elevator system

Abstract

To provide an elevator system capable of preventing users from being pulled in. [Solution] An elevator system according to one embodiment includes a camera that photographs users inside the car, a setting means that sets multiple detection areas on the photographed image to detect users near the car door, a detection means that detects users from the multiple detection areas, an identification means that, when a user is detected by the detection means, identifies a pull-in risk value that indicates the risk of the user being pulled into the car door, a determination means that compares the identified pull-in risk value with a risk threshold that is preset for the detection area in which the user detected by the detection means is located, and determines whether the user is at risk of being pulled into the car door, and a control means that, when the determination means determines that there is a risk of the user being pulled in, announces to the user inside the car that there is a risk of the user being pulled in.

Description

[Technical Field]

[0001] FIELD OF THE INVENTION An embodiment of the present invention relates to an elevator system. [Background technology]

[0002] When elevator doors open, a passenger's fingers may be pulled into the door pocket. Various technologies have been devised to prevent such accidents, but there is a need for technology that can more accurately prevent passengers from being pulled in. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6271776 [Patent Document 2] Patent No. 7155201 [Patent Document 3] Patent No. 6638696 Summary of the Invention [Problem to be solved by the invention]

[0004] The problem to be solved by the present invention is to provide an elevator system that can prevent users from being pulled in. [Means for solving the problem]

[0005] An elevator system according to one embodiment includes a camera that photographs a user inside a car; a setting means that sets a plurality of detection areas on an image photographed by the camera to detect users near the car door; a detection means that detects users from the plurality of detection areas; an identification means that, when a user is detected by the detection means, identifies a pull-in risk value that indicates the risk of the user being pulled into the car door; a determination means that compares the identified pull-in risk value with a predetermined risk threshold for the detection area in which the user detected by the detection means is located, and determines whether the user is at risk of being pulled into the car door; and a control means that, when the determination means determines that there is a risk of the user being pulled in, announces to the user inside the car that there is a risk of the user being pulled in. [Brief explanation of the drawings]

[0006] [Figure 1] FIG. 1 is a diagram showing a schematic configuration example of an elevator system according to an embodiment. [Figure 2] FIG. 2 is a diagram for explaining an example of a detection area that is set when the car door opening / closing system is a two-door single swing system. [Figure 3] FIG. 3 is a diagram for explaining another example of the detection area set when the car door opening / closing system is a two-door single swing system. [Figure 4] FIG. 4 is a diagram for explaining an example of a detection area that is set when the car door opens and closes in a double-opening central manner. [Figure 5] FIG. 5 is a flowchart showing an example of a procedure for a detection area setting process according to the embodiment. [Figure 6] FIG. 6 is a diagram for explaining the orientation of the user's body and the position of his / her hands relative to the car door, which are detected by the user detection unit according to the embodiment. [Figure 7] FIG. 7 is a diagram showing an example of a risk value table according to the embodiment. [Figure 8]FIG. 8 is a diagram showing another example of the risk value table according to the embodiment. [Figure 9] FIG. 9 is a flowchart showing an example of a procedure for the drag-in risk determination process according to the embodiment. [Figure 10] FIG. 10 is a diagram specifically illustrating the entrapment risk determination process according to the embodiment. [Figure 11] FIG. 11 is another diagram specifically illustrating the entrapment risk determination process according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0007] Hereinafter, an embodiment will be described with reference to the drawings. It should be noted that the disclosure is merely an example, and the invention is not limited to the contents described in the following embodiments. Modifications that can be easily conceived by a person skilled in the art are naturally included in the scope of the disclosure. For clearer explanation, the size, shape, etc. of each part may be changed from the actual embodiment and shown schematically in the drawings. In multiple drawings, corresponding elements may be given the same reference numerals, and detailed explanations may be omitted.

[0008] 1 is a diagram showing a schematic configuration example of an elevator system according to one embodiment. Note that, although a single car will be described as an example here, the same configuration applies to multiple cars.

[0009] The elevator system of this embodiment has the function of detecting a user near the car door inside the car, determining whether there is a risk that the detected user's fingers, hands, etc. will be pulled into the door pocket, for example (i.e., determining whether there is a risk of a pulling-in accident occurring), and issuing a warning to the user inside the car if it is determined that there is a risk of a pulling-in accident occurring.

[0010] A camera 12 is installed above the entrance / exit of the car 11. Specifically, the camera 12 is installed in a panel 11a that covers the upper part of the entrance / exit of the car 11, with the lens portion facing directly downward, or toward the landing side or the car side. The camera 12 has, for example, a wide-angle lens or a fisheye lens, and continuously captures images at several frames per second (for example, 30 frames per second). The shooting range of the camera 12 needs to include at least the vicinity of the car door 13. Note that the location of the camera 12 does not have to be above the entrance / exit of the car 11, as long as it is near the car door 13.

[0011] At the landing 15 on each floor, a landing door 14 is installed at the arrival entrance of the car 11. The landing door 14 engages with the car door 13 when the car 11 arrives and opens and closes. The power source (door motor) is on the car 11 side, and the landing door 14 simply opens and closes following the car door 13.

[0012] Each image (video) continuously captured by the camera 12 is analyzed in real time by the image processing device 20. For convenience, the image processing device 20 is shown removed from the car 11 in Fig. 1, but in reality, the image processing device 20 is housed in the headboard 11a together with the camera 12.

[0013] The image processing device 20 includes a storage unit 21 and a control unit 22. The storage unit 21 is composed of a memory device such as a RAM. The storage unit 21 sequentially stores images captured by the camera 12. The storage unit 21 has a buffer area for temporarily storing data necessary for processing by the control unit 22. The storage unit 21 may store images that have been subjected to preprocessing such as distortion correction, enlargement / reduction, and partial cropping as preprocessing of the images captured by the camera 12. The storage unit 21 further stores a risk value table, which will be described later.

[0014] The control unit 22 uses images captured by the camera 12 to detect users near the car door 13, determines whether the detected users are at risk of being pulled into the car door 13, and if it determines that there is a risk of the users being pulled in, performs various controls to realize the function of warning the users in the car 11.

[0015] The control unit 22 includes a detection area setting unit 23 , a user detection unit 24 , a user attraction risk value specification unit 25 , a user attraction risk determination unit 26 , and a communication unit 27 .

[0016] The detection area setting unit 23 sets a detection area for detecting a user near the car door 13 on the image captured by the camera 12. In this embodiment, multiple types of detection areas with different risk thresholds (to be described later) are set according to the risk of being pulled into the car door 13. The detection area setting unit 23 sets multiple types of detection areas at different positions for each door opening / closing method (door type) of the car door 13.

[0017] FIG. 2 is a diagram for explaining the detection area set when the door opening / closing system of the car door 13 is a two-door single swing system.

[0018] The detection area setting unit 23 sets a high-risk detection area E1, a medium-risk detection area E2, and a low-risk detection area E3 near the car door 13 as areas where there is a risk of a dragging-in accident occurring (dragging-in risk).

[0019] The high-risk detection area E1 is an area where the risk of being pulled in is the highest, and is an area where the lowest risk threshold (first risk threshold) is set as a risk threshold to be described later.

[0020] When the car door 13 is a two-door single swing type, it is expected that a pull-in accident will occur at point P1, which corresponds to the door pocket (near the door pocket), or at point P2, which corresponds to the overlapping portion of the two doors. Therefore, the detection area setting unit 23 sets the area near points P1 and P2, where a pull-in accident is expected to occur, as a high-risk detection area E1. Specifically, the detection area setting unit 23 sets two areas extending, for example, by L1 (e.g., 10 cm) in all directions from points P1 and P2, where a pull-in accident is expected to occur, as high-risk detection areas E1a and E1b (i.e., square areas with sides of 2 × L1 centered on points P1 and P2). Hereinafter, the points where a pull-in accident is expected to occur (locations where pull-in accidents are likely to occur) will be referred to as "pull-in points." Hereinafter, although a case where L1 is 10 cm is given as an example, the value of L1 is not limited to this and may be set to any value.

[0021] The medium risk detection area E2 is an area with a lower risk of being drawn in than the high risk detection area E1, and is an area in which a risk threshold (second risk threshold) higher than the first risk threshold set in the high risk detection area E1 is set as the risk threshold described below.

[0022] The detection area setting unit 23 sets as medium-risk detection areas E2a and E2b an area extending L2 (for example, 30 cm) in the horizontal direction of the door from the end e1 on the doorstop side of the high-risk detection area E1a set near the retraction point P1, and an area extending L2 in the horizontal direction of the door from the end e2 on the doorstop side of the high-risk detection area E1b set near the retraction point P2. Note that although the case where L2 is 30 cm is given here as an example, the value of L2 is not limited to this and may be set to any value.

[0023] The low risk detection area E3 is an area with a lower risk of being drawn in than the medium risk detection area E2, and is an area in which a risk threshold (third risk threshold) higher than the second risk threshold set in the medium risk detection area E2 is set as the risk threshold described below.

[0024] The detection area setting unit 23 sets a low-risk detection area E3 around the perimeter L3 (e.g., 10 cm) of the high-risk detection area E1 and the medium-risk detection area E2. Note that the low-risk detection area set around the perimeter L3 of the high-risk detection area E1a and the medium-risk detection area E2a and the low-risk detection area set around the perimeter L3 of the high-risk detection area E1b and the medium-risk detection area E2b partially overlap and become one, so for convenience, they are shown as one low-risk detection area E3 in Figure 2, but in reality, two low-risk detection areas are set. Also, here, a case where L3 is 10 cm is given as an example, but the value of L3 is not limited to this and may be set to any value.

[0025] Here, the case where the detection area shown in Fig. 2 is set when the door opening / closing system of the car door 13 is a two-door, single-swing system has been shown, but this is not limiting. When the door opening / closing system of the car door 13 is a two-door, single-swing system, for example, the detection area shown in Fig. 3 may be set. The detection area shown in Fig. 3 differs from the detection area shown in Fig. 2 in that the car-side range of high-risk detection area E1b set near retraction point P2 is expanded to the same range as the car-side range of high-risk detection area E1a set near retraction point P1 and extends L1 toward the door pocket side (in other words, high-risk detection area E1b is set so as to contact both medium-risk detection areas E2a and E2b). Accordingly, medium-risk detection area E2b, which extends L2 from the doorstop-side end e2 of high-risk detection area E1b in the horizontal direction of the door, also differs from the detection area shown in Fig. 2 in that the car-side range of

[0026] FIG. 4 is a diagram for explaining the detection area that is set when the car door 13 is a center double swing type.

[0027] When the car door 13 is a center double-swing type, it is expected that a pull-in accident will occur at pull-in points P3 and P4, which correspond to the door pockets on both the left and right sides (near the door pockets). For this reason, the detection area setting unit 23 sets the area near the pull-in points P3 and P4 as a high-risk detection area E1. Specifically, the detection area setting unit 23 sets two areas extending, for example, by L1 (for example, 10 cm) in all directions from the pull-in points P3 and P4 as high-risk detection areas E1a and E1b (that is, square areas with sides of 2 × L1 length and centered at points P3 and P4).

[0028] In addition, the detection area setting unit 23 sets as medium risk detection areas E2a and E2b an area extending L2 (e.g., 30 cm) from the end e1 on the center side of the car door of the high risk detection area E1a set near the retraction point P3 in the horizontal direction of the door, and an area extending L2 from the end e2 on the center side of the car door of the high risk detection area E1b set near the retraction point P4 in the horizontal direction of the door.

[0029] Furthermore, the detection area setting unit 23 sets a low-risk detection area E3a within a perimeter L3 (e.g., 10 cm) of the high-risk detection area E1a and the medium-risk detection area E2a, and sets a low-risk detection area E3b within a perimeter L3 of the high-risk detection area E1b and the medium-risk detection area E2b.

[0030] In the case of FIG. 4, L1, L2, and L3 are not limited to the values ​​shown as an example, and may be set to any values.

[0031] 2 to 4 show the positions of the detection areas E1, E2, E3 before the doors are opened (while the doors are all closed), but the positions of the detection areas E1, E2, E3 may be set to move as the doors are opened.

[0032] For example, if the door opening / closing method of the car door 13 is a two-door single-swing method, the position of the high-risk detection area E1b shown in Figure 2 may be moved toward the door pocket as the door opens while maintaining the size of the area, and the positions of the medium-risk detection areas E2a, E2b and low-risk detection area E3 may be moved toward the door pocket as the door opens while gradually reducing the size of each area E2a, E2b, E3.

[0033] In addition, when the door opening / closing method of the car door 13 is a center double-swing method, the positions of the medium risk detection area E2a and low risk detection area E3a shown in Figure 4 may be moved toward the door pocket on the left side of the figure as the door opens while gradually reducing the size of each area E2a, E3a, and the positions of the medium risk detection area E2b and low risk detection area E3b may be moved toward the door pocket on the right side of the figure as the door opens while gradually reducing the size of each area E2b, E3b.

[0034] The detection area setting unit 23 may set multiple detection areas with different risk thresholds depending on the distance from the attraction point, rather than setting detection areas according to the door opening / closing method of the car door 13. For example, the detection area setting unit 23 may set multiple concentric detection areas with different risk thresholds centered on the attraction point. In this case, the risk threshold set for each of the multiple detection areas is set to a lower value for the detection area closer to the attraction point, and a higher value for the detection area farther from the attraction point. Although the case where multiple detection areas are set in a concentric circle shape has been illustrated as an example here, the multiple detection areas may be set to any shape as long as they gradually expand from the attraction point as the center.

[0035] Furthermore, the detection area setting unit 23 may set multiple detection areas with different risk thresholds according to the distance from the draw-in point in the horizontal direction of the door. For example, the detection area setting unit 23 may set multiple detection areas that are aligned in the horizontal direction of the door from the draw-in point and have different risk thresholds. In this case, too, the risk threshold set for each of the multiple detection areas is set so that the closer the detection area is to the draw-in point, the lower the value, and the farther the detection area is from the draw-in point, the higher the value.

[0036] Here, an example of the procedure of the detection area setting process executed by the detection area setting unit 23 will be described with reference to the flowchart of Fig. 5. This process is executed as part of the initial setting of the image processing device 20, for example.

[0037] First, the detection area setting unit 23 acquires door type information indicating the door opening / closing method of the car door 13 (step S1). The door type information may be acquired from the elevator control device 30 or from a mobile terminal (not shown) operated by a maintenance worker.

[0038] Next, the detection area setting unit 23 determines whether the door opening / closing method of the car door 13 indicated by the acquired door type information is a two-door single swing method or a central double swing method (step S2).

[0039] If the door opening / closing method of the car door 13 is a two-door single swing method (two-door single swing method in step S2), the detection area setting unit 23 detects a retraction point P1 corresponding to the door pocket and a retraction point P2 corresponding to the overlapping part of the two doors from the captured image (step S3).

[0040] Next, the detection area setting unit 23 sets two areas extending by L1 in all directions from the two attraction points P1 and P2 detected in the processing of step S3 as high-risk detection areas E1 on the captured image (step S4).

[0041] Next, the detection area setting unit 23 sets an area extending L2 horizontally from the door stop end of the high-risk detection area E1 set near the retraction point P1, and an area extending L2 horizontally from the door stop end of the high-risk detection area E1 set near the retraction point P2, as a medium-risk detection area E2 on the captured image (step S5).

[0042] Then, the detection area setting unit 23 sets two areas surrounding the two high-risk detection areas and the two medium-risk detection areas set on the captured image L3 as a low-risk detection area E3 on the captured image (step S6), and the detection area setting unit 23 ends the series of detection area setting processes.

[0043] On the other hand, if the door opening / closing method of the car door 13 is a double-door center type (double-door center type in step S2), the detection area setting unit 23 detects a retraction point P3 corresponding to the door pocket of the left car door 13 and a retraction point P4 corresponding to the door pocket of the right car door 13 from the captured image (step S7).

[0044] Next, the detection area setting unit 23 sets two areas extending by L1 in all directions from the two attraction points P3 and P4 detected in the processing of step S7 as high-risk detection areas E1 on the captured image (step S8).

[0045] Next, the detection area setting unit 23 sets, on the captured image, an area extending L2 from the end of the high-risk detection area E1 set near the retraction point P3 toward the center of the car door in the direction horizontal to the door, and an area extending L2 from the end of the high-risk detection area E1 set near the retraction point P4 toward the horizontal to the door, as a medium-risk detection area E2 (step S9).

[0046] Then, the detection area setting unit 23 sets two areas surrounding the two high-risk detection areas and the two medium-risk detection areas set on the captured image L3 as a low-risk detection area E3 on the captured image (step S10), and the detection area setting unit 23 ends the series of detection area setting processes.

[0047] According to the series of detection area setting processes described above, the detection areas E1 to E3 can be set at appropriate positions according to the door opening / closing method of the car door 13.

[0048] Returning to the explanation of Figure 1 again. The user detection unit 24 uses the image captured by the camera 12 to detect users near the car door 13. Specifically, the user detection unit 24 divides the image captured by the camera 12 into a large number of blocks, and performs a motion detection process to detect blocks with movement (moving blocks) from among the blocks by focusing on changes in color information (for example, changes in brightness values) of each block included in the detection area. According to this motion detection process, a user appearing in the image is detected as a collection of moving blocks.

[0049] Note that the user detection unit 24 may detect users not by the above-described motion detection process but by, for example, SSD (Single Shot Multibox Detector), which is one of the image recognition methods using supervised learning. In this case, the user detection unit 24 can efficiently detect users near the car door 13, including their positions, from newly captured images by learning images that show users near the car door 13 in advance as training data.

[0050] In addition, when a user is detected near the car door 13, the user detection unit 24 executes a process to detect the attributes (characteristics) of the user and the orientation of the user's body and hand position relative to the car door 13 (more specifically, the attraction point).

[0051] Examples of user attributes include adult, child, infant, elderly, drunk, etc.

[0052] The orientation of the user's body relative to the car door 13 may be, for example, facing forward, to the side, or backward.

[0053] Furthermore, the position of the user's hand relative to the car door 13 is the hand position x when the retraction point (closest to the user) is assumed to be 0°, as shown in Figure 6, for example, and indicates the position of the hand closer to the car door 13.

[0054] For example, Fig. 6(a) shows a case where the orientation of the user's body relative to the car door 13 (retraction point) is "sideways" and the position x of the user's hands (in this case, the position of the right hand) relative to the car door 13 (retraction point) is "x=0°". Fig. 6(b) shows a case where the orientation of the user's body relative to the car door 13 (retraction point) is "frontways" and the position x of the user's hands (in this case, the positions of both the right and left hands) relative to the car door 13 (retraction point) is "x=90°".

[0055] The user detection unit 24 may detect the user's attributes, the orientation of the user's body relative to the car door 13, and the position of the user's hands relative to the car door 13 simply by image recognition, or may detect the skeleton of a user near the car door 13 using a known keypoint technique such as Open Pose, and then detect the user's attributes, the orientation of the user's body relative to the car door 13, and the position of the user's hands relative to the car door 13.

[0056] When the user detection unit 24 detects a user near the car door 13, the entrapment risk value identification unit 25 identifies an entrapment risk value based on the attributes of the detected user and the body orientation and hand position of the user. More specifically, the entrapment risk value identification unit 25 refers to a risk value table corresponding to the attributes of the detected user from among the risk value tables stored in advance in the storage unit 21, and identifies an entrapment risk value corresponding to the body orientation and hand position of the user.

[0057] Here, the risk value tables stored in advance in the storage unit 21 will be described. The storage unit 21 stores in advance risk value tables prepared for each user attribute. FIG. 7 shows an example of a risk value table t1 in which the user attribute is "adult." FIG. 8 shows an example of a risk value table t2 in which the user attribute is "child." Note that the various attraction risk values ​​defined in the risk value tables t1 and t2 shown in FIGS. 7 and 8 are merely examples, and are not limited to these values.

[0058] The entrainment risk value is a value indicating the risk (danger) of an entrainment accident occurring, and the higher the value, the higher the risk of an entrainment accident occurring.

[0059] From the perspective of a user's attributes, the draw-in risk value is set according to whether the user can determine for himself or herself whether there is a risk of a draw-in accident occurring. Specifically, the draw-in risk value is set higher the attribute that is expected to make it difficult for the user to determine for himself or herself whether there is a risk of a draw-in accident occurring. For this reason, the draw-in risk value defined in the risk value table t2 shown in FIG. 8 where the user attribute is "child" is set higher overall than the draw-in risk value defined in the risk value table t1 shown in FIG. 7 where the user attribute is "adult."

[0060] The risk value table defines a risk value of being pulled in according to the orientation of the user's body relative to the car door 13 and the position of the user's hand relative to the car door 13. The entrapment risk value is set in terms of the orientation of the user's body relative to the car door 13, depending on whether the user is looking at the car door 13 or not, and the more the user's body orientation is expected to indicate that the user is not looking at the car door 13, the higher the entrapment risk value is set.

[0061] For example, it is desirable that the entrainment risk value corresponding to a body orientation relative to the car door 13 being "backward" be set to a value equal to or greater than the entrainment risk values ​​corresponding to a body orientation relative to the car door 13 being "front" or "side." It is also desirable that the entrainment risk value corresponding to a body orientation relative to the car door 13 being "side" be set to a value equal to or greater than the entrainment risk value corresponding to a body orientation relative to the car door 13 being "front."

[0062] From the viewpoint of the position of the user's hand relative to the car door 13, the closer the position of the user's hand is to the car door 13, the higher the entrapment risk value is set.

[0063] For example, the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "45°≦x<0°" is desirably set to a value equal to or greater than the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "90°≦x<45°", "135°≦x<90°", or "180°≦x<135°". Furthermore, the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "90°≦x<45°" is desirably set to a value equal to or greater than the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "135°≦x<90°" or "180°≦x<135°". Furthermore, the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "135°≦x<90°" is desirably set to a value equal to or greater than the entrapment risk value corresponding to the user's hand position x relative to the car door 13 being "180°≦x<135°".

[0064] However, it is desirable that the entrainment risk value set in the risk value table t2 indicating that the user attribute is "child," and corresponding to a hand position x relative to the car door 13 of "45°≦x<0°," be set higher when the body is facing "side" relative to the car door 13 than when the body is facing "front." This is because it is expected that children will approach the car door 13 out of curiosity, and the risk of being entrained when they are trying to touch the car door 13 out of curiosity is higher than the risk of being entrained when they are standing sideways to the car door 13 and their hands are in a position where they are likely to be naturally drawn in. This makes it possible to take care of children who approach the car door 13 out of curiosity, and to appropriately determine the risk of being entrained.

[0065] Here, an example of the process of identifying the attraction risk value by the attraction risk value identifying unit 25 will be described. For example, if the user detection unit 24 detects an "adult" user whose body is facing "forward" relative to the car door 13 and whose hand position relative to the car door 13 is "0°," the entrapment risk value identification unit 25 refers to the risk value table t1 in Figure 7 and identifies the entrapment risk value of the user as "0.9."

[0066] Furthermore, when the user detection unit 24 detects a "child" user whose body is facing "backwards" relative to the car door 13 and whose hands are positioned "90°" relative to the car door 13, the entrapment risk value identification unit 25 refers to the risk value table t2 of Figure 8 and identifies the entrapment risk value of the user as "4.5".

[0067] Returning to the explanation of Figure 1 again. The entrainment risk determination unit 26 compares the entrainment risk value of the predetermined user identified by the entrainment risk value identification unit 25 with a risk threshold preset for the detection area in which the predetermined user is detected, and determines whether the entrainment risk value is equal to or greater than the risk threshold. If the entrainment risk value is equal to or greater than the risk threshold, the entrainment risk determination unit 26 determines that there is a risk of a entrainment accident occurring (high risk). On the other hand, if the entrainment risk value is less than the risk threshold, the entrainment risk determination unit 26 determines that there is no risk of a entrainment accident occurring (low risk).

[0068] In this embodiment, as described above, the entrainment risk determination unit 26 determines that there is a risk of a entrainment accident occurring when the entrainment risk value is equal to or greater than the risk threshold, and determines that there is no risk of a entrainment accident occurring when the entrainment risk value is less than the risk threshold. However, without being limited to this, the entrainment risk determination unit 26 may determine that there is a risk of a entrainment accident occurring when the entrainment risk value is greater than the risk threshold, and may determine that there is no risk of a entrainment accident occurring when the entrainment risk value is equal to or less than the risk threshold.

[0069] The communication unit 27 acquires door information indicating the current door open / closed state of the car doors 13 from the elevator control device 30. Furthermore, when the entrainment risk determination unit 26 determines that there is a risk of an entrainment accident occurring, the communication unit 27 outputs a request signal corresponding to the current door open / closed state of the car doors 13 to the elevator control device 30.

[0070] Specifically, when the current door opening / closing state of the car doors 13 is during the door opening operation, the communication unit 27 outputs to the elevator control device 30 a first request signal to slow down the door opening speed of the car doors 13 from the current speed, and a second request signal to notify the passengers in the car 11 that there is a risk of being pulled in. Furthermore, when the current door opening / closing state of the car doors 13 is not during the door opening operation, the communication unit 27 outputs only the second request signal described above to the elevator control device 30.

[0071] The elevator control device 30 is made up of a computer equipped with a CPU, ROM, RAM, etc. The elevator control device 30 controls the raising and lowering operation of the car 11 and the operation of various devices installed in the car 11. The elevator control device 30 stores door type information indicating the door opening and closing method of the car door 13 installed in the car 11 in a memory (not shown), and outputs this to the image processing device 20 in response to a request from the image processing device 20.

[0072] The elevator control device 30 includes a door opening / closing control unit 31 and a notification control unit 32. The door opening / closing control unit 31 controls the opening and closing of the car door 13 when the car 11 arrives at the landing 15. In detail, the door opening / closing control unit 31 opens the car door 13 when the car 11 arrives at the landing 15, and closes the door after a predetermined time has elapsed.

[0073] Furthermore, in response to a request from the image processing device 20, the door opening / closing control unit 31 outputs door information indicating the current door open / closed state of the car door 13 to the image processing device 20. Furthermore, when the door opening / closing control unit 31 receives the above-mentioned first request signal from the image processing device 20, it controls the door opening speed of the car door 13 to be slower than the current speed.

[0074] The notification control unit 32 broadcasts an announcement through a speaker (not shown) installed inside the elevator 11 to notify (warn) passengers of the risk of being pulled in and causing an accident. Specifically, when the notification control unit 32 receives the second request signal from the image processing device 20, it broadcasts an announcement inside the elevator 11 to warn of the risk of being pulled in and causing an accident, such as "Your hand may be pulled into the door. Please move away from the door."

[0075] Next, an example of the procedure for the entrainment risk determination process executed by the image processing device 20 will be described with reference to the flowchart in Fig. 9. This process is executed, for example, while the car 11 is moving from one floor to another floor, or when the car arrives at a certain floor.

[0076] First, the user detection unit 24 of the image processing device 20 determines whether or not a user is present near the car door 13 based on the image captured by the camera 12 (in other words, detects whether or not a user is present near the car door 13) (step S11).

[0077] In the processing of step S11, if it is determined that there is no user near the car door 13 (No in step S11), the communication unit 27 of the image processing device 20 acquires door information indicating the current door open / close state of the car door 13 from the elevator control device 30 (step S12), and determines whether the car door 13 is in a fully open state (in other words, whether the door opening operation of the car door 13 has been completed) based on the acquired door information (step S13).

[0078] In the processing of step S13, if it is determined that the car door 13 is in a fully open state (in other words, the door opening operation of the car door 13 has been completed) (Yes in step S13), the image processing device 20 terminates the series of pull-in risk determination processes.

[0079] On the other hand, if it is determined in the processing of step S13 that the car door 13 is not in a fully open state (in other words, the door opening operation of the car door 13 has not been completed) (No in step S13), the image processing device 20 executes the processing of step S11 again.

[0080] Here, if it is determined in the processing of step S11 that a user is present near the car door 13 (Yes in step S11), the user detection unit 24 designates one of the detected users as a target user, and detects the attributes of the target user, as well as the body orientation and hand position of the target user relative to the car door 13 (step S14).

[0081] Next, the attraction risk value identification unit 25 identifies an attraction risk value corresponding to the target user based on the risk value table stored in the memory unit 21 that corresponds to the attributes of the target user detected in the processing of step S14 and the body orientation and hand position of the target user detected in the processing of step S14 (step S15).

[0082] Next, the attraction risk determination unit 26 compares the risk threshold set for the detection area in which the target user was detected in the processing of step S11 with the attraction risk value identified in the processing of step S15, and determines whether the attraction risk value is equal to or greater than the risk threshold (step S16).

[0083] In the processing of step S16, if it is determined that the entrapment risk value is not greater than the risk threshold, that is, that the entrapment risk value is less than the risk threshold (No in step S16), the entrapment risk determination unit 26 determines that there is no risk of the target user being entrapped by the car door 13, and determines whether all users detected in the processing of step S11 are target users (step S17).

[0084] If it is determined in the processing of step S17 that not all users are target users (No in step S17), users detected in the processing of step S11 who have not yet been designated as target users are designated as new target users, and the processing of step S14 is executed again.

[0085] On the other hand, if it is determined in the process of step S17 that all users have been set as target users (Yes in step S17), the image processing device 20 executes the process of step S12 again.

[0086] In the processing of step S16 described above, if it is determined that the entrainment risk value is equal to or greater than the risk threshold (Yes in step S16), the entrainment risk determination unit 26 determines that there is a risk that the target user will be entrained by the car door 13 and that an entrainment accident may occur, and executes processing to warn the user in the car 11. Specifically, first, similar to the processing of step S12, the communication unit 27 acquires door information from the elevator control device 30 (step S18), and determines whether or not the car door 13 is in the door-opening operation based on the acquired door information (step S19).

[0087] In the processing of step S19, if it is determined that the car doors 13 are in the door-opening operation (Yes in step S19), the communication unit 27 outputs to the elevator control device 30 a first request signal to slow down the door-opening speed of the car doors 13 from the current speed, and a second request signal to announce to the users in the car 11 that there is a risk of a draw-in accident (step S20). In this case, in response to the request from the image processing device 20, the elevator control device 30 slows down the door-opening speed of the car doors 13 and issues an announcement from a speaker (not shown) in the car 11 that there is a risk of a draw-in accident.

[0088] On the other hand, if it is determined in the processing of step S19 that the car door 13 is not in the door-opening operation (No in step S19), the communication unit 27 outputs only the second request signal described above to the elevator control device 30 (step S21). In this case, in response to the request from the image processing device 20, the elevator control device 30 issues an announcement from a speaker (not shown) in the car 11 to the effect that there is a risk of a dragging accident occurring.

[0089] When the processing of step S20 or S21 is executed, the image processing device 20 executes the processing of step S12 again. Thereafter, as described above, when it is determined in the processing of step S13 that the car door 13 is in the fully open state (in other words, the door opening operation of the car door 13 has been completed), the image processing device 20 ends the series of pull-in risk determination processing.

[0090] Here, when it is determined that the entrapment risk value is equal to or greater than the risk threshold, the image processing device 20 determines whether or not the car door 13 is in the door-opening operation based on the door information acquired from the elevator control device 30, and performs different processing when the car door 13 is in the door-opening operation and when the car door 13 is not in the door-opening operation. However, this is not limited to this, and when it is determined that the entrapment risk value is equal to or greater than the risk threshold, the image processing device 20 may perform processing to output to the elevator control device 30 a request signal to notify users in the car 11 that there is a risk of an entrapment accident occurring, and a request signal to slow down the door-opening speed of the car door 13 below normal, regardless of the current door open / close state of the car door 13.

[0091] Fig. 10 shows an example of a case where it is determined that there is no risk of being drawn in when the series of processes for determining the risk of being drawn in shown in Fig. 9 is executed. Fig. 11 shows an example of a case where it is determined that there is a risk of being drawn in when the series of processes for determining the risk of being drawn in shown in Fig. 9 is executed.

[0092] 10, it is assumed that a "child" whose body is facing "forward" with respect to the car door 13 and whose hand position is "90°" relative to the car door 13 is detected from a low-risk detection area E3 for which a risk threshold of "5" has been set, and that the entrainment risk value corresponding to the child is identified as "4.5." Also, it is assumed that in FIG. 10, an "adult" whose body is facing "forward" with respect to the car door 13 and whose hand position is "0°" relative to the car door 13 is detected from a high-risk detection area E1 for which a risk threshold of "1" has been set, and that the entrainment risk value corresponding to the adult is identified as "0.9."

[0093] In this case, in order to determine whether or not there is a risk that a child detected from the low-risk detection area E3 will be drawn into the car door 13, the entrainment risk determination unit 26 compares the entrainment risk value "4.5" corresponding to the child with the risk threshold "5" set for the low-risk detection area E3, and determines that the entrainment risk value is not greater than or equal to the risk threshold, i.e., that there is no risk that the child will be drawn into the car door 13.

[0094] In addition, in order to determine whether or not an adult detected from the high-risk detection area E1 is at risk of being pulled into the car door 13, the pull-in risk determination unit 26 compares the pull-in risk value "0.9" corresponding to the adult with the risk threshold "1" set for the high-risk detection area E1, and determines that the pull-in risk value is not greater than or equal to the risk threshold, that is, that there is no risk of the adult being pulled into the car door 13.

[0095] 11, it is assumed that a "child" whose body is facing "forward" with respect to the car door 13 and whose hands are positioned at "90 degrees" relative to the car door 13 is detected from the medium risk detection area E2, for which a risk threshold of "3" has been set, and that the entrainment risk value corresponding to the child is identified as "4.5." Also, in FIG. 11, it is assumed that an "adult" whose body is facing "backward" with respect to the car door 13 and whose hands are positioned at "45 degrees" relative to the car door 13 is detected from the high risk detection area E1, for which a risk threshold of "1" has been set, and that the entrainment risk value corresponding to the adult is identified as "1.5."

[0096] In this case, in order to determine whether or not a child detected from the medium risk detection area E2 is at risk of being drawn into the car door 13, the entrainment risk determination unit 26 compares the entrainment risk value "4.5" corresponding to the child with the risk threshold "3" set for the medium risk detection area E2, and determines that the entrainment risk value is greater than or equal to the risk threshold, i.e., that there is a risk that the child will be drawn into the car door 13.

[0097] In addition, in order to determine whether or not an adult detected from the high-risk detection area E1 is at risk of being pulled into the car door 13, the pull-in risk determination unit 26 compares the pull-in risk value "1.5" corresponding to the adult with the risk threshold "1" set for the high-risk detection area E1, and determines that the pull-in risk value is equal to or greater than the risk threshold, i.e., that there is a risk that the adult will be pulled into the car door 13.

[0098] 10 and 11, when one user is located across multiple detection areas (that is, when one user is detected from multiple detection areas), the entrainment risk determination unit 26 uses the lower risk threshold among the risk thresholds set for the multiple detection areas to determine whether or not there is a risk of the user being entrained by the car door 13. This makes it possible to more reliably prevent entrainment accidents from occurring.

[0099] As described above, the elevator system of this embodiment sets multiple detection areas E1, E2, E3 on the image captured by the camera 12 to detect users near the car door 13, and when a user is detected near the car door 13, it identifies an entrapment risk value indicating the risk of the user being pulled into the car door 13, compares the identified entrapment risk value with a pre-set risk threshold for the detection area in which the user is detected, determines whether or not the user is at risk of being pulled into the car door 13, and if it is determined that there is a risk of the user being pulled into the car door 13, it announces to the user in the car 11 that there is a risk of being pulled into the car door 13.

[0100] This makes it possible to perform control according to the risk of a user being pulled in that is detected near the car door 13, and it is possible to prevent excessive control (for example, issuing a warning announcement, controlling the car door 13 to slow down the door opening speed, etc.) to prevent a user from being pulled in while ensuring the safety of the user. In other words, it is possible to appropriately determine whether there is a risk of being pulled in, and prevent a decrease in the operating efficiency of the car 11 while ensuring the safety of the user.

[0101] In this embodiment, the draw-in risk value is specified based on the user's attributes, the user's body orientation, and the user's hand position, and whether the user is looking at the car door 13 is determined only by the "body orientation," but this is not limited to this, and the draw-in risk value may be specified based on the user's attributes, the user's body orientation, the user's face orientation, and the user's hand position, and whether the user is looking at the car door 13 may be determined by the "body orientation" and the "face orientation." This makes it possible to more accurately determine whether the user is looking at the car door 13, and to more accurately identify the user's risk of being drawn in.

[0102] In addition, the risk value table pre-stored in the memory unit 21 may be updated based on, for example, data from past (during operation) entrainment accidents (for example, data indicating the entrainment risk value identified by the entrainment risk value identification unit 25 when an entrainment accident occurred, data indicating the judgment result by the entrainment risk judgment unit 26 using the entrainment risk value, data indicating the output results of various request signals by the communication unit 27, etc.).

[0103] This makes it possible to verify whether a past entrainment accident occurred after appropriate countermeasures were taken, and if the verification result indicates that the countermeasures were insufficient, to take countermeasures such as updating the entrainment risk value to be larger than the current value. The above-mentioned update process may be performed, for example, using artificial intelligence installed in the image processing device 20, or may be performed by a maintenance worker.

[0104] In addition, the detection area setting unit 23 may change the size of each detection area E1, E2, E3 and the risk threshold set for each detection area E1, E2, E3 depending on the type of building in which the elevator 11 is installed and the time of day.

[0105] For example, if the building in which the elevator 11 is installed is a "nursing home," there is a high possibility of an accident occurring in which the elevator is pulled in, so the detection area setting unit 23 may set the size of each detection area E1, E2, E3 to be larger than that of a typical building, and may also set the risk threshold for each detection area E1, E2, E3 to be lower (in other words, the risk value of being pulled in may be set to be more likely to be above the risk threshold).

[0106] Furthermore, since there is a high possibility of a pull-in accident occurring during "times when there are many children (for example, when they are commuting to school)," the detection area setting unit 23 may increase the size of each detection area E1, E2, E3 compared to normal times, and may set the risk threshold for each detection area E1, E2, E3 lower (in other words, the pull-in risk value may be set so that it is more likely to be equal to or greater than the risk threshold).

[0107] In this embodiment, it is assumed that a risk value table is prepared for each user attribute, but a risk value table may also be prepared for each user attribute and congestion level. As an example, a risk value table corresponding to a congestion level of "low" and a risk value table corresponding to a congestion level of "crowded" may be prepared as risk value tables corresponding to the user attribute "child." In this case, it is desirable that all of the attraction risk values ​​specified in the risk value table corresponding to the congestion level of "low" be set higher than the attraction risk values ​​specified in the risk value table corresponding to the congestion level of "crowded."

[0108] This makes it possible to reliably warn "children" who approach the car door 13 out of curiosity even when the car 11 is "quiet," thereby preventing accidents in which children are pulled in.

[0109] In this embodiment, when the entrapment risk determination unit 26 determines that there is a risk of being entrapped by the car door 13, the communication unit 27 outputs a second request signal to the elevator control device 30 to broadcast an announcement inside the car 11 saying, "Your hand may be entrapped by the door. Please move away from the door." However, this is not limited to this, and the communication unit 27 may output a request signal to the elevator control device 30 to broadcast a special announcement inside the car 11 depending on the current situation near the car door 13.

[0110] For example, in a situation where entrainment risk determination unit 26 determines that there is no entrainment risk for a certain adult user, but determines that there is a risk of a child (e.g., an infant) being held by the adult user being entrained, communication unit 27 may output a third request signal, different from the above-described second request signal, to elevator control device 30 in order to broadcast a special announcement inside car 11 saying, "Be careful not to get your baby pulled into the door." This makes it possible to warn users with more specific content, and to more effectively prevent entrainment accidents from occurring.

[0111] According to the embodiment described above, it is possible to provide an elevator system that can prevent users from being pulled in.

[0112] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the inventions and their equivalents as defined in the claims. [Explanation of symbols]

[0113] 11...car, 11a...curtain board, 12...camera, 13...car door, 14...landing door, 15...landing, 20...image processing device, 21...memory unit, 22...control unit, 23...detection area setting unit, 24...user detection unit, 25...draw-in risk value identification unit, 26...draw-in risk determination unit, 27...communication unit, 30...elevator control device, 31...door opening / closing control unit, 32...notification control unit.

Claims

1. A camera that photographs passengers inside the car; a setting means for setting a plurality of detection areas on the image captured by the camera to detect a user near the door of the elevator car; a detection means for detecting a user from the plurality of detection areas; an identification means for identifying a pull-in risk value indicating a risk of the user being pulled into the car door when the detection means detects the user; a determination means for comparing the identified entrapment risk value with a predetermined risk threshold for the detection area in which the user detected by the detection means is located, and determining whether or not there is a risk that the user will be entrapped by the car door; a control means for announcing to a user in the elevator car that there is a risk of being pulled in when the determination means determines that there is a risk of being pulled in; An elevator system comprising:

2. The identification means determining a risk of being pulled in by the user based on the detected attributes of the user and the user's body orientation and hand position relative to the elevator door; 10. The elevator system of claim 1.

3. The determination means If the identified entrainment risk value is equal to or greater than the risk threshold, it is determined that there is an entrainment risk.

10. The elevator system of claim 1.

4. When the door opening / closing method of the car door is a two-door single-opening method, the setting means a first detection area in which a first risk threshold, which is the lowest risk threshold, is set in the vicinity of a door pocket of the car door and in the vicinity of an overlapping portion of two doors constituting the car door; an area extending from the end of the door stop side of the first detection area set near the door pocket in the horizontal direction of the door, and an area extending from the end of the door stop side of the first detection area set near the overlapping portion of the two doors in the horizontal direction of the door, are set as second detection areas to which a second risk threshold higher than the first risk threshold is set; a third detection area is set around the first detection area and the second detection area, and a third risk threshold higher than the second risk threshold is set therein; 4. The elevator system of claim 3.

5. When the door opening / closing method of the car door is a central double-opening method, the setting means a first detection area in which a first risk threshold, which is the lowest risk threshold, is set, is set near two door pockets of the car door; an area extending from an end of a first detection area set near one door pocket toward a horizontal direction of the door from a door center side end, and an area extending from an end of a first detection area set near the other door pocket toward a horizontal direction of the door from a door center side end, as second detection areas for which a second risk threshold higher than the first risk threshold is set; a third detection area is set around the first detection area and the second detection area, and a third risk threshold higher than the second risk threshold is set therein; 4. The elevator system of claim 3.

6. the setting means sets a plurality of detection areas having different risk thresholds according to the distance from a position where entrapment is likely to occur; 4. The elevator system of claim 3.

7. The setting means sets a plurality of detection areas having different risk thresholds depending on the distance in the horizontal direction of the door from the position where the door is likely to be pulled in.

7. The elevator system of claim 6.

8. The risk threshold set for each of the plurality of detection areas is set to a lower value as the detection area approaches a position where the entrapment is likely to occur, and is set to a higher value as the detection area is farther from the position where the entrapment is likely to occur.

7. The elevator system of claim 6.

9. The determination means If the detected user is located across multiple detection areas, a lower risk threshold among the risk thresholds set for the multiple detection areas is used to determine whether or not there is a risk of the user being pulled into the car door.

10. The elevator system of claim 1.

10. When a plurality of users are detected by the detection means, the identification means identifies, for each of the detected users, a pull-in risk value indicating a risk of the user being pulled into the car door, The determination means determines whether or not there is a risk of each detected user being pulled into the car door based on the pull-in risk value identified for each detected user and a risk threshold set for the detection area in which each detected user is located, When it is determined that at least one of the detected users is at risk of being pulled in, the control means issues an announcement inside the elevator car that there is a risk of being pulled in.

10. The elevator system of claim 1.

11. The control means When the determination means determines that there is a risk of being pulled in, the control unit further performs control to slow down the door opening speed of the car door compared to normal.

10. The elevator system of claim 1.

12. The setting means The positions of the plurality of detection areas are moved in accordance with the door opening operation of the car door.

10. The elevator system of claim 1.

13. The elevator car further includes a storage means for storing a plurality of risk value tables in which a risk value of being pulled in based on the orientation of the user's body and the position of the user's hands relative to the door of the elevator car is set for each attribute of the user.

10. The elevator system of claim 1.

14. The determination means determines whether or not there is a risk that the user detected by the detection means will be pulled into the door of the car while the car is moving from one floor to another floor or when the car arrives at a certain floor.

10. The elevator system of claim 1.

Images