JP2010195537A - Monitoring device in car of elevator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To correctly detect the existence or nonexistence of a suspicious object without erroneously detecting a poster or the like in a car by automatically setting a detection area when detecting a suspicious object by using an image photographed by a monitoring camera. <P>SOLUTION: A monitoring camera 21 is installed in the car 11. The camera 21 is adjusted to photograph an image of the inside of the cage in accordance with the height of the car door 12. A detection area determination part 24 analyzes the image photographed by the camera 12 to determine a detection area of a suspicious object using an upper end or a lower end of the photographed image as a reference. An image difference detection part 25 compares a real-time image which is photographed during the operation of the elevator with a basic image stored in a basic image storage part 23 to detect the existence or nonexistence of a suspicious object from a difference between the images based on the detection area determined by the detection area determination part 24. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an elevator car monitoring apparatus for monitoring the inside of a car using a monitoring camera.

  In recent years, social awareness of criminal acts targeting unspecified majority such as terrorism has increased, and surveillance cameras are used to detect suspicious objects that cannot be identified by suspicious persons or owners.

  In particular, in an elevator, when the door of a car is closed, it becomes a closed space that is difficult for others to reach, so there is a high risk of leaving dangerous goods or violent acts. Therefore, a system has been developed in which a surveillance camera is installed in the car and a suspicious object is automatically detected from an image taken by the surveillance camera.

  Such a system is called a “rage / stagnation detection system”. The stay detection function (referred to as stay detection), which is one of the functions, detects that the same object has stayed in the car for a certain period of time and issues a report. Further, for example, when a user falls in the car or stays in the car for some reason, the user can quickly find and deal with it.

  Here, as a technique of stay detection, as disclosed in Patent Document 1, Patent Document 2, and the like, a method of detecting a state in which the same object stays in a preset region for a certain period of time. Is common.

  In addition, a method (recorded as a difference detection method) that records in advance an image taken inside the car in an unattended state as a basic image and digitizes the difference from the image taken during driving is also common. . In this case, the value of the difference is greater than or equal to a certain value, and it is determined that a stay is present when the difference continues for a certain time.

Japanese Patent Publication No. 5-27920 JP 2003-324726 A

  In Patent Documents 1 and 2 described above, if a paper or a poster is pasted on the side plate of the car, even if it is permitted, it may be erroneously detected as a suspicious object. In this case, it is sufficient to set the place where the stickers are pasted as outside the detection area, but it is very troublesome to perform such area setting for each property.

  In addition, even with the above difference detection method, if a sticker is pasted in the car, the difference value from the basic image becomes large, so that there is a possibility of erroneous detection that there is a suspicious object. In order to avoid this, it is only necessary to record the basic image of the car again after pasting a sticker or a poster, but there is a problem that rewriting the basic image each time is very troublesome.

  Therefore, an object of the present invention is to automatically set a detection area when detecting a suspicious object using an image taken by a surveillance camera, and a sticker sheet intentionally brought into a car. An object of the present invention is to provide an elevator car monitoring device that can accurately detect the presence or absence of a suspicious object without erroneously detecting the kind.

  An elevator car monitoring device according to the present invention is installed in a car, and a monitoring camera for photographing a car interior in accordance with the height of a car door provided at the entrance of the car. Basic image storage means for storing a basic image of the car photographed by the camera and analysis of the photographed image of the camera, and a detection area for the suspicious object is determined with reference to the upper end or the lower end of the photographed image. Based on the detection area determination means and the detection area determined by the detection area determination means, images taken at regular intervals by the imaging means during operation of the elevator and the basic information stored in the basic image storage means An image difference detecting means for comparing the images and detecting the presence / absence of a suspicious object from the difference between the images is provided.

  According to the present invention, when a suspicious object is detected using an image taken by a surveillance camera, the detection area can be automatically set. This makes it possible to accurately detect the presence or absence of a suspicious object without erroneously detecting stickers that are intentionally brought into the car.

FIG. 1 is a block diagram showing the configuration of the elevator car monitoring apparatus according to the first embodiment of the present invention. FIG. 2 is a schematic diagram illustrating a method for detecting a difference between a basic image and a real-time image according to the embodiment. FIG. 2A illustrates a basic image, and FIG. 2B illustrates a real-time image. FIG. 3 is a schematic diagram for explaining a detection region determination method in the embodiment. FIG. 4 is a schematic diagram when the basic image includes a ceiling surface in the embodiment. FIG. 5 is a schematic diagram for explaining a detection region determination method according to the second embodiment of the present invention. FIG. 6 is a schematic diagram for explaining a detection region determination method according to the third embodiment of the present invention. FIG. 7 is a schematic diagram for explaining a detection region determination method according to the fourth embodiment of the present invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram showing the configuration of the elevator car monitoring apparatus according to the first embodiment of the present invention. In the figure, 11 is an elevator car. The car 11 moves up and down in the hoistway through a rope by driving a hoisting machine (not shown).

  A car door 12 is attached to an entrance 11a of the car 11 so as to be freely opened and closed. The opening / closing operation of the car door 12 is controlled by a door control unit 13 installed at the upper part of the car 11.

  The door control unit 13 includes a door motor (not shown), its control circuit, a limit switch for detecting the open / closed state of the car door 12, and the like. The door control unit 13 opens and closes the car door 12 in response to an open / close control signal from the elevator operation control unit 14, and notifies the elevator operation control unit 14 of the open / closed state of the car door 12.

  The elevator operation control unit 14 performs elevator operation control, and includes a computer including a CPU, a ROM, a RAM, and the like. The elevator operation control unit 14 is installed in a machine room of a building.

  Here, a monitoring camera 21 is installed in the car 11. The camera 21 is installed on the surface 11 b facing the car door 12 with the photographing lens 21 a facing the car door 12 side. The reason for installing the camera 21 at such a position is to detect the characteristic shape of the car door 12 from the image of the car room photographed by the camera 21 and use this to determine the stay detection area. .

  Therefore, FIG. 1 shows an example in which the camera 21 is installed on the surface 11b facing the car door 12, but the ceiling surface 11c, side surface 11d, The camera 21 may be installed in 11e (see FIG. 3) or the like.

  During the operation of the elevator, the camera 21 sequentially transfers images taken inside the car 11 to the image processing device 22 at regular time intervals. The image processing device 22 includes a basic image storage unit 23, a detection area determination unit 24, and an image difference detection unit 25. In the example of FIG. 1, the elevator operation control unit 14 and the image processing device 22 are provided independently. However, the configuration may be incorporated in one control device.

  The basic image storage unit 23 stores a basic image of the car 11 photographed by the camera 21. Since this basic image is used as a determination criterion for stay detection, it is desirable that there is no user or suspicious object in the car 11. Therefore, usually, after confirming that there are no users or suspicious objects in the car 11, the most suitable basic image is selected from a plurality of images obtained by the shooting. It is desirable to store the image in the basic image storage unit 23 as an image.

  The detection area determination unit 24 analyzes an image captured by the camera 21 and determines an area for detecting a suspicious object in the captured image. The detection region determination method by the detection region determination unit 24 will be described in detail later.

  Based on the detection area determined by the detection area determination unit 24, the image difference detection unit 25 is an image (hereinafter referred to as a real-time image) and a basic image captured by the camera 21 during the operation of the elevator. A difference from the basic image stored in the storage unit 23 is detected.

  A difference detection method by the image difference detection unit 25 will be described with reference to FIG.

  2A and 2B are schematic diagrams showing a method for detecting a difference between a basic image and a real-time image. FIG. 2A shows a basic image, and FIG. 2B shows a real-time image.

  The basic image 30 is an image obtained by photographing the inside of the car 11 with the camera 21 in an unattended state, and is stored in advance in the basic image storage unit 23. On the other hand, the real-time image 31 is an image taken inside the car 11 by the camera 21 during the elevator operation. The image difference detection unit 25 extracts the images in the detection region 32 determined by the detection region determination unit 24 from the basic image 30 and the real-time image 31, and quantifies the difference between these images. In this case, if the area of the detection region 32 on the image is E1, and the area of the difference between the basic image 30 and the real-time image 31 is E2, E2 / E1 is obtained as the difference value.

  Here, when the state where the difference value is equal to or greater than a certain value continues for a certain time or longer, the image difference detecting unit 25 determines that a suspicious object (or suspicious person) is present in the car 11, and indicates that. The suspicious object detection signal shown is transmitted to the elevator operation control unit 14. When the suspicious object detection signal is received, the elevator operation control unit 14 stops the car 11 at the nearest floor, and then outputs a door door opening signal to the door control unit 13 to open the car door 12.

  Note that an alarm may be issued only when a suspicious object is detected. For this alarm, for example, a buzzer installed in a hall or a manager's room is sounded, a rotating light is turned on, or a communication line such as the Internet is used to send to an external elevator maintenance company or security company. Including reporting.

  Next, a detection area determination method by the detection area determination unit 24 will be described.

  FIG. 3 is a schematic diagram for explaining a detection region determination method according to the first embodiment.

  Here, the case where the detection area 32 is set for the basic image 30 will be described as an example. However, the detection area 32 needs to be set for the real-time image 31 in the same manner. However, in the real-time image 31, there is a possibility that a horizontal straight line, which will be described later, may be erroneously detected when a long object such as a heel pattern is placed sideways, for example, so the detection region 32 using the basic image 30 is detected. And coordinates corresponding to the detection region 32 are preferably applied to the real-time image 31.

  In FIG. 3, the car door 12 is a door called a two-door central opening type, and has a structure that opens and closes the entrance 11a by moving the two door panels 12a and 12b to the left and right in conjunction with each other. In general, a curtain plate 12 c is provided at the upper end of the car door 12. The curtain plate 12c is for hiding mechanism parts such as a door motor for driving the car door 12, a control device for the motor, and a limit switch for detecting the open / closed state of the car door 12.

  Now, the basic image 30 includes a curtain 12c provided at the upper end of the car door 12 in accordance with the height of the car door 12, a part of both side surfaces 11d and 11e in the car room, and a part of the floor surface 11f. It is assumed that the image is taken in such a manner.

  The detection area determination unit 24 recognizes the upper end 30U of the basic image 30 as the vicinity of the upper end of the car door 12, that is, the boundary with the curtain plate 12c, and a position lowered from the vertical direction by a predetermined distance LA. A virtual straight line 30LA is drawn in the horizontal direction.

  As shown in FIG. 4, when the basic image 30 includes a part of the ceiling surface 11 c in the car room, the basic image 30 has a predetermined length that is detected second from the top. Let the straight line be the upper end 30U. However, if a part of the ceiling surface 11c is included in the basic image 30, it is difficult to understand the boundary line with the car door 12, so that the car door does not include the ceiling surface 11c as in the example of FIG. The installation position of the camera 21 is preferably adjusted in advance according to the height of 12.

  Next, the detection area determination unit 24 recognizes the lower end 30D of the basic image 30 as the vicinity of the floor surface 11f in the car room, and surrounds the lower end 30D, the right end 30R, the left end 30L, and the virtual straight line 30LA. The determined area (shaded area) is determined as the suspicious object detection area 32.

  Here, the distance LA depends on how the actual car 11 is seen when a distance corresponding to about 1 to 1.5 m from the ceiling surface 11 c in the car room is projected on the captured image of the camera 21. It is determined. This is because a standard standard elevator has a cab height of just over 2 m, and stickers and posters are usually stuck at an average height of about 1 to 1.5 m from the floor 11 f. It is. Therefore, although the distance LA is set in advance, it actually differs depending on the height of the cab or the installation position of the camera 21.

  As described above, the image difference detection unit 25 extracts the images in the detection region 32 determined by the detection region determination unit 24 from the basic image 30 and the real-time image 31 and compares them, thereby determining the presence or absence of a suspicious object. To detect. In this case, even if a sticker or poster brought in by the elevator administrator or the like is put in the car 11, the image comparison can be performed in the detection area 32 except for these, thereby reducing the possibility of erroneous detection. can do.

  In this way, even if the operator does not set the area, the detection area 32 for the suspicious object is appropriately set so as to remove the stickers pasted in the car 11, and left in the car 11 from the captured image. The detected suspicious object or the suspicious person staying in the car 11 can be accurately detected.

(Second Embodiment)
Next, another detection region determination method by the detection region determination unit 24 will be described as a second embodiment of the present invention.

  In the first embodiment, the detection area of the suspicious object is determined based on the upper end portion of the captured image. In the second embodiment, the detection area of the suspicious object is determined based on the lower end portion of the captured image. To do.

  FIG. 5 is a schematic diagram for explaining a detection region determination method according to the second embodiment of the present invention. Here, a case where the detection area 32 is set for the basic image 30 will be described as an example. As described above, the coordinates of the detection area 32 set in the basic image 30 are applied to the real-time image 31.

  Now, the basic image 30 includes a curtain 12c provided at the upper end of the car door 12 in accordance with the height of the car door 12, a part of both side surfaces 11d and 11e in the car room, and a part of the floor surface 11f. It is assumed that the image is taken in such a manner.

  The detection area determination unit 24 recognizes the lower end portion 30D of the basic image 30 as the vicinity of the floor surface 11f in the car room, and virtually moves in a horizontal direction to a position that is raised in the vertical direction by a predetermined distance LB. Draw a straight line 30LB.

  Next, the detection region determination unit 24 detects the right end portion 30R and the left end portion 30L of the basic image 30, and a region (hatched line) surrounded by the end portions 30R, 30L, the lower end portion 30D, and the virtual straight line 30LB. (Part) is determined as the detection area 32 of the suspicious object.

  Here, the distance LB is determined in accordance with how the actual car 11 looks when a distance corresponding to a value obtained by adding about 60 cm to the depth dimension of the car room is projected on the captured image of the camera 21. The This is because, in consideration of the effect that the depth of the cab in the image looks shorter than the actual depth due to the perspective difference, an average height of about 50 to 100 cm from the floor surface 11f is determined as the detection region. is there. Accordingly, the distance LB is set in advance, but actually takes different values depending on the depth of the cab or the installation position of the camera 21.

  As described above, the image difference detection unit 25 extracts the images in the detection region 32 determined by the detection region determination unit 24 from the basic image 30 and the real-time image 31 and compares them, thereby determining the presence or absence of a suspicious object. To detect. In this case, even if a sticker or poster brought in by the elevator administrator or the like is put in the car 11, the image comparison can be performed in the detection area 32 except for these, thereby reducing the possibility of erroneous detection. can do.

  As described above, even when the detection area 32 is determined based on the lower end portion of the photographed image, the suspicious object is detected so as to remove the stickers attached to the car 11 as in the first embodiment. By appropriately setting the area 32, it is possible to accurately detect a suspicious object left in the car 11 or a suspicious person staying in the car 11 from the captured image.

(Third embodiment)
Next, another detection region determination method by the detection region determination unit 24 will be described as a third embodiment of the present invention.

  In the first embodiment, the detection area of the suspicious object is determined based on the upper end portion of the photographed image. In the third embodiment, the suspicious object is determined based on the upper end portion of the car door on the photographed image. The detection area is determined.

  FIG. 6 is a schematic diagram for explaining a detection region determination method according to the third embodiment of the present invention. Here, a case where the detection area 32 is set for the basic image 30 will be described as an example. As described above, the coordinates of the detection area 32 set in the basic image 30 are applied to the real-time image 31.

  Now, the basic image 30 includes a curtain 12c provided at the upper end of the car door 12 in accordance with the height of the car door 12, a part of both side surfaces 11d and 11e in the car room, and a part of the floor surface 11f. It is assumed that the image is taken in such a manner.

  The detection area determination unit 24 recognizes the upper end portion 30U of the basic image 30 as the vicinity of the upper end portion of the car door 12, that is, the boundary portion with the curtain plate 12c, and the center point of the upper end portion 30U, that is, the left and right ends of the basic image 30. A point P1 that is equidistant from 30L and 30R is detected.

  Next, the detection area determination unit 24 searches in the vertical direction from the point P1, and detects the intersection P2 with the straight line 30MA detected first.

  Next, the detection area determination unit 24 searches the straight line 30MA from the intersection point P2 in the horizontal direction, and detects a straight line 30MB closest to the intersection point P2 among the straight lines extending downward from the straight line 30MA. The intersection point P3 between the straight line 30MB and the straight line 30MA is detected.

  Next, the detection region determination unit 24 draws a virtual straight line 30LC in the horizontal direction at a position that is lower than the intersection P3 by a predetermined distance LC. As described with reference to FIG. 4, when the basic image 30 includes a part of the ceiling surface 11 c in the car room, it is detected second from the top including the upper end portion of the basic image 30. A straight line having a predetermined length is defined as the upper end 30U.

  Next, the detection area determination unit 24 recognizes the lower end 30D of the basic image 30 as the vicinity of the floor surface 11f in the car room, and surrounds the lower end 30D, the right end 30R, the left end 30L, and the virtual straight line 30LC. The determined area (shaded area) is determined as the suspicious object detection area 32.

  The distance LC is determined according to how the actual car 11 is viewed when a distance corresponding to a value obtained by subtracting about 100 cm from the height of the car door 12 is projected on the captured image of the camera 21. This is because, as already mentioned, the height of a typical car door is a little over 2 meters, and stickers and posters are usually stuck at a height of about 1 to 1.5 meters from the floor 11f. It is. That is, the distance LC is set in advance, but actually takes a different numerical value depending on the size of the car door 12 and the installation position of the camera.

  Here, a description will be given of which part in the actual passenger car cabin the straight lines 30MA and 30MB described above correspond to.

  In FIG. 6, the car door 12 is a two-door central opening type and has a structure that opens and closes the entrance 11a by moving the two door panels 12a and 12b to the left and right in conjunction with each other. A curtain plate 12 c is provided at the upper end of the car door 12. Accordingly, the straight line 30MA first detected in the vertical direction from the center point P1 of the upper end portion 30U of the basic image 30 is a boundary line between the curtain 12c and the car door 12 in the actual car room, and is the intersection. It can be seen that P2 is one point on the boundary line.

  Further, the straight line 30MB from the intersection point P2 is searched in the horizontal direction, and the straight line 30MB closest to the intersection point P2 among the straight lines extending downward from the straight line 30MA is two sheets constituting the car door 12. This is a boundary line between the door panels 12a and 12b.

  Here, the reason for searching on the straight line 30MA in the horizontal direction may be that the upper end center point P1 of the basic image 30 that is a captured image is biased to either the left or right depending on the mounting position or mounting angle of the camera 21. Because. In this way, the detection area 32 can be determined by detecting a T-shaped figure formed by the curtain 12c and the car door 12.

  As described above, the image difference detection unit 25 extracts the images in the detection region 32 determined by the detection region determination unit 24 from the basic image 30 and the real-time image 31 and compares them, thereby determining the presence or absence of a suspicious object. To detect. In this case, even if a sticker or poster brought in by the elevator administrator or the like is put in the car 11, the image comparison can be performed in the detection area 32 except for these, thereby reducing the possibility of erroneous detection. can do.

  In FIG. 6, the description has been given by taking the two-door central opening type car door 12 as an example. However, even in the case of the two-door single-opening type in which the two door panels 12 a and 12 b move in the same direction, the same as above. The boundary line between the two door panels 12a and 12b can be detected by the method.

  In this manner, the detection of the upper end portion of the car door 12 in the captured image and the determination of the detection region 32 with reference to the upper end portion of the car door 12 can be performed as in the first embodiment. The suspicious object detection area 32 is appropriately set so as to remove the stickers attached to the car 11 and the suspicious object left in the car 11 from the photographed image or the suspicious object staying in the car 11. The person can be accurately detected.

(Fourth embodiment)
Next, another detection region determination method by the detection region determination unit 24 will be described as a fourth embodiment of the present invention.

  In the third embodiment, the detection area of the suspicious object is determined based on the upper end portion of the car door on the photographed image. However, in the fourth embodiment, the lower end portion of the car door on the photographed image is used as a reference. Thus, the detection area of the suspicious object is determined.

  FIG. 7 is a schematic diagram for explaining a detection region determination method according to the fourth embodiment of the present invention. Here, a case where the detection area 32 is set for the basic image 30 will be described as an example. As described above, the coordinates of the detection area 32 set in the basic image 30 are applied to the real-time image 31.

  Now, the basic image 30 includes a curtain 12c provided at the upper end of the car door 12 in accordance with the height of the car door 12, a part of both side surfaces 11d and 11e in the car room, and a part of the floor surface 11f. It is assumed that the image is taken in such a manner.

  The detection area determination unit 24 recognizes the lower end 30D of the basic image 30 as the vicinity of the floor surface 11f in the car room, and the center point of the straight line constituting the lower end 30D, that is, the left and right ends 30L and 30R of the basic image 30. The point P4 which is equidistant from is detected.

  Next, the detection area determination unit 24 searches in the vertical direction from the intersection point P4 and detects the intersection point P5 with the straight line 30MC detected first.

  Next, the detection area determination unit 24 searches the straight line 30MC from the intersection point P5 in the horizontal direction, and detects the straight line 30MD closest to the intersection point P5 among the straight lines extending upward from the straight line 30MC. . The detection area determination unit 24 detects an intersection P6 between the straight line 30MD and the straight line 30MC.

  Next, the detection area determination unit 24 draws a virtual straight line 30LD in the horizontal direction at a position that is raised by a predetermined distance LD from P6.

  Next, the detection region determination unit 24 detects the right end portion 30R and the left end portion 30L of the basic image 30, and a region (hatched line) surrounded by the end portions 30R, 30D, the lower end portion 30D, and the virtual straight line 30LD. (Part) is determined as the detection area 32 of the suspicious object.

  The distance LD is a distance corresponding to about 20 cm to 50 cm from the actual floor surface 11f, and is determined according to the appearance when projected on the captured image of the camera 21. That is, the distance LD is set in advance, but actually differs depending on the installation position of the camera 21 and the like.

  Here, a description will be given of which part in the actual passenger car cabin the straight lines 30MC and 30MD described above correspond to.

  In FIG. 7, the car door 12 is a two-door central opening type, and has a structure in which the doorway 11a is opened and closed by moving the two door panels 12a and 12b to the left and right in conjunction with each other. Therefore, the straight line 30MC first detected in the vertical direction from the center point P4 of the lower end 30D of the basic image 30 is a boundary line between the car door 12 and the floor surface 11f, and the intersection point P5 is on the boundary line. It turns out that it is 1 point.

  Further, the straight line 30MC that searches the horizontal line 30MC from the intersection point P5 in the horizontal direction and extends upward from the straight line 30MC is the straight line 30MD that is closest to the intersection point P5. This is a boundary line between the door panels 12a and 12b.

  Here, the reason for searching on the straight line 30MC in the horizontal direction is that the lower end center point P4 of the basic image 30 that is a captured image may be biased to either the left or right depending on the mounting position or mounting angle of the camera 21. Because. In this way, the detection region 32 can be determined by detecting an inverted T-shaped figure composed of the floor surface 11f and the car door 12.

  As described above, the image difference detection unit 25 extracts the images in the detection region 32 determined by the detection region determination unit 24 from the basic image 30 and the real-time image 31 and compares them, thereby determining the presence or absence of a suspicious object. To detect. In this case, even if a sticker or poster brought in by the elevator administrator or the like is put in the car 11, the image comparison can be performed in the detection area 32 except for these, thereby reducing the possibility of erroneous detection. can do.

  In FIG. 7, the description has been given by taking the two-door central opening type car door 12 as an example. However, even in the case of the two-door single-opening configuration in which the two door panels 12 a and 12 b move in the same direction, the same as above. Thus, the boundary line between the two door panels 12a and 12b can be detected.

  In this way, the lower end of the car door 12 in the photographed image is detected, and the detection area 32 is determined based on the lower end of the car door 12, as in the first embodiment. The suspicious object detection area 32 is appropriately set so as to remove the stickers pasted in the car 11, and the suspicious object left in the car 11 from the captured image or the suspicious object staying in the car 11. The person can be accurately detected.

  As described above, according to each embodiment, in a stay detection system that detects a suspicious object left in an elevator car or a suspicious person staying in the car, a sticker or a poster is easily pasted. The location can be automatically outside the detection area. As a result, the possibility of erroneous detection of stay can be reduced, and the effort for setting the area can be reduced.

  In the third and fourth embodiments, the description is given on the assumption that the car has a two-door car door. However, the car door has a one-door system, a three-door system, or a four-door system. However, the same method as described above can be used. This is because the upper end portion of the car door is detected in the third embodiment and the lower end portion of the car door is detected in the fourth embodiment, so that it is only necessary to distinguish between the car door and the others.

  That is, for example, in the case of a single door type car door, the boundary line between the car door and the car side surface may be detected. Moreover, what is necessary is just to detect the boundary line between adjacent door panels, if it has two or more door panels.

  In the third embodiment, the boundary between the curtain plate and the car door is detected. However, if there is no curtain plate, the boundary between the ceiling surface of the car room and the car door is detected. There is no problem.

  In the detection region determination method described with reference to FIGS. 3 to 7, the virtual straight lines (30LA, 30LB, 30LC, and 30LD) drawn on the captured image are described as being drawn in the horizontal direction. It may be drawn so as to be parallel to the boundary line between the curtain and the car door or the boundary line between the car floor and the car door. In this way, it is possible to reduce the influence of image distortion and the like caused by differences in camera installation position and angle of view.

  In short, the present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. Moreover, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  DESCRIPTION OF SYMBOLS 11 ... Car, 11a ... Entrance / exit, 11b ... Car door facing surface, 11c ... Ceiling surface, 11d, 11e ... Side surface, 11f ... Floor surface, 12 ... Car door, 12a, 12b ... Door panel, 12c ... Curtain board, 13 DESCRIPTION OF SYMBOLS ... Door control part, 14 ... Elevator operation control part, 21 ... Camera, 22 ... Image processing apparatus, 23 ... Basic image storage part, 24 ... Detection area determination part, 25 ... Image difference detection part, 30 ... Basic image, 31 ... Real-time image, 32 ... detection area.

Claims (9)

A surveillance camera that is installed in the car and photographs the interior of the car according to the height of the car door provided at the entrance of the car.
Basic image storage means for storing a basic image of the car taken by the camera;
A detection area determining means for analyzing a captured image of the camera and determining a detection area of a suspicious object based on the upper end or the lower end of the captured image;
Based on the detection area determined by the detection area determination means, the image captured at regular intervals by the imaging means during operation of the elevator and the basic image stored in the basic image storage means, An elevator car monitoring device comprising: an image difference detecting means for detecting the presence or absence of a suspicious object from the difference between the two.   The detection area determining means recognizes the upper end portion of the captured image as being near the upper end portion of the car door, and horizontally draws it at a position that is lower than the upper end portion of the captured image by a preset first distance. The elevator car monitoring apparatus according to claim 1, wherein a region surrounded by the virtual line, the lower end portion of the captured image, and the left and right end portions of the captured image is determined as a suspicious object detection region.   The first distance corresponds to a distance when a distance from the ceiling surface in the car room to an average height at which the paper is pasted is projected onto the photographed image. Elevator car monitoring device.   The detection area determination means recognizes the lower end portion of the captured image as being near the floor surface in the car room, and horizontally draws it at a position that is a predetermined second distance above the lower end portion of the captured image. 2. The elevator in-car monitoring device according to claim 1, wherein a region surrounded by the virtual line, the lower end of the captured image, and the left and right ends of the captured image is determined as a suspicious object detection region.   The second distance corresponds to a distance obtained by projecting, on the captured image, a distance obtained by adding an average height at which stickers are pasted from the floor surface of the car room to the depth dimension of the car. The elevator in-car monitoring device according to claim 4.   The detection area determination means recognizes the upper end portion of the captured image as the vicinity of the upper end portion of the car door, and first detects a horizontal straight line first detected downward from the center point of the upper end portion of the captured image. Among the straight lines extending downward from the first straight line, the straight line closest to the center point is defined as a second straight line, and is preset from the intersection of the first straight line and the second straight line. A region surrounded by a virtual line drawn in the horizontal direction at a position lowered by the third distance, the lower end of the captured image, and the left and right ends of the captured image is determined as a suspicious object detection region. The elevator in-car monitoring device according to claim 1.   7. The elevator car according to claim 6, wherein the third distance corresponds to a distance when a distance obtained by subtracting a predetermined value from the height of the car door is projected on the photographed image. Inside monitoring device.   The detection area determining means recognizes the lower end portion of the captured image as being near the floor surface in the car room, and a third horizontal straight line first detected upward from the center point of the lower end portion of the captured image is a third. Among the straight lines extending upward from the third straight line, the straight line closest to the center point is defined as a fourth straight line, and is preset from the intersection of the third straight line and the fourth straight line. A region surrounded by a virtual line drawn in the horizontal direction at the position raised by the fourth distance, the lower end portion of the captured image, and the left and right end portions of the captured image is determined as a suspicious object detection region. The elevator in-car monitoring device according to claim 1.   9. The elevator car according to claim 8, wherein the fourth distance corresponds to a distance when a distance from a floor surface in the car room to a predetermined height is projected on the photographed image. Monitoring device.

Images