided distance measuring area, and sets a violence determining threshold suitable for a distance between the camera and the passenger, and detects the abnormal action of the passenger by comparing a variation quantity in the motion of the passenger in the car with the violence determining threshold. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an elevator car monitoring system, and more particularly to an elevator car monitoring system that detects abnormal behavior of passengers in an elevator car.

  As a conventional technique related to an elevator car monitoring system, for example, a technique described in Patent Document 1 is known. This prior art calculates the direction and magnitude of movement of each point in an image from two images taken by a camera installed in the car at different shooting times by a predetermined time, and the direction and magnitude of movement of each point. Then, the motion calculation means for calculating the amount of variation in the movement of at least one of the person's movement direction and magnitude is compared with the amount of variation in the person's movement and the person's rampage determination threshold. Is to determine.

JP 2006-276969 A

  Since the above-described conventional elevator car monitoring system according to the related art monitors the inside of the car with a single camera, the distance from the camera to the person cannot be grasped. Normally, the amount of variation in movement of a person close to the camera is large and the amount of movement of a person far from the camera is small. Since the distance cannot be grasped correctly, it is not possible to set a rampage judgment threshold suitable for the distance from the camera and compare it with that threshold, making it impossible to make a precise rampage judgment. Have.

  The object of the present invention is to solve the above-mentioned problems of the prior art, and even if the distance from the camera installed in the elevator car to the person changes, it is possible to accurately determine the violence of the person who is a passenger. An object of the present invention is to provide an elevator car monitoring system.

  According to the present invention, the object is to detect the movement of a passenger in the car, and to detect the abnormal behavior of the passenger in the elevator car by comparing the variation amount of the movement with the passenger's rampage determination threshold. In the monitoring system, a camera installed in the elevator car and photographing a passenger in the car, a mirror that is photographed from the side of the passenger from the side of the car, and the image is reflected in the camera, and An image processing device for detecting a passenger rampage from a recorded image of the camera, means for registering an area where the mirror of the recorded image taken by the camera is reflected in the image processing device as a distance measurement area, and the distance The distance between the camera and the passenger from the means for dividing and registering the measurement area image and the passenger image detected from the divided recorded image of the distance measurement area. Determined is accomplished by setting the rage determination threshold suitable to the distance between the camera and the passenger.

  According to the present invention, since a rampage determination threshold suitable for the distance between the camera and the passenger is set to detect the abnormal behavior of the passenger, even if the passenger in the car is close to the camera, the far position Even in this case, the rampage determination threshold value can be changed to an appropriate value, so that the rampage detection with high accuracy can be performed.

It is a block diagram which shows the structure of the elevator car monitoring system by one Embodiment of this invention. It is the figure which looked at the state which installed the mirror for distance measurement in the cage from the back of the cage. It is the figure which looked at the state which installed the distance measuring mirror and the security camera in the cage | basket | car from the side of a cage | basket | car. It is a figure which shows an example of the state in the cage | basket | car imaged with the security camera, and demonstrates the setting of a distance measurement area. It is a figure which shows the state which divided | segmented the distance measurement area according to the distance from a camera. It is a flowchart which sets a rampage detection determination threshold value from the image of a distance measurement area. It is a flowchart explaining the processing operation of the image processing apparatus which comprises an elevator car monitoring apparatus. It is a block diagram which shows the structure of the elevator car monitoring apparatus comprised in an image processing apparatus.

  Embodiments of an elevator car monitoring system according to the present invention will be described below in detail with reference to the drawings.

  FIG. 1 is a block diagram showing the configuration of an elevator car monitoring system according to an embodiment of the present invention.

  As shown in FIG. 1, the elevator car monitoring system according to the embodiment of the present invention is connected to an elevator car 1, a security camera 2 that is installed in the car 1 and photographs passengers in the car, and a security camera 2. An image processing device 3 that monitors the movement of passengers in the car 1, a recorder connected to the image processing device 3, an image recording device 4 that records images taken by the security camera 2, and an image processing device 3. A portable maintenance device 5 for setting a range of a distance measurement area for an image taken by the security camera 2, a maintenance monitoring device 6 connected to the image processing device 3, an elevator control panel 7, and a maintenance monitoring device 6 and a control center 8 connected via a communication line.

  In the above description, the image processing device 3 detects a passenger rampage from a recorded image taken by the security camera 2, and an elevator car monitoring device is configured therein.

  FIG. 8 is a block diagram showing the configuration of the elevator car monitoring device configured in the image processing device 3.

  As shown in FIG. 8, the elevator car monitoring device receives an image captured by the security camera 2 and calculates a motion of the passenger 31 and a variation in the motion of the passenger calculated by the motion calculator 31. A correction unit 32 that corrects the amount, a door open / close determination unit 33 that is a door open / close determination unit that determines the open / closed state of the door that communicates with the inside and outside of the car 1 that is a closed space, and the correction unit 32 determines the amount of variation in passenger movement. A statistical model database (DB) 34 storing a statistical model used for correction and a correction result of the variation amount of the passenger movement from the correction unit 32 are received, and the variation amount of the passenger movement and the passenger rampage determination threshold value are received. And a ramp determination unit 35 that is a ramp determination unit that determines passenger rampage.

  In the above description, the image in the car photographed by the security camera 2 is output to the motion calculation unit 31. The motion calculation unit 31 calculates the direction and magnitude of the movement of each point in the image from two images taken by the security camera 2 with different shooting times for a predetermined time, and based on the direction and magnitude of the movement of each point. A variation amount of the movement of at least one of the direction or magnitude of the movement of the passenger is calculated. The two images whose photographing times are different by a predetermined time may be two images before and after consecutive photographing. For example, in the case of an image photographed by the NTSC system using a video camera, photographing is performed every 1/30 seconds. In this case, the two images that are different from each other for a predetermined time can be two consecutive images. However, by thinning out the middle image or changing the shooting interval, Or every second. The motion calculation unit 31 includes a storage area and can appropriately store the received image. The variation amount of the passenger motion calculated by the motion calculation unit 31 is output to the correction unit 32.

  The inside of the car 1 is a closed space, and the door open / close determining unit 33 is a door open / close determining unit that determines an open / closed state of the door that communicates with the inside and outside of the car 1 in the closed space. The door open / close determination unit 33 determines the door open / close state based on a door close signal representing the door close state from the elevator control panel 7 and outputs the result to the motion calculation unit 31 and the correction unit 32. The motion calculation unit 31 calculates the amount of variation in passenger movement, excluding the passenger entry / exit data into the car 1, based on the determination result of the door opening / closing determination unit 33. The correction unit 32 is based on a statistical model composed of a database of average values and variance values of passenger movements classified for each door open / closed state, and the passenger movement variations or passengers used in the rampage determination unit 35 It is a correction means which corrects the rampage determination threshold. Then, the correction unit 32 corrects the amount of variation in passenger movement using the statistical model stored in the statistical model DB 34 and outputs the result to the rampage determination unit 35. The rampage determination unit 35 is rampage determination means for comparing the amount of variation in passenger movement with the passenger rampage determination threshold value to determine passenger rampage. The rampage determination unit 35 performs threshold processing on the corrected variation in the movement of the passenger to determine whether the passenger is rampant.

  FIG. 7 is a flowchart for explaining the processing operation of the image processing device 3 constituting the elevator car monitoring device, which will be described next.

(1) When an image is captured by the security camera 2 and the image is input, the door open / close determining unit 33 determines the open / closed state of the door based on a door close signal from the elevator control panel 7 indicating the door closed state (step) S10, S11).

(2) Next, the motion calculation unit 31 calculates an optical flow from two images whose shooting times are different by a predetermined time. This optical flow is an apparent velocity distribution representing the direction of movement and the magnitude of movement at each point of the image. In this way, the embodiment of the present invention pays attention to the amount of variation in passenger motion that correlates with the rampage operation, and determines that there is rampage when the variation is large. Is not erroneously determined, and stable rampage detection can be realized (step S12).

(3) Thereafter, the motion calculation unit 31 calculates the amount of variation in passenger movement from the optical flow and passes it to the correction unit 32. The variation amount of the movement of the passenger is, for example, a variation in the direction of the movement of the passenger or a variation in the magnitude of the movement of the passenger. The movement direction obtained at each point of the image and the variance value of the movement size. Or it calculates | requires as a standard deviation (step S13).

(4) Next, the correction unit 32 uses the statistical model stored in the statistical model database 34 to correct the amount of variation in passenger movement received from the motion calculation unit 31, and install the security camera 2. Modify the value to reduce the dependency on the environment. A statistical model is prepared in advance for each monitoring area (inside the car) and each door open / closed state. For example, after the installation of the security camera 2, the amount of variation in passenger movement is calculated on a trial basis with respect to normal passenger movement, and the temporal average value M and variance value σ are stored as a statistical model. When the variation amount of the passenger movement calculated in the process of step S13 is V, the correction value V ′ is
V ′ = (VM) ÷ σ (1)
(Step S14).

(5) Thereafter, the rampage determination unit 35 performs threshold processing on the correction value V ′ and determines whether there is a passenger rampage. In the embodiment of the present invention, the ramp determination unit 35 determines that there is rampage when the correction value V ′ is greater than or equal to a threshold value (step S15).

  The processes in steps S10 to S15 described above are repeatedly executed at predetermined time intervals, and the rampage of passengers in the car is constantly monitored.

  Next, a description will be given of means for changing the determination threshold for detecting the ramp to an appropriate one according to the distance between the passenger and the camera.

  2 is a view of the distance measuring mirror 9 installed in the car as viewed from the rear of the car. FIG. 3 is a view of the distance measuring mirror 9 and the security camera 2 installed in the car as viewed from the side of the car. FIG.

  Usually, as shown in FIG. 3, the security camera 2 is installed on the right side of the upper rear portion of the car so that the entire car can be photographed. As shown in FIG. 2, the distance measuring mirror 9 is attached to the upper side of the car side opposite to the security camera 2 so that the mirror surface is obliquely downward so that passengers can be seen. In the embodiment of the present invention, the distance measuring mirror 9 is attached to the side plate of the car so that the mirror surface has a rectangular shape and the longitudinal direction of the mirror surface is substantially parallel to the depth direction of the car. One long side of the mirror surface is substantially in contact with the car side plate surface, and the other long side is separated from the car side plate surface. The angle between the short side of the mirror surface, that is, the direction perpendicular to the longitudinal direction, and the car floor surface is set to be larger than 0 degree and smaller than 90 degrees so that the mirror surface faces obliquely downward. Since the distance measuring mirror 9 is attached in this way, the position at which the person appears in the distance measuring mirror 9 also changes depending on the positions before and after the passenger stands in the car.

  FIG. 4 is a diagram illustrating an example of the state in the car taken by the security camera 2 and illustrating the setting of the distance measurement area. FIG. 5 is a diagram illustrating a state in which the distance measurement area is divided according to the distance from the camera. .

  The distance measuring mirror 9 is installed so as to appear in an image taken by the security camera 2. Then, in the image taken by the security camera 2, as shown in FIG. 4, the passenger 11 and the passenger 12 shown in the distance measuring mirror 9 are shown. This image is sent to the portable maintenance device 5 via the image processing device 3. A maintenance person possessing the portable maintenance device 5 selects a portion of the image taken by the security camera 2 in which the distance measuring mirror 9 is reflected from the portable maintenance device 5, and uses that portion as the distance measurement area 10 for the image processing device 3. Register with. At this time, the maintenance person possessing the portable maintenance device 5 divides and registers the registered distance measurement area 10 into a plurality of areas according to the distance from the security camera 2, as shown in FIG.

  FIG. 6 is a flowchart for setting the rampage detection determination threshold from the image of the distance measurement area, which will be described next. The processing here is processing performed by the correction unit of the elevator car monitoring device configured in the image processing device 3.

(1) First, when the optical flow in the distance measurement area 10 is detected, it is determined whether or not the detected optical flow is detected in the area 1 of the distance measurement area 10, and the optical flow is detected in the area 1 Then, the rampage determination threshold is changed to the setting for short distance, and the process here is terminated (steps S1 to S3).

(2) If the optical flow is not detected in the area 1 in the determination in step S2, it is determined whether or not the detected optical flow is detected in the area 2 of the distance measurement area 10, and the optical flow is in the area 2 Is detected, the rampage determination threshold is changed to the setting for medium distance, and the process here is terminated (steps S4 and S5).

(3) If the optical flow is not detected in the area 2 in the determination in step S4, it is determined whether or not the detected optical flow is detected in the area 3 of the distance measurement area 10, and the optical flow is in the area 3 If it has been detected, the rampage determination threshold is changed to the setting for long distance, and the process here is terminated (steps S6 and S7).

(4) If the optical flow is not detected in the area 3 in the determination in step S6, the optical flow is not detected anywhere in the areas 1 to 3 of the distance measurement area 10. The standard rampage determination threshold is set, and the process here is terminated (step S8).

  In the embodiment of the present invention, by performing the processing as described above, it is possible to set a rampage detection determination threshold suitable for the distance between the passenger in the car and the security camera 2, thereby improving the rampage detection accuracy. Can do.

  Normally, even if the person who is a passenger in the car is moving in the same manner, the amount of variation in the movement of the person drawn from the image becomes smaller as the person moves away from the security camera 2, so that the violence determination is performed in the above-described processing. The threshold value is set so as to decrease in the order of short distance, medium distance, and long distance.

DESCRIPTION OF SYMBOLS 1 Car 2 Security camera 3 Image processing apparatus 4 Image recording apparatus 5 Portable maintenance apparatus 6 Maintenance monitoring apparatus 7 Elevator control panel 8 Control center 9 Distance measuring mirror 10 Distance measuring area 11 Passenger 12 Passenger reflected in distance measuring mirror 31 Motion calculation unit 32 Correction unit 33 Door open / close determination unit 34 Statistical model database (DB)
35 Rampage judgment part

Claims (2)

In the elevator car monitoring system that detects the movement of passengers in the car and detects the abnormal behavior of the passengers in the elevator car by comparing the fluctuation amount of the movement and the passenger rampage threshold,
A camera that is installed in the elevator car and photographs passengers in the car, a mirror that captures the passengers from the side of the car, and a picture that is captured by the camera, and a recorded image of the camera An image processing device for detecting a passenger rampage, a means for registering in the image processing device as an area for distance measurement the area where the mirror of the recorded image taken by the camera is shown, and the area for distance measurement A means for dividing and registering an image, and determining a distance between the camera and the passenger from a passenger image detected from the recorded image of the divided area for distance measurement, and a rampage determination threshold suitable for the distance between the camera and the passenger An elevator car monitoring system characterized by setting.   2. The elevator car monitoring apparatus according to claim 1, wherein the mirror is installed on an upper part of the car side plate so that a mirror surface faces obliquely downward.