A visual mechanical external damage identification method for power transmission channel and related device

Abstract

The present invention provides a method and related device for visualizing mechanical damage identification of a power transmission channel, and relates to the technical field of visualizing mechanical damage identification of a power transmission channel. By establishing a visual sample library of a power transmission channel, the detection targets and mechanical damage of the tower are marked in the visual sample library of the power transmission channel; the target detection model is trained using the target framework in combination with the visual sample library of the power transmission channel; the pixel information of the tower detection frame is obtained in the inference stage of the target detection model; the tower spacing, width and height are obtained and calculated in combination with the pixel information to obtain the tower ratio information; the tower height information and the span physical data information of the target tower are obtained through a preset sensing device and combined with the tower ratio and the first preset condition for the first data elimination; the pixel information of all current mechanical damage detection frames is obtained in combination with the tower ratio and the second preset condition for the second data elimination. The technical effect of improving the recognition accuracy in the target recognition process is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of visual mechanical damage identification of power transmission channels, and in particular to a visual mechanical damage identification method for power transmission channels and related devices. Background Art

[0002] In recent years, with the development of artificial intelligence technology, the "online monitoring + AI recognition" model replacement, which uses the visible light image of the online monitoring device as visual perception and image recognition as a detection method, will become the development direction of the intelligent early warning system for power transmission line protection against external force damage. 1. Complex scenes lead to many misidentifications. In complex scenes, traditional AI recognition technology may have misidentifications. This misidentification may be caused by factors such as lighting conditions, shadows, occlusions, and mirror reflections. For example, in wooded areas or busy streets in cities, monitoring devices may mistakenly identify trees, people or other objects as mechanical external damage objects, resulting in false alarms; 2. There are more low-risk detection results far away from poles and channels. In places far away from poles and channels, traditional AI recognition technology may produce a large number of low-risk detection results. This is because the image information observed from a long distance may not be clear enough, and the detailed features of the object cannot be accurately captured, resulting in a lack of credibility in the recognition results. These low-risk detection results may interfere with the judgment of operators.

[0003] Therefore, how to improve the recognition accuracy in the target recognition process has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In order to improve the recognition accuracy during the target recognition process, the present application provides a method and related device for visualizing mechanical external damage recognition of a power transmission channel.

[0005] In the first aspect, the present application provides a method for visually identifying mechanical damage in a power transmission channel using the following technical solutions:

[0006] A method for visualizing mechanical damage identification in a power transmission channel, comprising:

[0007] Establishing a transmission channel visualization sample library, marking detection targets and mechanical external damage of the tower in the transmission channel visualization sample library;

[0008] Using the target framework in combination with the transmission channel visualization sample library to train a target detection model;

[0009] Acquiring pixel information of the tower detection frame in the inference stage of the target detection model;

[0010] Obtaining the tower spacing, width and height and calculating them in combination with the pixel information to obtain the tower ratio information;

[0011] Obtaining tower height information and span physical data information of the target tower through a preset sensor device and performing a first data elimination in combination with the tower-tower ratio and a first preset condition;

[0012] The pixel information of all current mechanical damage detection frames is obtained and combined with the Tuta ratio and the second preset condition to perform a second data elimination.

[0013] Optionally, the step of establishing a transmission channel visualization sample library and marking the detection targets of the tower and mechanical external damage in the transmission channel visualization sample library includes:

[0014] Establishing a transmission channel visualization sample library by acquiring image information collected by a preset sensor device;

[0015] Acquire preset screening conditions, wherein the preset screening conditions include geographical conditions and lighting conditions;

[0016] Performing data cleaning and bbox annotation in the power transmission channel visualization sample library according to the preset screening conditions;

[0017] Obtain the current transmission channel visualization sample library and mark the detection targets and mechanical external damage of the tower.

[0018] Optionally, the pixel information of the tower detection frame includes: pixel width information and pixel height information.

[0019] Optionally, the step of obtaining the tower spacing, width and height and calculating them in combination with the pixel information to obtain the tower ratio information includes:

[0020] Get the tower spacing, width and height information, and set the upper angle coordinate of the tower (T i x1,T i y1), right angle coordinate (T i x2,T i y2), lower corner coordinate (T i x3,T i y3), the left corner coordinate is (T i x4,T i y4), the center coordinates of the image are (Cx, Cy) and the filtering logic is:

[0021] ;

[0022] Among them, T min That is, the annotation frame selected for calculating the tower ratio, i=1,2,...,n is the set sequence of tower annotation frames for model reasoning;

[0023] Use the tower record to obtain the height of the tower where the video surveillance equipment is located, which is Hi The tower in the picture is H i+1 , calculate the Tuta ratio ptR as:

[0024] .

[0025] Optionally, the step of obtaining the tower height information and the span physical data information of the target tower by a preset sensing device and performing a first data elimination in combination with the tower-tower ratio and the first preset condition includes:

[0026] Obtain tower height information and span physical data information of the target tower through a preset sensor device;

[0027] Obtaining a distance threshold corresponding to the first preset condition;

[0028] Using the tower ratio in combination with the tower height information and the span physical data information to calculate the physical space distance;

[0029] The first data elimination is performed according to the distance threshold and the physical space distance.

[0030] Optionally, the step of obtaining pixel information of all current mechanical damage detection frames and combining the Tutabi ratio and a second preset condition to perform a second data elimination includes:

[0031] Obtain a second preset condition, and obtain a preset mechanical equipment size table in the second preset condition;

[0032] Obtain pixel information of all current mechanical damage detection frames and generate a set of mechanical damage detection frame size information using the Tutabi ratio;

[0033] A second data elimination is performed based on the mechanical damage detection frame size information set and the preset mechanical equipment size table.

[0034] In a second aspect, the present application provides a device for visualizing mechanical damage identification of a power transmission channel, wherein the visualizing mechanical damage identification of a power transmission channel comprises:

[0035] A sample library module, used to establish a transmission channel visualization sample library, in which detection targets of the tower and mechanical external damage are marked;

[0036] A model training module, used for training a target detection model using a target framework in combination with the power transmission channel visualization sample library;

[0037] A pixel information module, used to obtain pixel information of the tower detection frame during the inference stage of the target detection model;

[0038] A figure-to-tower ratio information module, used to obtain the tower spacing, width and height and calculate them in combination with the pixel information to obtain the figure-to-tower ratio information;

[0039] A first data elimination module, used for obtaining the tower height information and the span physical data information of the target tower through a preset sensor device and performing the first data elimination in combination with the tower-tower ratio and the first preset condition;

[0040] The second data elimination module is used to obtain the pixel information of all current mechanical external damage detection frames and combine the picture tower ratio and the second preset condition to perform the second data elimination.

[0041] In a third aspect, the present application provides a device for visualizing mechanical damage to a power transmission channel, the device comprising: a memory and a processor, the processor executing the method described above when running a program for visualizing mechanical damage to a power transmission channel stored in the memory.

[0042] In a fourth aspect, the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a computer, enable the computer to execute the method described above.

[0043] In summary, the present application includes the following beneficial technical effects:

[0044] This application establishes a transmission channel visualization sample library, and marks the detection targets and mechanical external damage of the tower in the transmission channel visualization sample library; uses the target framework combined with the transmission channel visualization sample library to train the target detection model; obtains the pixel information of the tower detection frame in the inference stage of the target detection model; obtains the tower spacing, width and height combined with the pixel information to calculate to obtain the tower ratio information; obtains the tower height information and the span physical data information of the target tower through the preset sensor device and combines the tower ratio and the first preset condition to perform the first data elimination; obtains the pixel information of all current mechanical external damage detection frames combined with the tower ratio and the second preset condition to perform the second data elimination. The technical effect of improving the recognition accuracy in the target recognition process is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the structure of a visual mechanical external damage identification device for a power transmission channel in a hardware operating environment involved in an embodiment of the present application;

[0046] Figure 2 It is a flow chart of the first embodiment of the method for visualizing mechanical damage identification of a power transmission channel of the present application;

[0047] Figure 3 It is a structural block diagram of the first embodiment of the visual mechanical damage identification device for the power transmission channel of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below through the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0049] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a visual mechanical damage identification device for a power transmission channel in the hardware operating environment involved in an embodiment of the present application.

[0050] like Figure 1 As shown, the visual mechanical external damage identification device of the power transmission channel may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RandomAccess Memory, RAM), or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0051] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the device for visualizing mechanical damage to the power transmission channel, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0052] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a power transmission channel visual mechanical damage identification program.

[0053] exist Figure 1In the power transmission channel visual mechanical external rupture identification device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the present application can be set in the power transmission channel visual mechanical external rupture identification device, and the power transmission channel visual mechanical external rupture identification device calls the power transmission channel visual mechanical external rupture identification program stored in the memory 1005 through the processor 1001, and executes the power transmission channel visual mechanical external rupture identification method provided in the embodiment of the present application.

[0054] The present application embodiment provides a method for visually identifying mechanical damage in a power transmission channel, referring to Figure 2 , Figure 2 This is a flow chart of the first embodiment of the method for visualizing mechanical damage identification of a power transmission channel of the present application.

[0055] In this embodiment, the method for visualizing mechanical damage identification of a power transmission channel includes the following steps:

[0056] Step S10: Establish a power transmission channel visualization sample library, and mark the detection targets and mechanical external damage of the tower in the power transmission channel visualization sample library.

[0057] It should be noted that the steps of establishing a transmission channel visualization sample library and marking the detection targets of the tower and mechanical external damage in the transmission channel visualization sample library include: establishing the transmission channel visualization sample library by acquiring image information collected by a preset sensor device; acquiring preset filtering conditions, the preset filtering conditions include geographical conditions and lighting conditions; performing data cleaning and bbox marking in the transmission channel visualization sample library according to the preset filtering conditions; acquiring the current transmission channel visualization sample library and marking the detection targets of the tower and mechanical external damage.

[0058] In the specific implementation, a transmission channel visualization sample library is established, the detection targets and mechanical damage of the tower are annotated, and the target detection model is trained. The sample library is obtained by taking and annotating the video surveillance equipment installed on the transmission line tower, including: considering the diversity of regions and lighting conditions, collecting samples from various cities and companies in multiple regions such as mountains, plains, suburbs, and urban areas, and under multiple lighting conditions such as early morning, evening, rain and fog, cleaning and bbox annotating the sample data, and obtaining about 20,000 tower annotations and 14 categories of more than 40,000 mechanical damage annotations; based on the above sample conditions, a suitable target detection model is selected for training. Since the tower accounts for a large proportion of pixels in the pictures taken by the video surveillance equipment, the mAP50 for the recognition of this large target is about 0.8425;

[0059] Step S20: Use the target framework in combination with the power transmission channel visualization sample library to train the target detection model.

[0060] Step S30: acquiring pixel information of the tower detection frame in the inference stage of the target detection model.

[0061] It can be understood that the pixel information of the tower detection frame includes: pixel width information and pixel height information.

[0062] Step S40: Obtain the tower spacing, width and height and calculate in combination with pixel information to obtain the tower ratio information.

[0063] It is understandable that it is necessary to obtain the inventory of the pole tower where the video surveillance equipment is located, and obtain the physical data of the pole tower height and span respectively; since multiple pole towers are likely to appear within the field of view of the online detection device, and the identification frames of the pole towers may overlap or be misplaced, it is necessary to screen the validity of multiple pole tower identification frames in the reasoning stage according to the requirements of the device installation.

[0064] In a specific implementation, the step of obtaining the tower spacing, width and height and combining the pixel information to calculate to obtain the tower ratio information includes: obtaining the tower spacing, width and height information, setting the upper angle coordinate of the tower (T i x1,T i y1), right angle coordinate (T i x2,T i y2), lower corner coordinate (T i x3,T i y3), the left corner coordinate is (T i x4,T i y4) The center coordinates of the image are (Cx, Cy) and the filtering logic is:

[0065] ;

[0066] Among them, T min That is, the selected annotation box used to calculate the tower ratio, i=1,2,...,n is the set sequence of tower annotation boxes for model reasoning; the tower record is used to obtain the height of the tower where the video surveillance equipment is located, which is H i The tower in the picture is H i+1 , calculate the Tuta ratio ptR as:

[0067] .

[0068] It should be noted that, in this embodiment, the physical data of the tower height and the span are obtained by obtaining the tower record where the video surveillance equipment is located.

[0069] It can be understood that this embodiment refers to the X-axis center coordinates of the photo to filter the recognition frame so that the center of the monitored channel is aligned with the center of the picture, thereby ensuring the accuracy and effectiveness of the recognition. Assume that the coordinates of the upper right, lower left corners of the tower are (Ti x1,T i y1),(T i x2,T i y2),(T i x3,T i y3),(T i x4,T i y4)i=1,2,…,n, the image center coordinates Center is (Cx,Cy) and the filtering logic is:

[0070] ;

[0071] In the above formula, T min That is, the annotation box selected for calculating the tower ratio, i=1,2,...,n is the set sequence of tower annotation boxes for model reasoning.

[0072] Use the tower record to obtain the height of the tower where the video surveillance equipment is located, which is H i The tower in the picture is H i+1 , calculate the ptR (unit: meter / pixel), the calculation method is:

[0073] ;

[0074] In the above formula, ptR is the Tuta ratio data for calculating the spatial distance.

[0075] Step S50: obtaining the tower height information and span physical data information of the target tower through a preset sensor device and performing the first data elimination in combination with the tower-to-tower ratio and the first preset condition.

[0076] It should be noted that the step of obtaining the tower height information and the physical data information of the span of the target tower through a preset sensing device and performing the first data elimination in combination with the figure-tower ratio and the first preset condition includes: obtaining the tower height information and the physical data information of the span of the target tower through a preset sensing device; obtaining the distance threshold corresponding to the first preset condition; using the figure-tower ratio in combination with the tower height information and the physical data information of the span to calculate the physical space distance; and performing the first data elimination according to the distance threshold and the physical space distance.

[0077] It should be noted that this embodiment uses the tower ratio parameter to calculate the physical space distance between the external damage edge and the tower, and excludes detection results that exceed the set threshold, including: excluding mechanical external damage detection results that are completely within the tower detection frame and whose vertical ratio to the tower is less than a certain threshold; setting a distance threshold to exclude mechanical external damage detection results that exceed the threshold above, right, below, and left of the tower detection frame.

[0078] Assume that the coordinates of the upper right, lower left corners of the tower are (T min x1, Tmin y1), (T min x2, T min y2),(T min x3, T min y3), (T min x4, T min y4), the four coordinates of the mechanical outer broken top right bottom left are: (P i x1, P i y1), (P i x2,P i y2), (P i x3, P i y3), (P i x4, P i y4) i=1,2,…,n , compare the coordinate values ​​of the detection frame and the four vertices of the tower one by one, and eliminate those that meet [[P i x n <T min x n && P i y n <T min y n ] n=1,2,3,4 ] i=1,2,…,n Regular detection box.

[0079] Assume that the coordinates of the upper right, lower left corners of the tower are (T min x1, T min y1), (T min x2, T min y2),(T min x3, T min y3), (T min x4, T min y4), the four coordinates of the mechanical outer broken top right bottom left are: (P i x1, P i y1), (P i x2,P i y2), (P i x3, P i y3), (P i x4, P i y4) i=1,2,…,n , set the distance threshold to TR top , TR right , TR bottom ,TR left :

[0080] 1) Set the upper distance threshold to: TR top , rule: if (Tmin y1 > P i y3 and ptR (T min y1 - P i y3)>= TR top );

[0081] 2) Set the right distance threshold to: TR right , rule: if (P i x1 > T min x2 and ptR (P i x1 -T min x2) >= TR right );

[0082] 3) Set the bottom distance threshold to: TR bottom , rule: if (P i y1 > T min y3 and ptR (P i y1 -T min y3) >= TR bottom );

[0083] 4) Set the left distance threshold to: TR left , rule: if (T min x1 > P i x2 and ptR (T min x1 - P i x2)>= TR left ) .

[0084] Step S60: Obtain pixel information of all current mechanical damage detection frames and combine it with the Tutabi ratio and the second preset condition to perform a second data elimination.

[0085] It should be noted that the first and second descriptions in the first and second preset conditions appearing in this embodiment are only used to distinguish two different preset conditions, and there is no distinction in order and importance. This embodiment does not limit the specific naming.

[0086] It can be understood that the step of obtaining pixel information of all current mechanical external damage detection frames in combination with Tutabi and the second preset condition to perform the second data elimination includes:

[0087] Obtain a second preset condition, and obtain a preset mechanical equipment size table in the second preset condition; obtain pixel information of all current mechanical external damage detection frames and use Tutabi to generate a mechanical external damage detection frame size information set; perform a second data elimination based on the mechanical external damage detection frame size information set combined with the preset mechanical equipment size table.

[0088] It should be noted that this calculation method assumes that the equipment and the tower are in a two-dimensional plane at the same depth. The pixel height and width of the detection frame can obtain a more realistic physical height and width, and then according to the type of equipment outside the mechanical device, establish the "Table 1 Preset Mechanical Equipment Size Table". If the obtained physical size is not within the preset mechanical size range, it means that the mechanical equipment and the tower are not in the same depth plane and are far away. Assume that the physical width of the mechanical equipment size is W and the height is H. The four coordinates of the upper right, lower left of the mechanical equipment detection frame are: (P i x1,P i y1),(P i x2,P i y2),(P i x3,P i y3),(P i x4,P i y4), the estimated size is calculated using the "Tutabi" parameter:

[0089] ;

[0090] Then W i , H i Compare with the preset mechanical physical height and width range to eliminate the detection frame that does not conform to the normal size of the object.

[0091] Table 1 Preset mechanical equipment size table

[0092]

[0093] It can be understood that, 1. The method provided in this embodiment can greatly reduce the false detection rate caused by "online monitoring + AI recognition" and avoid unnecessary interference and false alarms, thereby reducing the line patrol work intensity of the front-line team and achieving the purpose of improving the efficiency and effectiveness of transmission channel patrol.

[0094] 2. The spatial distance measurement method provided in this embodiment does not require changes or adjustments to the existing AI model. On the one hand, it can reduce the software and hardware investment costs of high-end artificial intelligence training equipment. On the other hand, it can non-invasively improve the online monitoring effect of the model, making it have a broader application base.

[0095] This embodiment establishes a transmission channel visualization sample library, and marks the detection targets and mechanical external damage of the tower in the transmission channel visualization sample library; uses the target framework combined with the transmission channel visualization sample library to train the target detection model; obtains the pixel information of the tower detection frame in the inference stage of the target detection model; obtains the tower spacing, width and height combined with the pixel information to calculate to obtain the tower ratio information; obtains the tower height information and the span physical data information of the target tower through the preset sensor device and combines the tower ratio and the first preset condition to perform the first data elimination; obtains the pixel information of all current mechanical external damage detection frames combined with the tower ratio and the second preset condition to perform the second data elimination. The technical effect of improving the recognition accuracy in the target recognition process is achieved.

[0096] In addition, an embodiment of the present application also proposes a computer-readable storage medium, on which is stored a program for visualizing mechanical external rupture identification of a power transmission channel. When the program for visualizing mechanical external rupture identification of a power transmission channel is executed by a processor, the steps of the method for visualizing mechanical external rupture identification of a power transmission channel as described above are implemented.

[0097] Reference Figure 3 , Figure 3 This is a structural block diagram of the first embodiment of the visual mechanical damage identification device for the power transmission channel of the present application.

[0098] like Figure 3 As shown, the visual mechanical damage identification device for a power transmission channel proposed in the embodiment of the present application includes:

[0099] The sample library module 10 is used to establish a transmission channel visualization sample library, and mark the detection targets of the tower and mechanical external damage in the transmission channel visualization sample library;

[0100] A model training module 20 is used to train a target detection model using a target framework combined with a transmission channel visualization sample library;

[0101] The pixel information module 30 is used to obtain the pixel information of the tower detection frame in the inference stage of the target detection model;

[0102] The figure-to-tower ratio information module 40 is used to obtain the tower spacing, width and height and calculate them in combination with pixel information to obtain the figure-to-tower ratio information;

[0103] A first data elimination module 50 is used to obtain the tower height information and the span physical data information of the target tower through a preset sensor device and perform the first data elimination in combination with the tower-tower ratio and the first preset condition;

[0104] The second data elimination module 60 is used to obtain pixel information of all current mechanical damage detection frames and combine the picture tower ratio and the second preset condition to perform the second data elimination.

[0105] It should be understood that the above is only an example and does not constitute any limitation on the technical solution of the present application. In specific applications, technicians in this field can make settings as needed, and the present application does not impose any limitation on this.

[0106] This embodiment establishes a transmission channel visualization sample library, and marks the detection targets and mechanical external damage of the tower in the transmission channel visualization sample library; uses the target framework combined with the transmission channel visualization sample library to train the target detection model; obtains the pixel information of the tower detection frame in the inference stage of the target detection model; obtains the tower spacing, width and height combined with the pixel information to calculate to obtain the tower ratio information; obtains the tower height information and the span physical data information of the target tower through the preset sensor device and combines the tower ratio and the first preset condition to perform the first data elimination; obtains the pixel information of all current mechanical external damage detection frames combined with the tower ratio and the second preset condition to perform the second data elimination. The technical effect of improving the recognition accuracy in the target recognition process is achieved.

[0107] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present application. In practical applications, technicians in this field can select part or all of it according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0108] In addition, for technical details not fully described in this embodiment, reference can be made to the method for visualizing mechanical damage identification of a power transmission channel provided in any embodiment of the present application, and will not be repeated here.

[0109] In addition, it should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0110] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory (ROM) / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of each embodiment of the present application.

[0112] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for visualizing mechanical damage identification in a power transmission channel, characterized in that: include: Establishing a transmission channel visualization sample library, marking detection targets and mechanical external damage of the tower in the transmission channel visualization sample library; Training a target detection model in combination with the power transmission channel visualization sample library; Acquiring pixel information of the tower detection frame in the inference stage of the target detection model; The actual height of the tower is obtained by a preset method and combined with the pixel information to perform calculation to obtain the tower ratio information; Obtaining tower height information and span physical data information of the target tower through a preset sensor device and performing a first data elimination in combination with the tower-tower ratio and a first preset condition; Obtain pixel information of all current mechanical external damage detection frames and combine the Tutabi ratio and the second preset condition to perform a second data elimination; The step of obtaining the tower height information and the span physical data information of the target tower by a preset sensing device and performing the first data elimination in combination with the tower-tower ratio and the first preset condition includes: Obtain tower height information and span physical data information of the target tower through a preset sensor device; Obtaining a distance threshold corresponding to the first preset condition; Using the figure-tower ratio in combination with the tower height information and the span physical data information to calculate the physical space distance between the mechanical outer breaking edge and the tower pole; Performing a first data elimination according to the distance threshold and the physical space distance; The step of obtaining pixel information of all current mechanical damage detection frames and combining the Tutabi ratio and the second preset condition to perform a second data elimination includes: Obtain a second preset condition, and obtain a preset mechanical equipment size table in the second preset condition; Obtain pixel information of all current mechanical damage detection frames and generate a set of mechanical damage detection frame size information using the Tutabi ratio; A second data elimination is performed based on the mechanical damage detection frame size information set and the preset mechanical equipment size table.

2. The method for visualizing mechanical damage identification of a power transmission channel according to claim 1, characterized in that: The step of establishing a transmission channel visualization sample library and marking the detection targets of the tower and mechanical external damage in the transmission channel visualization sample library includes: Establishing a transmission channel visualization sample library by acquiring image information collected by a preset sensor device; Acquire preset screening conditions, wherein the preset screening conditions include geographical conditions and lighting conditions; Performing data cleaning and bbox annotation in the power transmission channel visualization sample library according to the preset screening conditions; Obtain the current transmission channel visualization sample library and mark the detection targets and mechanical external damage of the tower.

3. The method for visualizing mechanical damage identification of a power transmission channel according to claim 1, characterized in that: The pixel information of the tower detection frame includes: pixel width information and pixel height information.

4. The method for visualizing mechanical damage identification of a power transmission channel according to claim 3, characterized in that: The step of obtaining the tower height information in a preset manner and performing calculation in combination with the pixel information to obtain the tower ratio information comprises: Assume the upper angle coordinate of the tower (T i x1, T i y1), right angle coordinate (T i x2, T i y2), lower corner coordinate (T i x3, T i y3), the left corner coordinate is (T i x4, T i y4), the image center coordinates Center is (Cx, Cy) and the filtering logic is: ； Among them, T min That is, the annotation frame selected for calculating the tower ratio, i=1,2,...,n is the set sequence of tower annotation frames for model reasoning; Use the tower record to obtain the actual height H of the tower where the video surveillance equipment is located i ; The tower in the picture is H i+1 , calculate the Tuta ratio ptR as: 。 5. A visual mechanical damage identification device for a power transmission channel, characterized in that: Executing the method according to claim 1, comprising: A sample library module, used to establish a transmission channel visualization sample library, in which detection targets of the tower and mechanical external damage are marked; A model training module, used for training a target detection model using a target framework in combination with the power transmission channel visualization sample library; A pixel information module, used to obtain pixel information of the tower detection frame during the inference stage of the target detection model; A tower ratio information module, used to obtain tower height information in a preset manner and calculate in combination with the pixel information to obtain tower ratio information; A first data elimination module, used for obtaining the tower height information and the span physical data information of the target tower through a preset sensor device and performing the first data elimination in combination with the tower-tower ratio and the first preset condition; The second data elimination module is used to obtain the pixel information of all current mechanical external damage detection frames and combine the picture tower ratio and the second preset condition to perform the second data elimination.

6. A visual mechanical damage identification device for a power transmission channel, characterized in that: The device comprises: a memory and a processor, and the processor executes the method according to any one of claims 1 to 4 when running a power transmission channel visual mechanical damage identification program stored in the memory.

7. A computer-readable storage medium, characterized in that: The method comprises instructions, which, when executed on a computer, cause the computer to execute the method according to any one of claims 1 to 4.

Images