Elevator long object inspection device

Abstract

Provided is a long-object inspection device capable of more accurately detecting an abnormality in a long object of an elevator. In an elevator (1), a camera (14) captures an image of an upper part or a lower part of a car (6) so that a long object and a sheave are included in the range of the captured image. A long-object inspection device (13) includes a sheave detection unit (18), a long-object detection unit (19), and a determination unit (17). The sheave detection unit (18) performs image processing for detecting a sheave part in the image captured by the camera (14). The long-object detection unit (19) refers to the position of the sheave part detected by the sheave detection unit (18) and performs image processing for detecting a long-object part in the image captured by the camera (14). The determination unit (17) determines if there is any abnormality in the long object on the basis of the detected state of the long object.

Description

[Technical Field]

[0001] The present disclosure relates to a long object inspection device for an elevator. [Background technology]

[0002] Patent Document 1 discloses an example of a long object inspection device for an elevator. The long object inspection device includes a camera, an image processing device, and a determination device. The camera is mounted on a car traveling in a hoistway. The camera captures an image above the car so that the main rope, which is a long object, is included in the image. The image processing device performs image processing to extract the main rope portion from the image captured by the camera. The determination device uses a processed image, which is an image processed by the image processing device, to determine whether the main rope is caught on an object in the hoistway. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2015-20863 Summary of the Invention [Problem to be solved by the invention]

[0004] In an elevator hoistway, in addition to long objects such as the main rope that are subject to detection of abnormalities such as snagging, there are also long objects that are not subject to detection of abnormalities. Furthermore, the interior wall of the hoistway may have a linear texture. Therefore, if the long object inspection device of Patent Document 1 erroneously detects a long object or linear texture that is not subject to detection as a long object that is subject to detection, it may erroneously detect an abnormality such as a snagged long object.

[0005] The present disclosure relates to solving such problems, and provides a long object inspection device that can more accurately detect abnormalities in long objects in elevators. [Means for solving the problem]

[0006] The long object inspection device of the present disclosure is a long object inspection device that detects abnormalities in long objects wrapped around a trough in an elevator car shaft through which the car travels, and comprises: a photographing unit that photographs one or both of the above and below the car so that the long object and the trough are included in the photographing range; an image processing unit that performs image processing to detect portions of the long object in the image photographed by the photographing unit; and a judgment unit that judges whether or not there is an abnormality in the long object based on the state of the long object detected by the image processing unit, and the image processing unit comprises: a trough detection unit that performs image processing to detect portions of the trough in the image photographed by the photographing unit; and a long object detection unit that performs image processing to detect portions of the long object in the image photographed by the photographing unit by referring to the position of the trough portion detected by the trough detection unit. The long object inspection device of the present disclosure is a long object inspection device that detects abnormalities in long objects wrapped around a trough in an elevator car shaft through which the car travels, and is equipped with an image processing unit that performs image processing to detect portions of the long object in images captured by a photographing unit that photographs one or both of the above and below the car so that the long object and the trough are included in the photographing range, and a judgment unit that judges whether or not there is an abnormality in the long object based on the state of the long object detected by the image processing unit, and the image processing unit is equipped with a trough detection unit that performs image processing to detect portions of the trough in the images captured by the photographing unit, and a long object detection unit that performs image processing to detect portions of the long object in the images captured by the photographing unit by referring to the position of the trough portion detected by the trough detection unit. [Effects of the Invention]

[0007] The long object inspection device according to the present disclosure can more accurately detect abnormalities in long objects in elevators. [Brief explanation of the drawings]

[0008] [Figure 1]1 is a side view showing the configuration of an elevator according to a first embodiment. [Figure 2] 1 is a block diagram showing the functions of a long object inspection device according to a first embodiment. [Figure 3] 3 is a diagram showing an example of an image captured by a camera according to the first embodiment. FIG. [Figure 4] FIG. 2 is a side view showing a state in which an abnormality occurs in the governor rope of the elevator according to the first embodiment. [Figure 5] FIG. 2 is a diagram showing an example of an image captured by a camera when an abnormality occurs in a governor rope of the elevator according to the first embodiment. [Figure 6] FIG. 2 is a diagram showing a location where an abnormality has occurred in the governor rope of the elevator according to the first embodiment. [Figure 7] FIG. 2 is a diagram showing a location where an abnormality has occurred in the governor rope of the elevator according to the first embodiment. [Figure 8] 5 is a flowchart showing an example of the operation of the long object inspection device according to the first embodiment. [Figure 9] 1 is a hardware configuration diagram of a main part of a long object inspection device according to a first embodiment. [Figure 10] FIG. 10 is a block diagram showing the functions of a long object inspection device according to a second embodiment. [Figure 11] FIG. 10 is a diagram showing an example of a panoramic image of a hoistway according to the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] The following describes embodiments of the subject matter of the present disclosure with reference to the accompanying drawings. In each drawing, identical or corresponding parts are designated by the same reference numerals, and redundant explanations are appropriately simplified or omitted. Note that the subject matter of the present disclosure is not limited to the following embodiments, and any component of the embodiments may be modified or omitted within the scope of the gist of the present disclosure.

[0010] Embodiment 1 FIG. 1 is a side view showing the configuration of an elevator 1 according to the first embodiment.

[0011] An elevator 1 is applied to a building having multiple floors. FIG. 1 shows an example of the elevator 1 as seen from the side. A hoistway 2 for the elevator 1 is provided in the building. The hoistway 2 is a vertically long space that spans multiple floors. A pit 3 is provided at the bottom end of the hoistway 2. The elevator 1 includes a hoisting machine 4, a main rope 5, a car 6, a counterweight 7, and a control panel 8.

[0012] The hoisting machine 4 has a sheave and a motor. The sheave of the hoisting machine 4 is connected to the rotating shaft of the motor of the hoisting machine 4. The motor of the hoisting machine 4 is a device that generates a driving force to rotate the sheave of the hoisting machine 4.

[0013] The main ropes 5 are wound around a sheave of the hoisting machine 4. The main ropes 5 support the load of the car 6 by suspending the car 6 in the hoistway 2. The main ropes 5 support the load of the counterweight 7 by suspending the counterweight 7 in the hoistway 2. In this example, the main ropes 5 are wound around a return pulley 9. The main ropes 5 support the load of the car 6 on one side of the return pulley 9. The main ropes 5 support the load of the car 6 on the other side of the return pulley 9.

[0014] The car 6 and counterweight 7 run in opposite directions in the hoistway 2 as the main rope 5 moves with the rotation of the sheave of the hoisting machine 4. The car 6 is a device that transports passengers and the like between multiple floors by traveling up and down inside the hoistway 2. The counterweight 7 is a device that balances the load on the main rope 5 between the car 6 and a pulley such as a return sheave 9 around which the main rope 5 is wound.

[0015] The control panel 8 is a device that controls the operation of the elevator 1. The control panel 8 controls, for example, the running of the car 6. The control panel 8 is equipped with a function for acquiring the position of the car 6 in the elevator shaft 2, and the like.

[0016] The elevator 1 is equipped with a governor 10, a governor rope 11, and a tension wheel 12. The governor 10 is a device that suppresses excessive traveling speed of the car 6. The governor 10 has a pulley. The governor rope 11 is wound around the pulley of the governor 10. Both ends of the governor rope 11 are attached to the car 6. The governor rope 11 is wound around a tension wheel 12. The tension wheel 12 is a pulley that applies tension to the governor rope 11. The tension wheel 12 is installed in, for example, the pit 3. The pulley of the governor 10 rotates in conjunction with the movement of the car 6 via the governor rope 11 connected to the car 6. The governor 10 suppresses excessive traveling speed of the car 6 when the rotation speed of the pulley is excessive.

[0017] A long object inspection device 13 is applied to an elevator 1. The long object inspection device 13 is a device that detects abnormalities in long objects in a hoistway 2. The long object that the long object inspection device 13 detects abnormalities in is equipment that is long in one direction. When there is no abnormality, the longitudinal direction of the long object is parallel to the traveling direction of the car 6. In this example, the long object moves in the hoistway 2 as the elevator 1 operates. The long object is, for example, the main rope 5 or the governor rope 11. The long object may also be a counter rope (not shown) that compensates for the imbalance between the weight of the main rope 5 on the car 6 side and the weight of the counterweight 7 side of the main rope 5 caused by the movement of the main rope 5. The long object may also be a control cable that communicates electrical signals or supplies power. The long object may also be, for example, a strand rope, a belt, or a chain. The long object is wound around a sheave. The sheave is, for example, a pulley such as a sheave of a hoist 4, a pulley of a governor 10, a return pulley 9, a tension pulley 12, a deflector pulley, or a hoist pulley. The sheave includes, for example, a portion through which a long object passes or turns around. The long object inspection device 13 includes a camera 14 and an information processing device 15.

[0018] Camera 14 is equipped with a function for photographing the interior of hoistway 2. Camera 14 is an example of a photographing unit. Camera 14 photographs the interior of hoistway 2 so that the photographing range includes the long object and the trough on which the long object is wound. In this example, camera 14 is attached to the underside of the floor of car 6. Camera 14 photographs, for example, the area below car 6. Camera 14 may also be attached to the upper side of the ceiling of car 6. In this case, camera 14 photographs, for example, the area above car 6. Camera 14 may also be installed inside hoistway 2, in which case camera 14 photographs, for example, the area below hoistway 2.

[0019] The photographing unit may be a camera that photographs both above and below the car 6. The photographing unit may also include multiple cameras. In this case, the photographing unit may include a camera that photographs above the car 6 and a camera that photographs below the car 6. The long object inspection device 13 may also use a camera of an external device as the photographing unit. In other words, the information processing device 15 of the long object inspection device 13 may detect abnormalities in long objects using images photographed by a camera of an external device.

[0020] The information processing device 15 is a part that is responsible for processing information regarding the detection of abnormalities in long objects. The information processing device 15 is connected to the camera 14 so as to be able to acquire images captured by the camera 14. The information processing device 15 is provided, for example, on the top of the car 6.

[0021] FIG. 2 is a block diagram showing the functions of the long object inspection device 13 according to the first embodiment.

[0022] The information processing device 15 includes an image processing unit 16 and a determination unit 17. The image processing unit 16 is a unit equipped with a function of performing image processing to detect long objects from images captured by the camera 14. The image processing unit 16 is equipped with a function of acquiring images captured by the camera 14. The determination unit 17 is a unit equipped with a function of determining whether or not there is an abnormality in the long object based on the state of the long object detected by the image processing unit 16. The determination unit 17 is equipped with a function of outputting the determination result of whether or not there is an abnormality to the control panel 8. Image processing unit 16The detector includes a grooved wheel detector 18 and a long object detector 19.

[0023] The trough detection unit 18 has a function of performing image processing on the image captured by the camera 14 to detect the trough portion around which the long object is wound.

[0024] The trough detection unit 18 detects trough parts using, for example, a template matching technique. The trough detection unit 18 calculates the similarity of each part of the image captured by the camera 14 by comparing it with a preset template image that represents a correct image of a trough. At this time, the trough detection unit 18 detects, for example, a part where the calculated similarity is equal to or greater than a preset similarity threshold as a trough part.

[0025] The trough detection unit 18 may learn the feature amounts of the image by a machine learning technique and identify the trough portion by a classifier using the feature amounts. The trough detection unit 18 may use, for example, HOG (Histogram of Oriented Gradients) or SIFT (Scale Invariant Feature Transform) as the feature amounts. The trough detection unit 18 may use, for example, SVM (Support Vector Machine) as the classifier. The trough detection unit 18 may identify the trough portion by a deep learning technique.

[0026] The trough detection unit 18 may also detect a marker attached to the trough and detect the trough based on the marker detection result. Markers attached to the trough include, for example, stickers that display a preset color or patterned code. The trough detection unit 18 may narrow down the area including the trough portion on the image captured by the camera 14 based on the type or installation information of the elevator 1. In this case, the trough detection unit 18 performs image processing to detect the trough portion in the narrowed down area. The information on the type of the elevator 1 includes, for example, information such as the model or model number of the elevator 1. The installation information of the elevator 1 includes information such as the installation position of the trough. The type or installation information of the elevator 1 may be preset in the trough detection unit 18, or may be acquired by the trough detection unit 18 from the control panel 8 of the elevator 1, etc.

[0027] The long object detection unit 19 refers to the position of the groove portion detected by the groove detection unit 18 and performs image processing to detect the long object portion in the image captured by the camera 14.

[0028] The long object detection unit 19 detects the portions of the long object by, for example, an edge detection technique. For example, the long object detection unit 19 detects, among the detected edges, an edge that passes through the groove portion as the portion of the long object. The long object detection unit 19 identifies the portion of the long object by using an edge detector to identify a linear object that extends from the groove portion detected by the groove detection unit 18 as its starting point. The long object extends linearly in the traveling direction of the car 6, for example, from the groove portion as its starting point.

[0029] The long object detection unit 19 may detect portions of a long object using a local similarity determination method. For example, the long object detection unit 19 starts from the trough portion detected by the trough detection unit 18 and sequentially extracts local images of a predetermined size from the images captured by the camera 14 along the running direction of the car 6. The long object detection unit 19 calculates the similarity between each portion of a pair of adjacent local images among the extracted local images. Here, a pair of adjacent local images may be, for example, a pair of adjacent local images adjacent to each other in the running direction of the car 6, or a pair of local images whose distance in the running direction of the car 6 on the images captured by the camera 14 is shorter than a predetermined distance. The long object detection unit 19 detects portions of a long object by preferentially tracking portions of adjacent local images among the extracted local images that have a higher similarity. As a method for detecting portions of a long object sequentially from the trough portion in this way, the long object detection unit 19 may use a tracking filter used in image recognition processing, etc. The long object detection unit 19 detects the portion of the long object by tracking the portion of the long object from the position of the groove wheel using, for example, a particle filter or a Kalman filter.

[0030] The determination unit 17 determines whether or not there is an abnormality in the long object based on the state of the long object detected by the long object detection unit 19. The determination unit 17 determines whether or not there is an abnormality in the long object based on, for example, the position, orientation, or shape of the detected long object. The determination unit 17 outputs the determination result regarding whether or not there is an abnormality in the long object to, for example, the control panel 8.

[0031] The determination unit 17 may calculate the reliability of the detection of a long object by the long object detection unit 19. For example, when the long object detection unit 19 detects a long object using an edge detector, similarity between local images, a tracking filter, or the like, the determination unit 17 calculates the reliability of the detection based on edge strength or likelihood, etc. The determination unit 17 outputs information about the calculated reliability to the control panel 8, for example, together with the determination result.

[0032] The determination unit 17 may also calculate the position in the hoistway 2 of the location where an abnormality has been detected for the long object. The determination unit 17 may calculate the position of the location in the hoistway 2 based on the position of the location where the abnormality has been detected on the image. At this time, the determination unit 17 may calculate the position of the location in the hoistway 2, for example, using the part of the trough detected by the trough detection unit 18 as a reference. At this time, the determination unit 17 may also use information on the position of the car 6 when the image was captured. The determination unit 17 outputs the information on the calculated position of the location where the abnormality has been detected to the control panel 8, for example, together with the determination result.

[0033] The control panel 8 may control the operation of the elevator 1 according to the judgment result input from the judgment unit 17. For example, when the control panel 8 receives a judgment result indicating no abnormality, the control panel 8 continues the operation of the elevator 1 without stopping it. On the other hand, when the control panel 8 receives a judgment result indicating an abnormality, the control panel 8 stops the operation of the elevator 1. The control panel 8 may perform processing according to the reliability of the judgment result. For example, when the control panel 8 receives a judgment result indicating no abnormality with a reliability higher than a preset threshold, the control panel 8 continues the operation of the elevator 1 without requiring confirmation by a person such as a maintenance worker. If the elevator 1 is stopped at this time, the control panel 8 may resume the operation of the elevator 1 without requiring confirmation by a person such as a maintenance worker. On the other hand, when the control panel 8 receives a judgment result indicating no abnormality with a reliability lower than a preset threshold, the control panel 8 may notify a maintenance worker or the like of the judgment result. In this case, the control panel 8 may stop the operation of the elevator 1 until a person such as a maintenance worker visits the site to check the condition of the long object. The control panel 8 may notify the reliability calculated by the judgment unit 17 or the location of the abnormality detection point along with the judgment result.

[0034] FIG. 3 is a diagram showing an example of an image captured by the camera 14 according to the first embodiment.

[0035] In this example, the long object inspection device 13 detects an abnormality in a governor rope 11, which is a long object. The long object inspection device 13 detects a tension wheel 12, which is an example of a groove wheel, and then detects an abnormality in the governor rope 11. In Figure 3, illustrations of structures other than the governor rope 11 and tension wheel 12 are omitted.

[0036] The long object inspection device 13 may also detect abnormalities in other pairs of long objects and sheaves. The pairs of long objects and sheaves that the long object inspection device 13 detects may be, for example, a pair of the governor rope 11 and the pulley of the governor 10, a pair of the main rope 5 and the sheave of the hoist 4, and a pair of the balancing rope and the pulley around which it is wound.

[0037] Next, an example of detection of abnormality in a long object by the long object inspection device 13 will be described with reference to FIGS. In this example, the long object inspection device 13 detects an abnormality in a long object when the governor rope 11 wound around the tension pulley 12 gets caught on a structure in the hoistway 2 or the car 6. The structure in the hoistway 2 includes, for example, the housing, support, frame, beam, column, or bracket of equipment in the hoistway 2.

[0038] FIG. 4 is a side view showing a state in which an abnormality occurs in the governor rope 11 of the elevator 1 according to the first embodiment. Figure 1 shows the elevator 1 in a state where there is no abnormality in the governor rope 11, while Figure 4 shows the elevator 1 in a state where the governor rope 11 is caught on a structure within the elevator shaft 2.

[0039] FIG. 5 is a diagram showing an example of an image captured by the camera 14 when an abnormality occurs in the governor rope 11 of the elevator 1 according to the first embodiment. Fig. 3 shows an example of an image taken when there is no abnormality in the governor rope 11, whereas Fig. 5 shows an example of an image taken when the governor rope 11 is caught on a structure inside the hoistway 2. Note that the structure inside the hoistway 2 is not shown.

[0040] FIG. 6 is a diagram showing a location where an abnormality has occurred in the governor rope 11 of the elevator 1 according to the first embodiment. An enlarged view of the portion where the governor rope 11 is caught on a structure is shown in Figure 6. In Figure 6, the governor rope 11 is shown by a dashed line when it is not caught on a structure in the hoistway 2. The state of the governor rope 11, such as its direction and position when it is not caught, is acquired in advance when no abnormality occurs. The state of the governor rope 11, such as its direction and position when it is not caught, is acquired in advance, for example, when the elevator 1 is installed or during regular inspection of the elevator 1. The acquired state is stored in the long object inspection device 13.

[0041] When the governor rope 11 is caught on a structure, the direction of the governor rope 11 extending from the tension pulley 12 forms an angle D with the direction of the governor rope 11 when it is not caught. The determination unit 17 calculates this angle D based on, for example, the detection result of the long object detection unit 19. For example, when the calculated angle D is equal to or greater than a preset angle threshold, the determination unit 17 determines that the governor rope 11 is caught on a structure and that an abnormality has occurred.

[0042] FIG. 7 is a diagram showing a location where an abnormality has occurred in the governor rope 11 of the elevator 1 according to the first embodiment. 7, a plurality of sampling points are shown on the detected long object. The sampling points are, for example, sampled at equal intervals in the traveling direction of the car 6.

[0043] The determination unit 17 may detect an abnormality in a long object based on the linearity of the detected long object. For example, the determination unit 17 may calculate the direction of a line segment connecting adjacent sampling points in the running direction of the car 6, and determine that an abnormality has occurred by assuming that the governor rope 11 is caught on a structure when the angle between the directions of the adjacent line segments is equal to or greater than a preset threshold value.

[0044] The long object inspection device 13 may detect other abnormalities in the long object being caught. The long object inspection device 13 may detect abnormalities such as cuts on the long object, damage such as strand breakage, or poor tension. The long object inspection device 13 may detect cuts on the long object based on, for example, the continuity of the detected long object. The long object inspection device 13 may detect localized damage based on, for example, a change in image similarity along the longitudinal direction of the detected long object. The long object inspection device 13 may detect poor tension based on, for example, the linearity of the detected long object.

[0045] Next, an example of the operation of the long object inspection device 13 will be described with reference to FIG. FIG. 8 is a flowchart showing an example of the operation of the long object inspection device 13 according to the first embodiment. The processing in Fig. 8 is performed, for example, when an earthquake is detected in the elevator 1. The processing in Fig. 8 may be performed, for example, constantly while the elevator 1 is in operation, or may be performed at predetermined regular or irregular times. Furthermore, the processing in Fig. 8 may be performed based on an operation by a maintenance worker or a manager, or may be performed based on the occurrence of an event detected in the elevator 1.

[0046] In step S1, the camera 14 takes an image of the hoistway 2. Thereafter, the processing of the long object inspection device 13 proceeds to step S2.

[0047] In step S2, the grooved wheel detection unit 18 performs processing to detect grooved wheels from the image captured by the camera 14. The grooved wheel detection unit 18 outputs information about the grooved wheel portion detected in the image. Thereafter, the processing of the long object inspection device 13 proceeds to step S3.

[0048] In step S3, the long object detection unit 19 performs a process of detecting the portion of the long object by referring to the position of the groove portion output by the groove detection unit 18. The long object detection unit 19 outputs information on the portion of the long object detected on the image. Thereafter, the process of the long object inspection device 13 proceeds to step S4.

[0049] In step S4, the determination unit 17 determines whether there is an abnormality, such as the long object being caught, based on the state of the part of the long object output by the long object detection unit 19. If there is an abnormality, the processing of the long object inspection device 13 proceeds to step S5. On the other hand, if there is no abnormality, the processing of the long object inspection device 13 proceeds to step S6.

[0050] In step S5, the determination unit 17 outputs an abnormality signal to the control panel 8. Thereafter, the processing of the long object inspection device 13 ends.

[0051] In step S6, the determination unit 17 outputs a normal signal to the control panel 8. Thereafter, the processing of the long object inspection device 13 ends.

[0052] As described above, the long object inspection device 13 according to the first embodiment detects abnormalities in long objects wound around a trough in the hoistway 2. In the elevator 1, the camera 14 is provided on the car 6 traveling in the hoistway 2. The camera 14 captures images of either above or below the car 6, or both, so that the long object and the trough are included in the captured image. The long object inspection device 13 includes an image processing unit 16 and a determination unit 17. The image processing unit 16 performs image processing to detect the long object portion in the image captured by the camera 14. The determination unit 17 determines whether or not there is an abnormality in the long object based on the state of the long object detected by the image processing unit 16. The image processing unit 16 includes a trough detection unit 18 and a long object detection unit 19. The trough detection unit 18 performs image processing to detect the trough portion in the image captured by the camera 14. The long object detection unit 19 refers to the position of the groove portion detected by the groove detection unit 18 and performs image processing to detect the long object portion in the image captured by the camera 14.

[0053] With this configuration, the long object portion is detected while referring to the detected trough portion. This allows for more accurate detection of the long object to be determined for abnormality, even when a long object or linear texture not subject to determination is present in the hoistway 2. This allows for more accurate detection of abnormalities, such as a long object getting caught on a structure in the elevator 1. When a disaster such as an earthquake occurs in a location where the elevator 1 is installed, the elevator 1 may make an emergency stop. In such a case, in order to restore the elevator 1, it is important to check whether a long object in the hoistway 2, such as the governor rope 11, main rope 5, balancing rope, or control cable, is caught on a structure in the hoistway 2 or a part of the car 6. Meanwhile, having maintenance personnel visit and inspect numerous elevators 1 in an earthquake-hit area one after another requires significant time and cost. In contrast, detecting a long object getting caught using a camera 14 that photographs the inside of the hoistway 2 reduces the time and cost required for post-earthquake response. Furthermore, since abnormalities in long objects are detected more accurately, the time and cost required for post-earthquake response is more effectively reduced.

[0054] In addition, the judgment unit 17 judges that there is an abnormality in the long object if the angle between the direction of extension on the image of the long object detected by the long object detection unit 19 and the direction of extension on the image of the long object when it is not caught on any structure in the elevator shaft 2 is greater than or equal to a predetermined angle threshold.

[0055] With this configuration, the presence or absence of a snag is determined based on the difference from the state of the elongated object in a normal state, so that the criteria for determining the presence or absence of an abnormality can be set more clearly.

[0056] Furthermore, the trough detection unit 18 calculates the degree of similarity by comparing a portion of the image captured by the camera 14 with a preset template image. The trough detection unit 18 may detect a position having a degree of similarity equal to or greater than a preset similarity threshold as a trough portion. Furthermore, the trough detection unit 18 may detect the trough portion from the image captured by the camera 14 using a machine learning technique.

[0057] With this configuration, the trough detection unit 18 can identify the position of the trough from the image captured by the camera 14.

[0058] Furthermore, the trough detection unit 18 narrows down the area including the trough part on the image captured by the camera 14 based on the type or installation information of the elevator 1. The trough detection unit 18 performs image processing to detect the trough part in the narrowed down area.

[0059] With this configuration, the trough detection unit 18 can more accurately identify the position of the trough from the image captured by the camera 14.

[0060] Furthermore, the long object detection unit 19 detects the long object portion by using an edge detector to identify a linear object extending from the groove portion detected by the groove detection unit 18 as a starting point.

[0061] With this configuration, the long object detection unit 19 can more accurately detect, from among a plurality of long objects in the elevator shaft 2, a long object that is the object to be determined for the presence or absence of an abnormality.

[0062] Furthermore, the long object detection unit 19 uses the grooved wheel portion detected by the grooved wheel detection unit 18 as a starting point and sequentially extracts local images of a preset size along the running direction of the car 6. The long object detection unit 19 detects the portion of the long object by sequentially tracking, with priority given to portions of the extracted local images that have a higher degree of similarity between adjacent local images.

[0063] With this configuration, the long object detection unit 19 can detect long objects by tracing them in order starting from the groove wheel, even when local edge detection is not possible due to the influence of external light or lighting.

[0064] Furthermore, the determination unit 17 calculates the reliability of the detection of the long object by the long object detection unit 19. The determination unit 17 outputs the calculated reliability together with the determination result of an abnormality in the long object.

[0065] With this configuration, a maintenance person or the like who receives the judgment result from the judgment unit 17 can determine the order of recovery measures based on the output reliability.

[0066] Furthermore, the determination unit 17 calculates the position in the hoistway 2 of the point where the abnormality of the long object is detected. The determination unit 17 outputs the calculated position in the hoistway 2 together with the determination result of the abnormality of the long object.

[0067] With this configuration, a maintenance worker or the like who receives the determination result from the determining unit 17 can quickly ascertain the location in the hoistway 2 of the part that needs restoration and perform the work.

[0068] Next, an example of the hardware configuration of the long object inspection device 13 will be described with reference to FIG. FIG. 9 is a hardware configuration diagram of the main part of the long object inspection device 13 according to the first embodiment.

[0069] Each function of the long object inspection device 13 can be realized by a processing circuit. The processing circuit includes at least one processor 100a and at least one memory 100b. The processing circuit may include at least one dedicated hardware 200 in addition to or in place of the processor 100a and the memory 100b.

[0070] When the processing circuit includes a processor 100a and a memory 100b, each function of the long object inspection device 13 is realized by software, firmware, or a combination of software and firmware. At least one of the software and firmware is written as a program. The program is stored in the memory 100b. The processor 100a realizes each function of the long object inspection device 13 by reading and executing the program stored in the memory 100b.

[0071] The processor 100a is also called a CPU (Central Processing Unit), processing device, arithmetic unit, microprocessor, microcomputer, or DSP. The memory 100b is configured by, for example, a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM, or EEPROM.

[0072] Where the processing circuitry comprises dedicated hardware 200, the processing circuitry may be implemented, for example, as a single circuit, multiple circuits, a programmed processor, parallel programmed processors, an ASIC, an FPGA, or a combination thereof.

[0073] Each function of the long object inspection device 13 can be realized by a processing circuit. Alternatively, each function of the long object inspection device 13 can be realized collectively by a processing circuit. Some of the functions of the long object inspection device 13 may be realized by dedicated hardware 200, and the other parts may be realized by software or firmware. In this way, the processing circuit realizes each function of the long object inspection device 13 by dedicated hardware 200, software, firmware, or a combination of these.

[0074] Embodiment 2 In the second embodiment, differences from the example disclosed in the first embodiment will be described in particular detail. For features not described in the second embodiment, any of the features of the example disclosed in the first embodiment may be adopted.

[0075] FIG. 10 is a block diagram showing the functions of a long object inspection device 13 according to the second embodiment.

[0076] The image processing unit 16 includes a panoramic image generating unit 20. The panoramic image generating unit 20 is a unit equipped with a function for generating a panoramic image using images sequentially captured by the camera 14 while the car 6 is ascending or descending. The panoramic image generating unit 20 generates a panoramic image by cutting out and stitching together portions of the sequentially captured images. The panoramic image generating unit 20 may generate a panoramic image by adding an image of the pit 3 captured by the camera 14 to the stitched images. In this example, the panoramic image is a single image that shows the entire elevator shaft 2. Note that, for example, if there is no groove around which the long object to be determined is wound in the pit 3, the panoramic image generating unit 20 may generate a panoramic image without adding an image of the pit 3.

[0077] The trough detection unit 18 detects the trough portion by using the panoramic image generated by the panoramic image generation unit 20. Furthermore, the long object detection unit 19 detects the long object portion by using the panoramic image generated by the panoramic image generation unit 20 and referring to the detected trough portion.

[0078] FIG. 11 is a diagram showing an example of a panoramic image of the elevator shaft 2 according to the second embodiment.

[0079] 11, the governor rope 11 in the image of the pit 3 is indicated by the reference symbol 11a. Also, the governor rope 11 in the portion where the images of the hoistway 2 are joined together is indicated by the reference symbol 11b.

[0080] In the panoramic image generated in this way, the governor rope 11, which is a long object, is shown as being connected throughout the entire hoistway 2. Therefore, the long object detection unit 19 can detect the governor rope 11 throughout the entire hoistway 2, starting from the tension pulley 12 detected by the groove detection unit 18. Long object inspection device 13 The image processing unit 16 may exclude the portion where the images of the elevator shaft 2 are joined together and the boundary of the image of the pit 3 from the targets for determining whether there is an abnormality in the long object.

[0081] As described above, the second embodiment Long object inspection device 13 The image processing unit 16 includes a panoramic image generating unit 20. The panoramic image generating unit 20 generates a panoramic image of the elevator shaft 2 spanning the traveling direction of the car 6 by stitching together at least a portion of the images sequentially taken by the camera 14 along the traveling direction of the car 6. The trough detection unit 18 performs image processing to detect trough portions in the panoramic image generated by the panoramic image generating unit 20. The long object detection unit 19 refers to the positions of the trough portions detected by the trough detection unit 18 and performs image processing to detect long object portions in the panoramic image generated by the panoramic image generating unit 20. Furthermore, the panoramic image generator 20 generates a panoramic image so as to add an image of the pit 3 at the bottom end of the elevator shaft 2.

[0082] With this configuration, the image processing unit 16 can process a single still image as a panoramic image. This reduces the calculation load on the image processing unit 16 and the amount of memory used to store the image. Furthermore, long objects without abnormalities are displayed in a straight line along the running direction of the car 6 on the panoramic image. This allows for clearer criteria for determining whether an abnormality, such as a snag, exists. [Industrial Applicability]

[0083] The long object inspection device according to the present disclosure can be applied to elevators. [Explanation of symbols]

[0084] REFERENCE SIGNS LIST 1 elevator, 2 hoistway, 3 pit, 4 hoist, 5 main rope, 6 car, 7 counterweight, 8 control panel, 9 return wheel, 10 governor, 11 governor rope, 12 tension wheel, 13 long object inspection device, 14 camera, 15 information processing device, 16 image processing unit, 17 determination unit, 18 groove wheel detection unit, 19 long object detection unit, 20 panoramic image generation unit, 100a processor, 100b memory, 200 dedicated hardware

Claims

1. This is a long object inspection device that detects abnormalities in long objects wound around a sheave in the elevator car's hoistway. an imaging unit that images one or both of the upper and lower sides of the car so that the elongated object and the grooved wheel are included in the imaging range; an image processing unit that performs image processing to detect the portion of the elongated object in the image captured by the imaging unit; a determination unit that determines whether or not there is an abnormality in the elongated object based on the state of the elongated object detected by the image processing unit; Equipped with The image processing unit a grooved wheel detection unit that performs image processing to detect the grooved wheel portion in the image captured by the imaging unit; a long object detection unit that performs image processing to detect the long object portion in the image captured by the imaging unit by referring to the position of the groove portion detected by the groove detection unit; A long object inspection device comprising:

2. This is a long object inspection device that detects abnormalities in long objects wound around a sheave in the elevator car's hoistway. an image processing unit that performs image processing to detect a portion of the elongated object in an image captured by an image capturing unit that captures an image of one or both of the upper and lower sides of the car so that the elongated object and the grooved wheel are included in the image capturing range; a determination unit that determines whether or not there is an abnormality in the elongated object based on the state of the elongated object detected by the image processing unit; Equipped with The image processing unit a grooved wheel detection unit that performs image processing to detect the grooved wheel portion in the image captured by the imaging unit; a long object detection unit that performs image processing to detect the long object portion in the image captured by the imaging unit by referring to the position of the groove portion detected by the groove detection unit; A long object inspection device comprising:

3. The determination unit determines that there is an abnormality in the long object when an angle formed between a direction in which the long object detected by the long object detection unit extends on an image captured by the photographing unit and a direction in which the long object extends on the image when not caught on a structure in the elevator shaft is equal to or greater than a preset angle threshold. The long object inspection device according to claim 1 or 2.

4. The trough wheel detection unit calculates a similarity between a portion of the image captured by the imaging unit and a predetermined template image, and detects a position having a similarity equal to or greater than a predetermined similarity threshold as a portion of the trough wheel. The long object inspection device according to claim 1 or 2.

5. The grooved wheel detection unit detects a grooved wheel portion from the image captured by the imaging unit using a machine learning technique. The long object inspection device according to claim 1 or 2.

6. The trough detection unit narrows down an area including the trough portion on the image captured by the photographing unit based on the type or installation information of the elevator, and performs image processing to detect the trough portion in the narrowed down area. The long object inspection device according to claim 1 or 2.

7. The elongated object detection unit detects the elongated object portion by identifying a linear object extending from the groove portion detected by the groove portion detection unit as a starting point using an edge detector. The long object inspection device according to claim 1 or 2.

8. the long object detection unit sequentially extracts local images of a preset size along the running direction of the car, starting from the trough portion detected by the trough detection unit, and sequentially tracks portions of the extracted local images that have a higher degree of similarity between adjacent local images with higher priority, thereby detecting the long object portion. The long object inspection device according to claim 1 or 2.

9. the determination unit calculates a reliability of the detection of the long object by the long object detection unit, and outputs the calculated reliability together with a determination result of an abnormality in the long object. The long object inspection device according to claim 1 or 2.

10. The determination unit calculates a position in the hoistway of a point where an abnormality is detected for the long object, and outputs the calculated position in the hoistway together with a determination result of the abnormality of the long object. The long object inspection device according to claim 1 or 2.

11. The photographing unit is provided in the car, The image processing unit a panoramic image generating unit that generates a panoramic image of the elevator shaft in the running direction of the car by stitching together at least a portion of the images sequentially captured by the photographing unit along the running direction of the car; Equipped with the grooved wheel detection unit performs image processing to detect the grooved wheel portion in the panoramic image generated by the panoramic image generation unit, the elongated object detection unit refers to the position of the grooved wheel portion detected by the grooved wheel detection unit, and performs image processing to detect the elongated object portion in the panoramic image generated by the panoramic image generation unit. The long object inspection device according to claim 1 or 2.

12. the panoramic image generator generates a panoramic image so as to add a pit image of a lower end of the elevator shaft. The long object inspection device according to claim 11.

Images