Method, device, equipment, medium and product for detecting bending state of vibration damping hammer

Abstract

The present disclosure provides a method, device, equipment, medium and product for detecting the bending state of an anti-vibration hammer, and relates to the technical field of power transmission line fault detection, and in particular to the anti-vibration hammer bending degree detection scenario in the technical field of anti-vibration hammer defect detection. The specific implementation scheme is: acquiring an anti-vibration hammer image; segmenting a first hammer body area and a second hammer body area in the anti-vibration hammer image; performing a straight line fitting on the first hammer body area to obtain a first fitting straight line, and performing a straight line fitting on the second hammer body area to obtain a second fitting straight line; calculating the angle between the first fitting straight line and the second fitting straight line, and determining the bending state of the anti-vibration hammer based on the angle. The present disclosure can more efficiently detect the bending state of the anti-vibration hammer.

Description

Technical Field

[0001] The present disclosure relates to the technical field of power transmission line fault detection, and in particular to a scenario of detecting the bending degree of an anti-vibration hammer in the technical field of anti-vibration hammer defect detection. Background Art

[0002] In recent years, the scale of power grids has gradually increased, and the maintenance of transmission lines has become particularly important. In order to reduce the vibration at the outlet of the overhead line clamp, anti-vibration hammers are installed on the overhead lines to consume the vibration energy transmitted to the overhead lines by the wind, reduce the vibration amplitude of the overhead lines, and reduce the damage to the overhead lines.

[0003] In practical applications, it is necessary to detect the state of the anti-vibration hammer, such as detecting the bending degree of the anti-vibration hammer, to prevent the anti-vibration hammer from failing. Summary of the invention

[0004] The present disclosure provides a method, device, equipment, medium and product for detecting the bending state of an anti-vibration hammer.

[0005] According to one aspect of the present disclosure, there is provided a method for detecting a state of an anti-vibration hammer, comprising:

[0006] Acquire an anti-vibration hammer image; segment a first hammer body region and a second hammer body region in the anti-vibration hammer image; perform straight line fitting on the first hammer body region to obtain a first fitting straight line, and perform straight line fitting on the second hammer body region to obtain a second fitting straight line; calculate an angle between the first fitting straight line and the second fitting straight line, and determine a bending state of the anti-vibration hammer based on the angle.

[0007] According to another aspect of the present disclosure, there is provided a device for detecting a state of an anti-vibration hammer, comprising:

[0008] An acquisition unit is used to acquire an anti-vibration hammer image; a segmentation unit is used to segment a first hammer body area and a second hammer body area in the anti-vibration hammer image; a fitting unit is used to perform straight line fitting on the first hammer body area to obtain a first fitting straight line, and perform straight line fitting on the second hammer body area to obtain a second fitting straight line; a calculation unit is used to calculate an angle between the first fitting straight line and the second fitting straight line, and determine a bending state of the anti-vibration hammer based on the angle.

[0009] According to another aspect of the present disclosure, there is provided an electronic device, including:

[0010] at least one processor; and

[0011] a memory communicatively connected to the at least one processor; wherein,

[0012] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor so that the at least one processor can perform the above method.

[0013] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the above method.

[0014] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein the computer program implements the above method when executed by a processor.

[0015] The method for detecting the bending state of the anti-vibration hammer provided by the present disclosure can detect the bending state of the anti-vibration hammer more efficiently.

[0016] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0018] Figure 1 is a flow chart of a method for detecting a bending state of an anti-vibration hammer according to an exemplary embodiment of the present disclosure;

[0019] Figure 2A A schematic diagram of an anti-vibration hammer image shown in an exemplary embodiment of the present disclosure is shown;

[0020] Figure 2B A schematic diagram of an anti-vibration hammer image after hammer body area segmentation is shown in an exemplary embodiment of the present disclosure;

[0021] Figure 2C A schematic diagram showing an angle between a first fitting straight line and a second fitting straight line shown in an exemplary embodiment of the present disclosure;

[0022] Figure 3 is a flow chart of a method for segmenting a hammer region according to an exemplary embodiment of the present disclosure;

[0023] Figure 4 A schematic diagram of an image annotation of an anti-vibration hammer shown in an exemplary embodiment of the present disclosure is shown;

[0024] Figure 5 is a block diagram of a device for detecting a bending state of an anti-vibration hammer according to an exemplary embodiment of the present disclosure;

[0025] Figure 6 It is a block diagram of an electronic device for implementing the method for detecting the bending state of the vibration-proof hammer according to the embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0027] The method for detecting the bending state of the anti-vibration hammer provided in the embodiment of the present disclosure is applied to the scene of detecting the state of the anti-vibration hammer in the inspection scene of the power grid transmission line. In the inspection scene of the power grid transmission line, for example, an aerial photography method can be used to collect the transmission line image, and the state of the anti-vibration hammer is detected based on the image features in the collected transmission line image.

[0028] The detection of the state of the anti-vibration hammer in the related technology mainly includes the following two methods. Among them, in one method, the target detection model is used to detect the state of the anti-vibration hammer. When the target detection model is used to detect the state of the anti-vibration hammer, the anti-vibration hammer defects are mainly directly marked, or the existence of anti-vibration hammer defects is identified based on image features, and then a secondary detection is performed to determine the state of the anti-vibration hammer. Among them, the anti-vibration hammer state detection using the target detection model can locate the position of the anti-vibration hammer and make a judgment on more obvious damage, but in the detection of the bending state of the anti-vibration hammer, whether it will be judged as bending depends entirely on the data annotation. Different data annotators have different determinations on the degree of bending. Some anti-vibration hammers with unclear bending may be partially detected or partially not detected. There is a situation where the bending definition is unclear. Therefore, the traditional target detection method is not suitable for judging the state detection of the anti-vibration hammer bending and deformation.

[0029] In another method, the key point detection method is used to detect the state of the anti-vibration hammer. In this method, the anti-vibration hammer area image target is detected based on the transmission line image to obtain the width and height of the bounding box of the anti-vibration hammer area image, and the existence state values ​​and coordinate values ​​of the three key points of the left vertex, right vertex, and midpoint of the wire clamp body of the bounding box of the anti-vibration hammer area image are detected. According to the existence state values ​​and coordinate values ​​of these three key points, it is judged whether the anti-vibration hammer has defects such as missing hammer head, loose hammer head and deformation, and the defect detection of the anti-vibration hammer is completed. However, this method has a good detection effect on the detection of loose deformation and missing hammer head, but when the anti-vibration hammer is bent, the hammer bodies at both ends sink at the same time into an inconspicuous inverted V shape, and the height difference between the left and right endpoints and the midpoint of the wire clamp body is not obvious, so it is impossible to better judge whether it is bent, and the hammer body of the anti-vibration hammer may be damaged and deformed. The key points are not unique and difficult to mark.

[0030] Therefore, at present, in the scheme of detecting the bending state of the anti-vibration hammer, the deep learning method is used for identification. Since there are a large number of inconsistent standards in the labels for bending annotation that rely on manual annotation, it is difficult to improve the final model effect and generalization.

[0031] In view of this, the embodiment of the present disclosure provides a method for detecting the bending state of the anti-vibration hammer, by performing image segmentation on the anti-vibration hammer image to obtain hammer body regions of different hammer bodies of the anti-vibration hammer. The bending angles between different hammer bodies are determined by the hammer body regions obtained by segmentation, so as to quantify the bending angles of the hammer body parts, and then determine the bending state of the anti-vibration hammer based on the bending angles.

[0032] The anti-vibration hammer generally includes two hammer bodies, which are sometimes referred to as a left hammer body and a right hammer body. For the convenience of description, the present disclosure refers to the two different hammer bodies as a first hammer body and a second hammer body.

[0033] As an exemplary embodiment, Figure 1 FIG. 1 is a flow chart of a method for detecting the bending state of an anti-vibration hammer according to an exemplary embodiment of the present disclosure. Figure 1 As shown, the process includes the following steps S101 to S104.

[0034] In step S101 , an anti-vibration hammer image is acquired.

[0035] The anti-vibration hammer image in the present disclosure may be an image of a power transmission line including an anti-vibration hammer, which is acquired by aerial photography in a power grid transmission line inspection scenario.

[0036] Figure 2A A schematic diagram of an anti-vibration hammer image shown in an exemplary embodiment of the present disclosure is shown.

[0037] In step S102 , a first hammer body region and a second hammer body region are segmented from the vibration-proof hammer image.

[0038] In the embodiment of the present disclosure, when performing hammer area segmentation to obtain the first hammer area and the second hammer area, an image segmentation algorithm can be used for image segmentation. For example, a semantic recognition-based method can be used to segment the first hammer area and the second hammer area in the anti-vibration hammer image.

[0039] Figure 2B A schematic diagram of an anti-vibration hammer image after hammer body area segmentation is shown in an exemplary embodiment of the present disclosure. Figure 2B The reference numerals 1 and 2 in the figure represent the first hammer body region and the second hammer body region, respectively.

[0040] In step S103, straight line fitting is performed on the first hammer body region to obtain a first fitting straight line, and straight line fitting is performed on the second hammer body region to obtain a second fitting straight line.

[0041] In the embodiment of the present disclosure, a straight line fitting algorithm can be used to perform straight line fitting. For example, an opencv straight line fitting algorithm can be used to perform straight line fitting on the first hammer body region and the second hammer body region respectively to obtain a first fitting straight line and a second fitting straight line.

[0042] Among them, opencv is a general algorithm. Opencv can be understood as a cross-platform computer vision and machine learning software library. Opencv can run on a specified operating system, including functions and providing programming language interfaces, which can realize image processing and computer vision processing.

[0043] In step S104 , the angle between the first fitting straight line and the second fitting straight line is calculated, and the bending state of the vibration-damping weight is determined based on the angle between the first fitting straight line and the second fitting straight line.

[0044] In the disclosed embodiment, the image of the anti-vibration hammer is segmented to obtain the first hammer body area and the second hammer body area, and the bending angle between the hammer bodies is determined by calculating the angle of the fitting straight line obtained by fitting the first hammer body area and the second hammer body area, and then the bending state between the hammer bodies of the anti-vibration hammer can be determined by the bending angle, which is not affected by the subjective influence of the artificially labeled data, thereby improving the accuracy of the anti-vibration hammer bending state detection. In addition, there is no need to manually label the bending defect data, which can reduce the cost and complexity of the anti-vibration hammer bending state detection.

[0045] In an exemplary embodiment of the present disclosure, when performing straight line fitting on the first hammer body region and the second hammer body region obtained by image segmentation, a currently available straight line fitting algorithm may be used, such as an ellipse fitting method or a centroid fitting method.

[0046] In the embodiment of the present disclosure, a straight line fitting method can be used to fit a direction line passing through the centroid of the hammer. In one method, in the embodiment of the present disclosure, a straight line fitting method can be used to fit a direction line passing through the centroid. For example, a straight line fitting of the direction line passing through the center of gravity is performed on the first hammer region to obtain a first fitting straight line, and a straight line fitting of the direction line passing through the center of gravity is performed on the second hammer region to obtain a second fitting straight line. In the embodiment of the present disclosure, since the anti-vibration hammer has a regular shape, a unique direction line passing through the center of gravity can be fitted. Therefore, by using a straight line fitting method passing through the direction line passing through the center of gravity to obtain the first fitting straight line and the second fitting straight line, a fitting straight line with a better fitting effect can be obtained.

[0047] The first fitting straight line can be understood as the center axis of the first hammer, and the second fitting straight line can be understood as the center axis of the second hammer.

[0048] In the disclosed embodiment, if the anti-vibration hammer is not bent, the central axis of the first hammer body and the central axis of the second hammer body are located on a straight line, that is, the angle between the first fitting straight line and the second fitting straight line can be understood as a horizontal angle of 180°. Therefore, after determining the angle between the first fitting straight line and the second fitting straight line, the angle difference between the angle and the horizontal angle can be defined as the bending angle between the first hammer body and the second hammer body. Figure 2C A schematic diagram showing the angle between a first fitting straight line and a second fitting straight line in an exemplary embodiment of the present disclosure is shown. For example, Figure 2C In the figure, the angle between the first fitting straight line and the second fitting straight line is 124°, and the bending angle between the first hammer body and the second hammer body is 180°-124°=56°.

[0049] In an exemplary embodiment of the present disclosure, a model for hammer region segmentation may be trained based on deep learning, and a first hammer region and a second hammer region may be segmented in an anti-vibration hammer image based on the model.

[0050] In the embodiment of the present disclosure, a model for segmenting the first hammer body area and the second hammer body area can be pre-trained. When performing the anti-vibration hammer bending state detection, the pre-trained model for segmenting the hammer body area is called, and the anti-vibration hammer image is input into the model to obtain the first hammer body area and the second hammer body area, so as to segment the hammer body area more efficiently and improve the efficiency of the anti-vibration hammer bending state detection.

[0051] In one implementation of the disclosed embodiment, the training input may be an original anti-vibration hammer image, and the output may be a model of the first hammer body area and the second hammer body area. In one example, the original anti-vibration hammer image may be, for example, an original aerial image collected.

[0052] However, due to the high resolution of the original anti-vibration hammer image, the segmentation model is very complicated to process, especially in the model training stage. Therefore, in another implementation of the disclosed embodiment, a two-level model can be trained, wherein the first-level model is a target detection model, which is used to detect the target detection frame that can include the anti-vibration hammer in the original anti-vibration hammer image. The second-level model is a hammer segmentation model, which is used to perform hammer region segmentation on the image in the target detection frame to obtain a first hammer region and a second hammer region.

[0053] As an exemplary embodiment, Figure 3 FIG. 1 is a flow chart of a method for segmenting a hammer region according to an exemplary embodiment of the present disclosure. Figure 3 As shown, the process includes the following steps S301 to S304.

[0054] In step S301, a pre-trained target detection model and a pre-trained hammer segmentation model are called.

[0055] In step S302, the anti-vibration hammer image is input into the target detection model to detect and obtain a target detection frame.

[0056] In step S303 , based on the target detection frame, a target image including a hammer body region and having a smaller size than the anti-vibration hammer image is determined in the anti-vibration hammer image.

[0057] In step S304, the target image is input into a hammer body segmentation model to perform hammer body segmentation on the hammer body region included in the target image to obtain a first hammer body region and a second hammer body region.

[0058] In the disclosed embodiment, the anti-vibration hammer image is input into the target detection model, and a target detection frame including the anti-vibration hammer is detected. Based on the target detection frame, a target image including a hammer body region and having a size smaller than the anti-vibration hammer image can be determined in the anti-vibration hammer image. Among them, the target image size is smaller than the size of the original anti-vibration hammer image, which can also be understood as most of the noise data in the anti-vibration hammer image is removed from the target image, and the resolution is also reduced. Therefore, the hammer body region segmentation based on the target image can reduce the processing complexity of the hammer body segmentation model.

[0059] Furthermore, the target detection model and hammer segmentation model in the embodiment of the present disclosure can be pre-trained using traditional deep learning methods.

[0060] The target detection model can be pre-trained in the following manner: collect an anti-vibration hammer image training data set, and perform data annotation on each anti-vibration hammer image in the anti-vibration hammer image training data set to obtain an annotation detection frame and hammer body region segmentation information. The annotation detection frame is used to provide target detection annotation of the anti-vibration hammer body. The hammer body region segmentation information is used to provide region segmentation annotation of the first hammer body region and the second hammer body region of the anti-vibration hammer. Figure 4 FIG. 1 shows a schematic diagram of an anti-vibration hammer image annotation in an exemplary embodiment of the present disclosure. Figure 4 As shown, the outer rectangular frame is the marking detection frame. Wherein, the numbers 1 and 2 represent the first hammer body area and the second hammer body area respectively.

[0061] In the embodiment of the present disclosure, the anti-vibration hammer image in the anti-vibration hammer image training data set can be used as an input feature, and the labeled detection box corresponding to the anti-vibration hammer image can be used as an output feature to train a target detection model.

[0062] Further, in an embodiment of the present disclosure, the anti-vibration hammer image in the anti-vibration hammer image training data set can be input into the target detection model to obtain the target detection frame corresponding to each anti-vibration hammer image. Then, based on the target detection frame, a target image including a hammer body region and having a size smaller than the anti-vibration hammer image is cropped from the corresponding anti-vibration hammer image. The target image is used as an input feature, and the first hammer body region and the second hammer body region marked as the hammer body region are used as output features to train and obtain a hammer body segmentation model.

[0063] In the disclosed embodiment, most of the noise data in the anti-vibration hammer image is removed from the target image, and thus the resolution is also reduced. Therefore, training the hammer segmentation model based on the target image can reduce the complexity of hammer segmentation model training.

[0064] Furthermore, in the embodiment of the present disclosure, during the training of the target detection model and the hammer segmentation model, the labeled bending angle data is not included in the training labeled data, so the subjective influence of human factors on the labeled data can be avoided. Therefore, the anti-vibration hammer bending state detection is performed based on the target detection model and the hammer segmentation model, so that the detection of the anti-vibration hammer is more robust, and the accuracy and recall rate of the detection of the anti-vibration hammer bending angle are also improved.

[0065] It can be further understood that, in practical applications, when performing a bending state detection of an anti-vibration hammer, the acquired anti-vibration hammer image can be an anti-vibration hammer image data set composed of multiple anti-vibration hammer images, and the anti-vibration hammers included in the anti-vibration hammer image data set can be the same or different. Therefore, in the embodiment of the present disclosure, for each anti-vibration hammer image in the anti-vibration hammer image data set, the hammer body region segmentation method involved in the above embodiment of the present disclosure can be used, and a straight line fitting can be performed based on the hammer body region segmentation result to determine the bending angle between the hammer bodies.

[0066] Furthermore, after the bending angle of each anti-vibration hammer image is determined in the embodiment of the present disclosure, the bending angle can be mapped to the anti-vibration hammer image. For example, if there are ten sets of anti-vibration hammer images in total, ten sets of bending angles are obtained, and the ten sets of bending angles need to be mapped to the corresponding anti-vibration hammer images. Among them, the mapped anti-vibration hammer image and bending angle can be provided to the corresponding business party that performs anti-vibration hammer bending state detection, and the business party determines the defect state of the anti-vibration hammer based on the bending angle.

[0067] In the disclosed embodiment, when detecting the bending state of the anti-vibration hammer, the original anti-vibration hammer image is input into the target detection model, and the target detection frame including the hammer body area is detected by the target detection model, and the target image smaller than the original anti-vibration hammer image is determined based on the target detection frame. The target image is input into the hammer body segmentation model, and the target image is segmented by the hammer body segmentation model to segment the hammer body area, and a straight line fitting is performed on the hammer body area to obtain a fitting line passing through the centroid of the anti-vibration hammer figure, and the state of the anti-vibration hammer is determined by the angle between the fitting lines, thereby realizing the angle quantification of the anti-vibration hammer bending angle detection. The bending degree detection based on the quantized angle is more concise and convenient in calculation compared with the traditional target detection method and the key point detection method.

[0068] Based on the same concept, an embodiment of the present disclosure also provides a device for detecting the bending state of an anti-vibration hammer.

[0069] It is understandable that the device for detecting the bending state of the anti-vibration hammer provided in the embodiment of the present disclosure includes hardware structures and / or software modules corresponding to the execution of each function in order to realize the above functions. In combination with the units and algorithm steps of each example disclosed in the embodiment of the present disclosure, the embodiment of the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the technical solution of the embodiment of the present disclosure.

[0070] As an exemplary embodiment, Figure 5is a block diagram of a device 500 for detecting a bending state of an anti-vibration hammer according to an exemplary embodiment of the present disclosure. Figure 5 As shown, the device 500 for detecting the bending state of the anti-vibration hammer includes an acquisition unit 501 , a segmentation unit 502 , a fitting unit 503 , and a calculation unit 504 .

[0071] The acquisition unit 501 is used to acquire an anti-vibration hammer image. The segmentation unit 502 is used to segment the anti-vibration hammer image into a first hammer region and a second hammer region. The fitting unit 503 is used to perform straight line fitting on the first hammer region to obtain a first fitting straight line, and perform straight line fitting on the second hammer region to obtain a second fitting straight line. The calculation unit 504 is used to calculate the angle between the first fitting straight line and the second fitting straight line, and determine the bending state of the anti-vibration hammer based on the angle.

[0072] The segmentation unit 502 segments the first hammer body region and the second hammer body region in the anti-vibration hammer image in the following manner: calling a pre-trained model for segmenting hammer body regions, and inputting the anti-vibration hammer image into the model to obtain the first hammer body region and the second hammer body region.

[0073] The segmentation unit 502 uses the following method to call a pre-trained model for segmenting the hammer area, and inputs the anti-vibration hammer image into the model to obtain the first hammer area and the second hammer area:

[0074] A pre-trained target detection model and a pre-trained hammer segmentation model are called; the anti-vibration hammer image is input into the target detection model to detect and obtain a target detection frame; based on the target detection frame, a target image including a hammer area and having a size smaller than the anti-vibration hammer image is determined in the anti-vibration hammer image; the target image is input into the hammer segmentation model to perform hammer segmentation on the hammer area included in the target image to obtain a first hammer area and a second hammer area.

[0075] The fitting unit 503 performs straight line fitting on the first hammer region to obtain a first fitting straight line, and performs straight line fitting on the second hammer region to obtain a second fitting straight line in the following manner:

[0076] A straight line fitting of the center of gravity direction line is performed on the first hammer body region to obtain a first fitting straight line, and a straight line fitting of the center of gravity direction line is performed on the second hammer body region to obtain a second fitting straight line.

[0077] Regarding the above-mentioned device involved in the present disclosure, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0078] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0079] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0080] Figure 6 A schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0081] like Figure 6 As shown, the device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0082] A number of components in the device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0083] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as a method for detecting the bending state of the anti-vibration hammer. For example, in some embodiments, the method for detecting the bending state of the anti-vibration hammer may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the method for detecting the bending state of the anti-vibration hammer described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to execute the method for detecting the bending state of the anti-vibration hammer in any other appropriate manner (for example, by means of firmware).

[0084] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0085] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0086] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0087] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0088] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0089] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0090] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0091] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for detecting the bending state of an anti-vibration hammer, comprising: Acquire an anti-vibration hammer image, wherein the anti-vibration hammer corresponding to the anti-vibration hammer image includes a first hammer body and a second hammer body; Inputting the anti-vibration hammer image into a target detection model to detect and obtain a target detection frame including the anti-vibration hammer; Based on the target detection frame, determining a target image including a hammer body area and having a size smaller than the anti-vibration hammer image in the anti-vibration hammer image; Performing image segmentation on the target image by using a hammer segmentation model to obtain a first hammer region and a second hammer region, wherein the first hammer region corresponds to the first hammer, and the second hammer region corresponds to the second hammer; Performing straight line fitting on the first hammer region to obtain a first fitting straight line, and performing straight line fitting on the second hammer region to obtain a second fitting straight line; calculating an angle between the first fitting straight line and the second fitting straight line, and determining a bending state of the anti-vibration hammer based on the angle; The acquired anti-vibration hammer image is an anti-vibration hammer image data set composed of multiple anti-vibration hammer images, the anti-vibration hammers included in the anti-vibration hammer image data set are the same, and the angle is an angle obtained after target detection, hammer area segmentation and straight line fitting are performed on each anti-vibration hammer image in the anti-vibration hammer image data set; The target detection model and the hammer segmentation model are pre-trained in the following manner: Collecting an anti-vibration hammer image training data set, and annotating each anti-vibration hammer image in the anti-vibration hammer image training data set to obtain an annotation detection frame and hammer region segmentation information, wherein the hammer region segmentation information is used to provide region segmentation annotation of a first hammer region and a second hammer region of the anti-vibration hammer; The anti-vibration hammer image in the anti-vibration hammer image training data set is used as the input feature, and the annotated detection box corresponding to the anti-vibration hammer image is used as the output feature to train the target detection model; The anti-vibration hammer images in the anti-vibration hammer image training data set are input into the target detection model to obtain the target detection frames corresponding to each anti-vibration hammer image. Based on the target detection frames corresponding to each anti-vibration hammer image, a target image including a hammer body area and smaller in size than the anti-vibration hammer image is cropped from the corresponding anti-vibration hammer image. The target image is used as an input feature, and the first hammer body area and the second hammer body area marked as the hammer body area are used as output features to train and obtain a hammer body segmentation model. Among them, the annotation detection box is used to provide target detection annotation of the anti-vibration hammer hammer body, and the hammer body area segmentation information is used to provide area segmentation annotation of the first hammer body area and the second hammer body area of ​​the anti-vibration hammer. The training annotation data does not include the annotated bending angle data.

2. The method according to claim 1, wherein: The performing straight line fitting on the first hammer region to obtain a first fitting straight line, and performing straight line fitting on the second hammer region to obtain a second fitting straight line, comprises: A straight line fitting of the center of gravity direction line is performed on the first hammer body region to obtain a first fitting straight line, and a straight line fitting of the center of gravity direction line is performed on the second hammer body region to obtain a second fitting straight line.

3. A device for detecting the bending state of an anti-vibration hammer, comprising: An acquisition unit, configured to acquire an anti-vibration hammer image, wherein the anti-vibration hammer corresponding to the anti-vibration hammer image includes a first hammer body and a second hammer body; a segmentation unit, configured to input the anti-vibration hammer image into a target detection model, detect and obtain a target detection frame including the anti-vibration hammer, determine a target image including a hammer body region and having a size smaller than the anti-vibration hammer image in the anti-vibration hammer image based on the target detection frame, and perform image segmentation on the target image using a hammer body segmentation model to obtain a first hammer body region and a second hammer body region, wherein the first hammer body region corresponds to the first hammer body, and the second hammer body region corresponds to the second hammer body; a fitting unit, configured to perform straight line fitting on the first hammer region to obtain a first fitting straight line, and perform straight line fitting on the second hammer region to obtain a second fitting straight line; a calculation unit, configured to calculate an angle between the first fitting straight line and the second fitting straight line, and determine a bending state of the anti-vibration hammer based on the angle; The acquired anti-vibration hammer image is an anti-vibration hammer image data set composed of multiple anti-vibration hammer images, the anti-vibration hammers included in the anti-vibration hammer image data set are the same, and the angle is an angle obtained after target detection, hammer area segmentation and straight line fitting are performed on each anti-vibration hammer image in the anti-vibration hammer image data set; The target detection model and the hammer segmentation model are pre-trained in the following manner: Collecting an anti-vibration hammer image training data set, and annotating each anti-vibration hammer image in the anti-vibration hammer image training data set to obtain an annotation detection frame and hammer region segmentation information, wherein the hammer region segmentation information is used to provide region segmentation annotation of a first hammer region and a second hammer region of the anti-vibration hammer; The anti-vibration hammer image in the anti-vibration hammer image training data set is used as the input feature, and the annotated detection box corresponding to the anti-vibration hammer image is used as the output feature to train the target detection model; The anti-vibration hammer images in the anti-vibration hammer image training data set are input into the target detection model to obtain the target detection frames corresponding to each anti-vibration hammer image. Based on the target detection frames corresponding to each anti-vibration hammer image, a target image including a hammer body area and smaller in size than the anti-vibration hammer image is cropped from the corresponding anti-vibration hammer image. The target image is used as an input feature, and the first hammer body area and the second hammer body area marked as the hammer body area are used as output features to train and obtain a hammer body segmentation model. Among them, the annotation detection box is used to provide target detection annotation of the anti-vibration hammer hammer body, and the hammer body area segmentation information is used to provide area segmentation annotation of the first hammer body area and the second hammer body area of ​​the anti-vibration hammer. The training annotation data does not include the annotated bending angle data.

4. The device according to claim 3, wherein: The fitting unit performs straight line fitting on the first hammer region to obtain a first fitting straight line, and performs straight line fitting on the second hammer region to obtain a second fitting straight line in the following manner: A straight line fitting of the center of gravity direction line is performed on the first hammer body region to obtain a first fitting straight line, and a straight line fitting of the center of gravity direction line is performed on the second hammer body region to obtain a second fitting straight line.

5. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 2.

6. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-2.

7. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 2.

Images