Diagnostic system for wire rope fracture, diagnostic method, and diagnostic program

Abstract

To provide a wire rope fracture diagnostic system, a diagnostic method, and a diagnostic program that reduce user burden by relying solely on image data.SOLUTION: In one embodiment of the present invention, a wire rope fracture diagnostic system is provided, configured to perform the following steps: a receiving step receives image data targeted for a wire rope acquired through an image pickup device; a removal step removes the background other than the wire rope in the image data; an extraction step extracts bare line parts of the wire rope in the image data the background of which has been removed; an input step inputs only the extracted bare line parts into a model; a difference step generates a difference image between a restoration image output from the model and the image consisting of only the bare line parts; a determination step determines an abnormality degree of the wire rope from the difference image; and a transmission step notifies the user of the determination result.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a wire rope breakage diagnosis system, a diagnosis method, and a diagnosis program. [Background technology]

[0002] Wire ropes combine high levels of strength and flexibility, and are used widely in structures such as bridges, elevators, and cargo handling machines such as cranes. However, especially when used with cranes, the wire rope is bent along the sheaves, which causes the radius of curvature to change between the inside and outside, causing the wires that make up the wire rope to rub against each other, resulting in wire breakage, known as fatigue failure. When a wire breaks, the strength of the entire wire rope decreases, and the wire rope itself may eventually break. Therefore, a simple method for early detection of wire breakage is desired.

[0003] Patent Document 1 describes the following wire rope inspection method. A magnetic field is applied to the wire rope to excite the wire rope, changes in the magnetic flux of the excited wire rope are detected, and a previously acquired abnormal waveform is superimposed on the detected signal waveform. The wire rope condition is inspected based on whether or not this superimposed waveform exceeds a predetermined range corresponding to noise in the detected signal waveform. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2023-027643 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the method disclosed in Patent Document 1, if the wire itself is not magnetized prior to measuring the magnetic flux, the signal is too weak to be measured, and is highly susceptible to the effects of noise. Furthermore, the need to magnetize the wire itself prior to measurement requires a lot of time for preparation, and the method does not allow for a simple and quick determination of whether or not a wire is broken.

[0006] In consideration of the above circumstances, the present invention aims to provide a wire rope break diagnosis system and a wire rope break diagnosis method that place a low burden on the user, by using only image data of the wire rope obtained by an image collection device to remove background images, extract only the wire portions, input a model, construct a restored image, create a difference image, and determine the degree of abnormality. [Means for solving the problem]

[0007] The first aspect of the present invention is A wire rope breakage diagnostic system, comprising: It is configured to perform the following steps: In the receiving step, image data of the wire rope obtained by the image capturing device is received, In the removing step, a background other than the wire rope is removed from the image data; In the extraction step, only the wire portion of the wire rope is extracted from the image data from which the background has been removed; In the input step, only the extracted wire portion is input into a model; In the subtraction step, a subtraction image is created between a restored image output from the model and an image of only the wire portion; In the determination step, a degree of abnormality of the wire rope is determined from the difference image; In the transmission step, the wire rope breakage diagnosis system notifies the user of the judgment result.

[0008] A second aspect of the present invention is A wire rope breakage diagnosis method, comprising: It is configured to perform the following steps: In the receiving step, image data of the wire rope obtained by the image capturing device is received, In the removing step, a background other than the wire rope is removed from the image data; In the extraction step, only the wire portion of the wire rope is extracted from the image data from which the background has been removed; In the input step, only the extracted wire portion is input into a model; In the subtraction step, a subtraction image is created between a restored image output from the model and an image of only the wire portion; In the determination step, a degree of abnormality of the wire rope is determined from the difference image; In the transmission step, the wire rope breakage diagnosis method notifies the user of the judgment result.

[0009] According to these, it is possible to easily diagnose the presence or absence of abnormalities such as broken wires in the wire rope by simply receiving image data of the wire rope obtained by an image capturing device, without measuring the magnetization and magnetic flux density of the wire rope. [Brief description of the drawings]

[0010] [Figure 1] 1 is an overall view of a wire rope breakage diagnosis system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a block diagram showing a hardware configuration of the server 3. [Diagram 3] 2 is a block diagram showing functions realized by a server 3. FIG. [Figure 4] FIG. 2 is a flow chart of the present embodiment. [Diagram 5] 1 is an example pixel classification model for background removal. [Figure 6] 13 is a diagram showing examples of an actual image, a restored image, and a difference image. [Figure 7] 13A and 13B are diagrams illustrating an example of a difference image heat map and an example of an abnormality degree determination diagram. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. Various characteristic features shown in the following embodiments can be combined with each other.

[0012] By the way, the program for realizing the software appearing in this embodiment is a computer program. The program may be provided as a non-transitory computer-readable recording medium, or may be provided so as to be downloadable from an external server, or may be provided so that the program is started on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0013] In the present embodiment, the term "unit" refers to, for example, hardware implemented by a circuit in a broad sense. hardware resources and software that can be specifically realized by these hardware resources In the present embodiment, various information is handled, and this information is represented by, for example, physical values ​​of signal values ​​representing voltage and current, high and low signal values ​​as a binary bit set consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculation can be performed on a circuit in a broad sense.

[0014] In addition, a circuit in the broad sense is a circuit realized by at least appropriately combining a circuit, circuitry, a processor, and a memory, etc. In other words, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0015] 1. Hardware Configuration In the first section, the hardware configuration of this embodiment will be described.

[0016] 1.1 Wire rope breakage diagnosis system 1 FIG. 1 is an overall view of a wire rope break diagnosis system 1 according to this embodiment. The wire rope break diagnosis system 1 includes an image pickup device 2 (e.g., image pickup devices 2-1, 2-2, . . . , 2-n) and a server 3, which are connected via a network. These components will be further described. Here, the system exemplified by the wire rope break diagnosis system 1 is made up of one or more devices or components.

[0017] 1.2 Image capture device 2 The image capturing device 2 targets the wire rope. Since most of the wire rope extends in a straight line when in use, an image capturing device that scans multiple wire ropes in that straight line is efficient and desirable. In addition, it is preferable that the image capturing device 2 is provided with an exposure device in consideration of the case where imaging is performed while scanning at high speed. The image capturing device 2 is, for example, a digital camera or a digital video camera. In this embodiment, since only the wire part is model input, a 4K (pixel number: 3840 x 2160) camera is preferable to facilitate this. A still image is preferable to reduce the amount of data transmitted to the server 3, and a video is preferable when the actual image is also to be viewed. In the case of a video, 30 FPS (frames per second) or less is preferable. More preferably, it is 5 FPS or less, specifically, for example, 5, 10, 15, 20, 25, or 30 FPS, and may be within a range between any two of the numerical values ​​exemplified here. This is because lowering the frame rate can suppress the amount of data transmitted to the server 3, and can reduce the load required for model input, etc.

[0018] The image capturing device 2 is connected to a communication unit 31 in a server 3 (described later) via a network. The image capturing device 2 is configured to transfer all captured image data to the server 3.

[0019] Moreover, an imaging device capable of measuring not only visible light but also ultraviolet or infrared light bands that cannot be perceived by humans may be used as the imaging device 2. By using such an imaging device 2, the wire rope breakage diagnosis system 1 according to the present embodiment can be implemented even in a dark field.

[0020] 1.3 Server 3 The server 3 is linked to all the image capturing devices 2. Considering the accumulation of teaching data in machine learning described later, a server with a larger data capacity is preferable, but as long as it can accommodate the amount of data transmitted from the image capturing devices 2, it is not particularly specified.

[0021] 2 is a block diagram showing a hardware configuration of the server 3. The server 3 has a communication unit 31, a storage unit 32, and a control unit 33, and these components are electrically connected via a communication bus 30 inside the server 3. Each component will be further described.

[0022] The communication unit 31 is preferably a wired communication means such as USB, IEEE1394, Thunderbolt, wired LAN network communication, etc., but may also include wireless LAN network communication, mobile communication such as 3G / LTE / 5G, Bluetooth (registered trademark) communication, etc. as necessary. In other words, it is more preferable to implement it as a collection of multiple communication means. The server 3 receives image data from the image capturing device 2 via the communication unit 31. Details will be described later.

[0023] The storage unit 32 temporarily stores image data received from the image capturing device 2 described above. This can be implemented, for example, as a storage device such as a solid state drive (SSD) that stores various programs related to the server 3 executed by the control unit 33, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the calculation of the program. It may also be a combination of these. In particular, when the "notification" to the user described in step S108 of FIG. 4 is given and the teacher data described below is to be saved, the image data is stored until "save to server 3" is selected.

[0024] The control unit 33 performs processing and control of the overall operation related to the image collecting device 3. The control unit 33 is, for example, a central processing unit (CPU) not shown. The control unit 33 reads out a predetermined program stored in the storage unit 32, and judges the degree of abnormality of the wire rope from the image obtained from the image capturing device 2, and performs a breakage diagnosis of the wire rope. That is, information processing by the software stored in the storage unit 32 can be specifically realized by the control unit 33, which is an example of hardware, and executed as a functional unit included in the control unit 33. These will be described in more detail in the next section. Note that the control unit 33 is not limited to being single, and may be implemented to have multiple control units 33 for each function. Also, a combination of them may be used.

[0025] 2. Functional configuration In this section, the functional configuration of the present embodiment will be described. As described above, information processing by the software stored in the storage unit 32 is specifically realized by the control unit 33, which is an example of hardware, and can be executed as a functional unit included in the control unit 33.

[0026] 3 is a block diagram showing functions realized by the server 3 (control unit 33). Specifically, the server 3 (control unit 33) includes a reception unit 331, a removal unit 332, an extraction unit 333, an input unit 334, a difference image creation unit 335, a determination unit 336, and a transmission unit 337.

[0027] The receiving unit 331 is configured to receive various types of information. For example, the receiving unit 331 receives all image data sent from the image capturing device 2. As described above, the image data may be a still image or a video.

[0028] The removal unit 332 is configured to remove background other than the wire rope, which is the target object, from all image data transmitted from the image capturing device 2. For example, the removal unit 332 calls a predetermined program stored in the storage unit 32, extracts features (edge ​​extraction) from all image data transmitted from the image capturing device 2, performs analysis, and recognizes the wire rope, which is the target object, and recognizes the rest as the background.

[0029] The above recognition is preferably performed by machine learning based on teacher data stored in the server 3 from time to time, and may include an algorithm for extracting the feature amount as necessary. The machine learning algorithm is not particularly limited, and k-nearest neighbor method, logistic regression, support vector machine, neural network, topic model, Gaussian mixture model, etc. may be appropriately adopted.

[0030] Furthermore, as a pixel classification model for background removal, a convolutional neural network with a pyramid pooling structure is more preferable because, as shown in Fig. 5, it is easier to obtain content information around the wire rope, and therefore the wire rope region can be classified faithfully.

[0031] The extraction unit 333 is configured to extract only the bare wire portion of the wire rope from the image from which the background has been removed by the removal unit 332. Since most parts of a wire rope extend in a straight line when in use, it is easy to apply a line detector to the bare wire portion. When applying the line detector, color clustering may be used in combination to divide the image into superpixel units (divide the image into small areas that reflect the positional relationship of pixels with similar colors). In this way, only the bare wire area can be extracted. Details will be described later.

[0032] The input unit 334 is configured to input to the model only the wire region portion extracted by the extraction unit 333. The model is a type of convolutional neural network represented by U-net AutoEncoder, and is preferably a model that uses images of unbroken wires (normal images) as trained teacher data.

[0033] By using the above model, a restored image is output. A difference image creation unit 335 creates a difference image between the restored image and the actual image from which the background has been removed by the removal unit 332. A judgment unit 336 uses this difference image to diagnose the degree of abnormality of the wire rope.

[0034] 7 shows a thermograph showing the degree of deviation from a so-called normal distribution for the difference image and a diagram showing the degree of abnormality calculated from the Mahalanobis distance (not the Euclidean distance). Specifically, the determination unit 336 is configured to determine a certain threshold value for the Mahalanobis distance and diagnose an abnormality when the value is exceeded.

[0035] After the above diagnosis, the transmission unit 337 is configured to report the diagnosis result to the user. At that time, it is preferable that the above difference image, the image with the background removed, etc. are sent to the user, and the wire rope breakage diagnosis system 1 does not end until the user checks it. In that case, the user can check the overall confirmation including the still images and videos sent from the image capture device 2 to the server 3.

[0036] 3. Wire rope breakage diagnosis method In this section, a diagnosis method of the wire rope break diagnosis system 1 described above will be described. This wire rope break diagnosis method includes the following steps: In the reception step, image data of the wire rope obtained by an image capturing device is received. In the removal step, background other than the wire rope is removed from the image data. In the extraction step, only the bare wire portion of the wire rope is extracted from the image data from which the background has been removed. In the input step, only the extracted bare wire portion is input to a model. In the difference step, a difference image is created between the restored image output from the model and the image data from which the background has been removed. In the judgment step, the degree of abnormality of the wire rope is judged from the difference image. In the transmission step, the judgment result is notified to a user.

[0037] 4 is a flowchart showing the flow of information processing executed by the wire rope breakage diagnosis system 1. Below, a case where a wire rope used in a loading machine such as a crane is the subject will be described along each flow of this flowchart.

[0038] When a user attempts to inspect or diagnose the condition (presence or absence of wire breakage, etc.) of a wire rope used in cargo handling machines such as container cranes, the image capturing device 2 is first started. The user may manually start and capture an image by operating the image capturing device 2. However, since the wires used in cranes are often installed in parallel and linear fashion at high altitudes, it is desirable from a safety standpoint that the image capturing device 2 be configured to scan and capture multiple wires while maintaining a certain distance from the wires. The server 3 is configured to accept all image data from the image capturing device 2 (step S101).

[0039] It is preferable that the image capturing device 2 is a 4K (pixel count: 3840 x 2160) camera. This is to prevent insufficient resolution, since only the wire region is extracted and input into the model as described above. However, since wire ropes used in loading and unloading machines such as container cranes move at high speeds, it is more preferable to make timely adjustments, such as limiting the shooting range and reducing the number of pixels to reduce the amount of data, taking into account the communication interface and transfer rate. Then, in the image obtained from the image capturing device 2, the wire rope and the background image are identified (step S102).

[0040] There are various methods for identifying in step S102. As described above, there are identification by machine learning using supervised data and identification using a convolutional neural network. In particular, by using a convolutional neural network such as the U-net Classifier, it is possible to directly learn the features, which can eliminate some of the work required for the user to extract features, thereby improving the accuracy of the wire rope breakage diagnosis system 1.

[0041] Furthermore, in the above convolutional neural network, it is more preferable to use a pyramid pooling structure as shown in Fig. 5. This improves the classification and identification of the wire rope region and the background. This improvement also improves the estimation of the wire diameter.

[0042] In wire ropes used in cargo handling machines such as cranes, the thickness can sometimes be reduced by 15%, which is the legal limit, before the entire rope wears out and a wire breaks. In such cases, it is useful to be able to improve the accuracy of estimating the wire diameter. In addition, it is possible that a wire may break inside the wire before the wire is thinned. Even in such a case, it is preferable to improve the accuracy of estimating the wire diameter.

[0043] The above-mentioned result of identifying the wire rope and the background image is used to remove the background image (step S103).From the background-removed image thus obtained (an image in which only the wire rope is included in the field of view), only the wire strand portion is further extracted (step S104).

[0044] In particular, wire ropes used in cargo handling machines such as cranes are coated with oil and grease during use. However, in the parts of the wire rope where wire breakage is likely to occur, the metal base is exposed due to wear, and a metallic luster is observed. For this reason, color clustering works effectively, making it easier to apply line detectors.

[0045] After determining the wire portion by clustering, the number of lines detected by the line detector for that wire portion is counted. A threshold is set for the number of lines, and only the portion exceeding the threshold is extracted as the straight line detection region.

[0046] One method for clustering by color is to divide the image into superpixel units (divide the image into small regions that reflect the positional relationship of pixels with similar colors). This allows for more accurate extraction of only the wire parts. There are various methods for superpixels themselves, but there is no particular limit to the method as long as it is image segmentation.

[0047] The clustering and line detector using the superpixels are possible measures to avoid the oil and grease parts applied to the wire rope, and contribute to improving the diagnostic accuracy without losing the features of the bare wire parts. As mentioned above, even for extracting only the bare wire parts, it is possible to use machine learning using supervised data or a convolutional neural network.

[0048] In this way, only the wire strand is input to the model (step S105). The model is one of the models used in convolutional neural networks, such as the U-net AutoEncoder, as described above. A model that is constructed by machine learning using images of unbroken wires (normal images) as training data is preferable.

[0049] After that, the restored image is output. The restored image is treated as a fake image, and the image of only the wire portion is treated as a real image, and each is visualized, and a difference image is created (step S106). The real image, restored image, and difference image shown on the right side of Fig. 6 were created using the example of a wire (only one wire) with only the wire region extracted, shown on the left side of Fig. 6.

[0050] The difference image diagram above is not visually intuitive for the user to judge, so a thermograph is created (left side of Fig. 7) showing the degree to which the difference image deviates from the so-called normal distribution, and the thermograph is also sent from the transmission unit 337 to the user together with the notification of the judgment result.

[0051] The right side of Fig. 7 shows the degree of anomaly calculated from the Mahalanobis distance (not the Euclidean distance). Specifically, a certain threshold value is set for the Mahalanobis distance, and if the value is exceeded, it is determined to be abnormal (step S107). The reason why it is preferable to use the Mahalanobis distance here is that it is possible to prevent so-called erroneous determination that an abnormality is determined when the deviation from the normal distribution exceeds a certain range but the deviation tendency is not particularly problematic.

[0052] In particular, the safety of wire ropes used in cargo handling machines such as container cranes must be guaranteed, but if the above-mentioned erroneous judgments occur frequently, containers may end up stranded along the coast. Therefore, an abnormality diagnosis is required to prevent the above-mentioned erroneous judgments as much as possible.

[0053] Then, as described above, the result of the judgment and the thermograph and an image of only the wire portion are sent to the user so that the user can finally decide whether to replace the wire rope or suspend the work (step S108). In order to prevent the user from having to look at and check all the data, it is preferable to limit the images to only those in which still images or videos corresponding to the portion exceeding the certain threshold can be confirmed. After the user receives the result of the judgment, the wire rope breakage diagnosis system 1 ends (step S109).

[0054] Preferably, the user not only receives the above-mentioned judgment result, but also selects all or a part of the image obtained by extracting only the wire region as training data for machine learning in the extraction unit 333. It is particularly preferable that the above-mentioned selection is made together with the notification of the abnormality degree judgment in step S108. Therefore, in practice, the wire rope breakage diagnosis system 1 ends after the user selects "save the training data in server 3" or "do not save in server 3" (step S109).

[0055] 4.Other The wire rope breakage diagnosis system 1 according to the present embodiment may adopt the following aspects: The wire rope breakage diagnosis system 1 judges the degree of abnormality in the judging step based on a Mahalanobis distance. It is known that there are various methods for determining the degree of abnormality using normal distribution, and the method is not particularly limited in this embodiment. However, in image diagnosis, even if the deviation from the normal distribution exceeds a certain range, it may not be a problem in consideration of the deviation tendency. Therefore, it is preferable to use the Mahalanobis distance instead of the Euclidean distance, since it is possible to perform the abnormality determination based on the above tendency.

[0056] In the removal step, the wire rope and the background are identified by image segmentation in the wire rope breakage diagnosis system 1. Although it is possible to identify the wire and the background by extracting (edge ​​extraction) features from all image data transmitted from the image capture device 2 and performing analysis, this identification is performed by creating an extraction algorithm by the researcher based on the researcher's knowledge. Therefore, by adopting image segmentation in deep learning, an optimal extraction algorithm is automatically generated, and it becomes possible to identify the wire and the background without arbitrariness and with little burden on the user. Therefore, it is more preferable.

[0057] This is a wire rope breakage diagnosis system 1 in which the image segmentation is performed based on pre-trained learned data. As mentioned above, image segmentation allows the creation of an optimal feature extraction algorithm, but it is preferable to have a large amount of so-called training data, which allows for more accurate identification of the wire and the background.

[0058] The image segmentation further includes a learning step, in which the learned data is generated or updated using the image data after the image segmentation as teaching data, in the wire rope break diagnosis system 1. By providing the learning step, the user does not need to prepare separate teaching data, and the data can be accumulated in the server 3. This is even more preferable since it further enables identification of the wire and the background.

[0059] In the extraction step, the wire portion is identified by a clustering algorithm in the wire rope breakage diagnosis system 1. As described in paragraph 0047 above, in extracting only the bare wire portion, it is possible to use machine learning using supervised data or a convolutional neural network, but this would incur costs for creating the supervised data. In particular, in the case of wire ropes used in cargo handling machines such as container cranes, the positional relationship between the image capture device 2 and the wire rope is almost constant, making it possible to obtain similar data, and therefore it is preferable to use the above-mentioned Superpixel-based clustering and line detector in combination to improve the accuracy of the wire rope breakage diagnosis system 1.

[0060] In the above embodiment, the configuration of the wire rope break diagnosis system 1 has been described, but a program for causing a computer to execute each step in the wire rope break diagnosis system 1 may be provided. Even when the operation of the wire rope breakage diagnosis system 1 has ended and the server 3 cannot be used, it is preferable that the user can diagnose wire breakage on an edge terminal or the like owned by the user himself / herself if there are images obtained from the image capturing device 2. This can also reduce the load on the server 3, which can be shared for other operations of cargo handling machines such as container cranes.

[0061] Finally, although various embodiments of the present invention have been described, these are presented as examples and are not intended to limit the scope of the invention. The novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. The embodiments and their modifications are included within the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0062] 1: Wire rope breakage diagnostic system 2: Image capture device 3: Server 30: Communication bus 31: Communications Department 32: Storage section 33: Control section 331: Reception 332:Removal section 333:Extraction part 334: Input section 335: Differential image creation unit 336: Judgment section 337: Transmission unit

Claims

1. A wire rope breakage diagnostic system, comprising: It is configured to perform the following steps: In the receiving step, image data of the wire rope obtained by the image capturing device is received, In the removing step, a background other than the wire rope is removed from the image data; In the extraction step, only the wire portion of the wire rope is extracted from the image data from which the background has been removed; In the input step, only the extracted wire portion is input into a model; In the subtraction step, a subtraction image is created between a restored image output from the model and an image of only the wire portion; In the determination step, a degree of abnormality of the wire rope is determined from the difference image; In the transmission step, the wire rope breakage diagnosis system notifies the user of the judgment result.

2. 2. The wire rope breakage diagnostic system according to claim 1, A wire rope breakage diagnosis system in which the degree of abnormality in the judgment step is determined based on Mahalanobis distance.

3. The wire rope breakage diagnosis system according to any one of claims 1 to 2, A wire rope breakage diagnosis system, wherein in the removal step, the wire rope and the background are identified by image segmentation.

4. 4. The wire rope breakage diagnostic system according to claim 3, A wire rope breakage diagnosis system, wherein the image segmentation is performed based on pre-trained learned data.

5. 5. The wire rope breakage diagnostic system according to claim 4, The image segmentation further comprises a learning step; A wire rope breakage detection system in which, in the learning step, the image data after the image segmentation is used as training data to generate or update the learned data.

6. The wire rope breakage diagnosis system according to any one of claims 1 to 2, A wire rope breakage diagnosis system in which, in the extraction step, the wire portion is identified using a clustering algorithm.

7. 4. The wire rope breakage diagnostic system according to claim 3, A wire rope breakage diagnosis system in which, in the extraction step, the wire portion is identified using a clustering algorithm.

8. 5. The wire rope breakage diagnostic system according to claim 4, A wire rope breakage diagnosis system in which, in the extraction step, the wire portion is identified using a clustering algorithm.

9. 6. The wire rope breakage diagnosis system according to claim 5, A wire rope breakage diagnosis system in which, in the extraction step, the wire portion is identified using a clustering algorithm.

10. A wire rope breakage diagnosis method, comprising: It is configured to perform the following steps: In the receiving step, image data of the wire rope obtained by the image capturing device is received, In the removing step, a background other than the wire rope is removed from the image data; In the extraction step, only the wire portion of the wire rope is extracted from the image data from which the background has been removed; In the input step, only the extracted wire portion is input into a model; In the subtraction step, a subtraction image is created between a restored image output from the model and an image of only the wire portion; In the determination step, a degree of abnormality of the wire rope is determined from the difference image; In the transmission step, the determination result is notified to a user.

11. The wire rope breakage diagnosis method according to claim 10, A wire rope breakage diagnosis method in which the degree of abnormality in the judgment step is determined based on Mahalanobis distance.

12. A wire rope breakage diagnosis method according to any one of claims 10 to 11, A wire rope breakage diagnosis method, in which in the removal step, the wire rope and the background are identified by image segmentation.

13. The wire rope breakage diagnosis method according to claim 12, A wire rope breakage diagnosis method, wherein the image segmentation is performed based on pre-trained learned data.

14. The wire rope breakage diagnosis method according to claim 13, The image segmentation further comprises a learning step; A wire rope breakage diagnosis method in which, in the learning step, the image data after the image segmentation is used as training data to generate or update the learned data.

15. A wire rope breakage diagnosis method according to any one of claims 10 to 11, A wire rope breakage diagnosis method, in which in the extraction step, the wire portion is identified using a clustering algorithm.

16. 13. A wire rope breakage diagnostic system according to claim 12, A wire rope breakage diagnosis method, in which in the extraction step, the wire portion is identified using a clustering algorithm.

17. 14. The wire rope breakage diagnostic system according to claim 13, A wire rope breakage diagnosis method, in which in the extraction step, the wire portion is identified using a clustering algorithm.

18. 15. A wire rope breakage diagnostic system according to claim 14, A wire rope breakage diagnosis method, in which in the extraction step, the wire portion is identified using a clustering algorithm.

19. A program, A program that causes a computer to execute each step of the wire rope breakage diagnosis method according to any one of claims 10 to 18.

Images