A method and system for image verification of low-voltage metering equipment in the power industry

Abstract

The present invention discloses an image verification method and system for low-voltage metering equipment in the electric power industry, comprising obtaining an image to be detected of a low-voltage metering equipment in the electric power industry, inputting the obtained image into a preset electric power image quality analysis model, performing image quality analysis, and outputting a final image quality analysis result, and then judging whether the quality of the image meets a preset quality standard. If it meets the preset quality standard, the image to be detected is input into a cloud for recognition to detect whether there is a fault in the electric power equipment. If it does not meet the preset quality detection standard, the image to be detected is re-acquired. Before recognizing the image to be detected, the quality of the image to be detected is first judged, low-quality images to be detected are screened out, and only high-quality images to be detected are recognized. By improving the quality of the image to be detected, the accuracy of the fault recognition result of the electric power equipment is improved.

Description

Technical Field

[0001] The present invention belongs to the field of power equipment safety, and in particular relates to a method for verifying the image acquisition integrity of low-voltage metering equipment in the power industry. Background Art

[0002] With the development of science and technology, the safety monitoring and fault analysis of power equipment are mostly analyzed through intelligent detection systems. The adoption of intelligent detection systems saves a lot of human resources, reduces the accident rate during the detection of power equipment, and greatly improves the efficiency of equipment detection. However, due to a series of quality problems in the collected images during the equipment detection process of the intelligent detection system, the analysis results are often inaccurate, and power equipment faults are missed or misdetected. Summary of the invention

[0003] In view of the problems existing in the above-mentioned prior art, the present invention provides an image verification method and system for low-voltage metering equipment in the electric power industry, which improves the accuracy of power equipment fault identification by performing quality inspection on the image before performing fault detection on the image.

[0004] A first aspect of an embodiment of the present invention provides an image verification method for low-voltage area metering equipment in the electric power industry, the method comprising:

[0005] Obtain images of low-voltage metering equipment to be inspected in the power industry;

[0006] Input the image to be detected of the metering equipment in the low-voltage area of ​​the power industry into the power image quality analysis model for quality analysis, and output the image quality analysis result;

[0007] Determining whether the image to be detected meets a preset quality standard based on the image quality analysis result;

[0008] If the preset quality standards are met, the image to be inspected will be identified in the cloud.

[0009] A second aspect of an embodiment of the present invention provides an image verification system for low-voltage area metering equipment in the electric power industry, the system comprising:

[0010] Image acquisition module, used to acquire the image to be tested of the metering equipment in the low-voltage area of ​​the power industry;

[0011] An image analysis module is used to input the image to be detected of the low-voltage metering equipment in the power industry into the power image quality analysis model for quality analysis, and output the image quality analysis result;

[0012] A result detection module, used to determine whether the image to be detected meets the preset quality standard based on the image quality analysis result;

[0013] If the preset quality standards are met, the image to be inspected will be identified in the cloud.

[0014] A third aspect of an embodiment of the present invention provides a terminal, the terminal comprising a processor and a memory, the memory storing at least one program code, the at least one program code being loaded and executed by the processor to implement the above-mentioned method for image verification of low-voltage metering equipment in the power industry.

[0015] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, in which at least one program code is stored. The at least one program code is loaded and executed by a processor to implement the above-mentioned method for image verification of low-voltage metering equipment in the power industry.

[0016] The present invention provides a method and system for image verification of low-voltage metering equipment in the electric power industry, which has the following beneficial effects:

[0017] The present invention discloses an image verification method and system for low-voltage metering equipment in the electric power industry, comprising obtaining an image to be detected of a low-voltage metering equipment in the electric power industry, inputting the obtained image into a preset electric power image quality analysis model, performing image quality analysis, and outputting a final image quality analysis result, and then judging whether the quality of the image meets a preset quality standard. If it meets the preset quality standard, the image to be detected is input into a cloud for recognition to detect whether there is a fault in the electric power equipment. If it does not meet the preset quality detection standard, the image to be detected is re-acquired. Before recognizing the image to be detected, the quality of the image to be detected is first judged, low-quality images to be detected are screened out, and only high-quality images to be detected are recognized. By improving the quality of the image to be detected, the accuracy of the fault recognition result of the electric power equipment is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is an overall flow chart of an image verification method for low-voltage area metering equipment in the power industry of the present invention;

[0020] Figure 2 is a flow chart of the screening process of the initial image set;

[0021] Figure 3 It is a flow chart of the analysis process of the model analysis results. DETAILED DESCRIPTION

[0022] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0023] An embodiment of the present invention provides an image verification method for low-voltage area metering equipment in the electric power industry, the method comprising:

[0024] Obtain images of low-voltage metering equipment to be inspected in the power industry;

[0025] Input the images to be inspected of the low-voltage metering equipment in the power industry into the power image quality analysis model for quality analysis, and output the image quality analysis results;

[0026] Based on the image quality analysis result, determine whether the image to be detected meets the preset quality standard;

[0027] If the preset quality standards are met, the image to be inspected will be identified in the cloud.

[0028] refer to Figure 1 In this embodiment, first, an image to be detected of a low-voltage metering device in the electric power industry is obtained through an image acquisition device, and the obtained image is input into a preset electric power image quality analysis model to perform image quality analysis, and the final image quality analysis result is output, and then it is determined whether the quality of the image meets the preset quality standard. If it meets the preset quality standard, the image to be detected is input into the cloud for identification to detect whether there is a fault in the electric power equipment. If it does not meet the preset quality detection standard, the image to be detected is re-acquired. In this embodiment, before identifying the image to be detected, the quality of the image to be detected is first determined, and low-quality images to be detected are screened out, and only high-quality images to be detected are identified. By improving the quality of the image to be detected, the accuracy of the fault identification result of the electric power equipment is improved.

[0029] Based on the above method, the above power image quality analysis model includes:

[0030] Obtain an initial set of images of metering equipment in low-voltage areas of the power industry;

[0031] Screen out the erroneous images in the initial image set and form the remaining images into a standard image set;

[0032] Input the standard image set into the network model to detect the image target and extract the power equipment in the standard image set;

[0033] Based on the power equipment in the extracted standard image set, the number of pixels occupied by the power equipment and the positions of the pixels are calculated;

[0034] Perform image quality analysis based on the number of pixels occupied by power equipment and the location of the pixels, and output the model analysis results;

[0035] Use OpenCV computer vision software to uniformly read and analyze standard data sets and output the reading and analysis results;

[0036] The model training results and the read analysis results are combined into the final output of the power image quality analysis model.

[0037] In this embodiment, the sample images are further read and analyzed by the penCV computer vision software, and the analysis results of the model are verified, thereby improving the accuracy of the power image quality analysis model. Specifically, the initial image set of the metering equipment in the low-voltage area of ​​the power industry is obtained, the erroneous images in the initial image set are screened out, and the remaining images are composed of a standard image set. The images in the standard image set are input into the network model for image target detection, and the power equipment in the standard image set is extracted, wherein the network model can determine the type and location of the power equipment, and then the proportion of the power equipment in the image and the location of the power equipment in the image are calculated based on the number of pixels and the pixel position in the area where the power equipment is located, and the image quality is further analyzed, and the model analysis results are output, and then the standard image set is uniformly read and analyzed by the OpenCV computer vision software, and the reading analysis results are output, and the model analysis results are verified by the reading analysis results, and the output results after verification are used as the final output results of the network model to complete the construction of the power image quality analysis model.

[0038] refer to Figure 2 ,Based on the above power image quality analysis model, the screening steps of the above initial image set are as follows:

[0039] S1: Convert the images in the initial image set into images of the same size;

[0040] S2: randomly select an image from the initial image set, and obtain the attributes of the selected image as the first attribute group;

[0041] S3: randomly select an image from the initial image set, obtain the attributes of the selected image and compare them with the attribute group. If the comparison passes, the selected image is divided into the attribute group that passes the comparison. If the comparison fails, the attributes of the selected image are used as a new attribute group.

[0042] S4: repeat step S3 until all images in the initial image set are compared;

[0043] S5: Select the attribute group with the largest number of images as the correct images, filter out the wrong images in the remaining attribute groups, and complete the screening of the initial image set.

[0044] In this embodiment, image target detection is performed, and the initial image set needs to be screened and the erroneous images therein need to be deleted. Specifically, first, all images in the initial image set are converted to the same size, and then an image is randomly selected from the initial image set, and the attributes of the selected image are obtained as the first attribute group, wherein the image attributes include image name, image format, image resolution, image size, image color channel, etc., and then an image is selected and the attributes are compared with the existing attribute groups in sequence. If the attributes are the same, the image is divided into the existing attribute group, and if the comparison fails, the image is divided into a new attribute group, and images are selected from the initial image set in sequence for attribute comparison. After all the images in the initial image set are compared, multiple attribute groups are obtained, and the attribute group with the largest number of images is selected as the standard attribute group, and the erroneous images in the remaining attribute groups are screened out, and the images in the standard attribute group are the standard image set.

[0045] Based on the above power image quality analysis model, the specific analysis process of the above model analysis results is as follows:

[0046] Use network model to extract power equipment in images;

[0047] If the power equipment in the image is extracted, the number of pixels occupied by the power equipment is obtained, and the ratio of the number of pixels occupied by the power equipment to the total number of pixels in the image is calculated;

[0048] If the ratio is within the preset threshold range, the image pixel points are represented by two-dimensional coordinates, and the distance between the power equipment and the edge of the image is calculated;

[0049] If the distance between the power equipment and the edge of the image meets the preset distance, the edge pixel points of the power equipment are obtained to describe the shape of the power equipment, and the similarity between the shape of the power equipment in the image and the preset template is calculated;

[0050] If the similarity meets the requirements, the model analysis results are output.

[0051] refer to Figure 3In this embodiment, a network model is used to extract the power equipment in the image. Specifically, the input channel of the network model includes a first branch entrance and a second branch entrance. When the image is input into the first branch entrance, the network model will identify the type of power equipment in the image and output a first result; when the image is input into the second branch entrance, the network model will set three prediction boxes to extract the position of the power equipment and output a second result; the first result and the second result are spliced ​​to extract the power equipment in the image. If the power equipment in the image is extracted, the number of pixels occupied by the power equipment is obtained, and the ratio of the number of pixels occupied by the power equipment to the total number of pixels in the image is calculated to determine the size of the power equipment in the image. If the ratio is within the preset threshold range, the image pixels are represented by two-dimensional coordinates to obtain the position of the power equipment in the image. In this embodiment, the threshold range is 0.3-0.6. If the position of the power equipment and the edge of the image meets the preset distance, the edge pixel points of the power equipment are obtained to describe the shape of the power equipment, and the similarity between the shape of the power equipment in the image and the preset template is calculated. The power equipment The position that meets the preset distance with the edge of the image is calculated as the closest distance between the center point of the power equipment and the edge of the image. If the required distance L satisfies L≥0.5S+0.1M, it meets the preset distance, where S represents the side length of the power equipment and M represents the side length of the image. Then, by obtaining the contour shape of the power equipment and calculating its similarity with the same type of power equipment in the preset template, the image integrity of the power equipment is judged. In this embodiment, the similarity is greater than 0.7. After the judgment is passed, the model analysis result of the image can be output. If any of the above conditions is not met, the analysis process is terminated and the image is acquired again.

[0052] Based on the above-mentioned power image quality analysis model, the specific process of using OpenCV computer vision software to uniformly read and analyze the standard image set is as follows:

[0053] The standard image set is stored in a unified directory, and the sample images in the standard image set are numbered in sequence;

[0054] Read the sample images in sequence based on the sample image numbers and perform feature analysis, classify the sample images according to the analysis results, and obtain a sample classification set;

[0055] The sample classification sets are numbered and the number of sample images in each sample classification set is specified. When the number of sample images reaches the specified value, samples are stopped from being added to the sample classification set.

[0056] The image features corresponding to the final classification set of the sample image are used as the reading analysis results.

[0057] In this embodiment, the standard image set is uniformly read and analyzed by OpenCV computer vision software to obtain the reading and analysis results. Specifically, the standard image set is first stored in a unified directory, and the storage range of the sample image is limited to facilitate the reading by the OpenCV computer vision software, and then the sample images in the standard image set are numbered in order, so that the analysis results correspond to the sample images one by one, which is convenient for subsequent statistical quantity, and the sample images are obtained according to the analysis results to obtain a sample classification set, wherein each sample classification set corresponds to an image feature, and the number of sample images of the sample classification set is specified. When the number of sample images reaches a specified value, the sample is stopped from being put into the sample classification set, and the image feature corresponding to the sample classification set finally obtained is used as the reading and analysis result.

[0058] Based on the above method, the above image quality analysis results include: whether the image includes power equipment, whether the power equipment in the image is complete, the state of the power equipment in the image, and the type of power equipment in the image.

[0059] Furthermore, the above-mentioned determination of whether the image to be detected meets the preset quality standard includes:

[0060] Read the quality analysis result of the image to be detected, and determine whether the image analysis result includes all the information;

[0061] If the image analysis result does not include all the image information, the image to be detected is acquired again;

[0062] If the image analysis result includes all image information, it is determined that the image information meets the preset standard;

[0063] If the image information does not meet the preset standard, the image to be detected is acquired again;

[0064] If the image information meets the preset standards, the image to be detected will be input into the cloud server for fault detection of low-voltage metering equipment in the power industry, and the fault analysis results will be sent to the mobile terminal.

[0065] In this embodiment, the subsequent operation on the image to be detected is determined by judging whether it meets the preset quality standard. Specifically, the quality analysis result of the image to be detected is read to judge whether the image analysis result includes all the image information, wherein the image quality analysis result includes: whether the image includes power equipment, whether the power equipment in the image is complete, the state of the power equipment in the image, and the type of power equipment in the image. If the quality analysis result of the image to be detected does not contain all the image information, the image to be detected is reacquired, otherwise it is judged whether the image information of the image to be detected meets the preset standard. In this embodiment, the type of power equipment in the image is manually judged to determine whether there are defects in the power equipment in the image and whether the power equipment needs to be repaired. If the image information of the image to be detected does not meet the preset standard, the image to be detected is reacquired, otherwise the image to be detected is input into the cloud server for fault detection of low-voltage metering equipment in the power industry, and the fault analysis result is sent to the mobile terminal.

[0066] The embodiment of the present invention provides an image verification system for low-voltage area metering equipment in the electric power industry, and the system includes:

[0067] An image acquisition module is used to acquire the image to be tested of the metering equipment in the low-voltage area of ​​the power industry;

[0068] An image analysis module is used to input the image to be tested of the low-voltage metering equipment in the power industry into the power image quality analysis model for quality analysis and output the image quality analysis results;

[0069] A result detection module is used to determine whether the image to be detected meets the preset quality standard based on the image quality analysis result;

[0070] If the preset quality standards are met, the image to be inspected will be identified in the cloud.

[0071] The image verification system for low-voltage substation metering equipment in the electric power industry provided in this embodiment and the image verification method embodiment for low-voltage substation metering equipment in the electric power industry provided in the above embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0072] An embodiment of the present invention provides a terminal, including a processor and a memory, wherein at least one program code is stored in the memory, and the at least one program code is loaded and executed by the processor to implement the above-mentioned method for image verification of low-voltage metering equipment in the power industry.

[0073] The terminal includes: at least one processor, a memory, a user interface and at least one network interface. The various components in the terminal are coupled together through a bus system. It can be understood that the bus system is used to achieve connection and communication between these components.

[0074] An embodiment of the present invention provides a computer-readable storage medium, in which at least one program code is stored. The at least one program code is loaded and executed by a processor to implement the above-mentioned method for image verification of low-voltage metering equipment in the power industry.

[0075] It is understood that the memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. The memory in the embodiment of the present invention is capable of storing data to support the operation of the terminal. Examples of such data include: any computer program for operating on the terminal, such as an operating system and an application program. Among them, the operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application program may include various applications.

[0076] The present invention is not limited to the above-mentioned specific implementation modes. Various changes made by ordinary technicians in this field based on the above-mentioned concepts without creative work are all within the protection scope of the present invention.

Claims

1. A method for image verification of low-voltage metering equipment in the power industry, characterized in that: The method comprises: Obtain images of low-voltage metering equipment to be inspected in the power industry; Input the images to be inspected of the low-voltage metering equipment in the power industry into the power image quality analysis model for quality analysis, and output the image quality analysis results; Determining whether the image to be detected meets a preset quality standard based on the image quality analysis result; If the preset quality standard is met, the image to be inspected will be identified in the cloud; The power image quality analysis model includes: Obtain an initial set of images of metering equipment in low-voltage areas of the power industry; Screening out erroneous images in the initial image set and forming the remaining images into a standard image set; Inputting the standard image set into the network model to perform image target detection, and extracting the power equipment in the standard image set; Based on the power equipment in the extracted standard image set, the number of pixels occupied by the power equipment and the positions of the pixels are calculated; Perform image quality analysis based on the number of pixels occupied by power equipment and the location of the pixels, and output the model analysis results; Using OpenCV computer vision software to uniformly read and analyze the standard data set, and output the reading and analysis results; Combining the model training result and the reading analysis result into the final output of the power image quality analysis model; The screening steps of the initial image set are as follows: S1: converting the images in the initial image set into images of the same size; S2: randomly selecting an image from the initial image set, and obtaining attributes of the selected image as a first attribute group; wherein the attributes of the image include image name, image format, image resolution, image size, and image color channel; S3: randomly select an image from the initial image set, obtain the attributes of the selected image and compare them with the attribute group. If the comparison passes, the selected image is divided into the attribute group that passes the comparison. If the comparison fails, the attributes of the selected image are used as a new attribute group. S4: repeat step S3 until all images in the initial image set are compared; S5: Select the attribute group with the largest number of images as the correct images, and filter out the wrong images in the remaining attribute groups to complete the screening of the initial image set; The specific analysis process of the model analysis results is as follows: Use network model to extract power equipment in images; If the power equipment in the image is extracted, the number of pixels occupied by the power equipment is obtained, and the ratio of the number of pixels occupied by the power equipment to the total number of pixels in the image is calculated; If the ratio is within a preset threshold range, the image pixel is represented by two-dimensional coordinates to obtain the position of the power equipment in the image; If the position of the power equipment and the edge of the image meets the preset distance, the edge pixel points of the power equipment are obtained to describe the shape of the power equipment, and the similarity between the shape of the power equipment in the image and the preset template is calculated; If the similarity meets the requirements, the model analysis results are output.

2. The method according to claim 1, characterized in that: The specific process of using OpenCV computer vision software to uniformly read and analyze the standard image set is as follows: The standard image set is stored in a unified directory, and the sample images in the standard image set are numbered in sequence; Read the sample images in sequence based on the sample image numbers and perform feature analysis, and classify the sample images according to the analysis results to obtain a sample classification set; The sample classification sets are numbered and the number of sample images in each sample classification set is specified. When the number of sample images reaches the specified value, samples are stopped from being added to the sample classification set. The image features corresponding to the final classification set of the sample image are used as the reading analysis results.

3. The method according to claim 1, characterized in that: The image quality analysis result includes: whether the image includes power equipment, whether the power equipment in the image is complete, the state of the power equipment in the image, and the type of the power equipment in the image.

4. The method according to claim 1, characterized in that: The determining whether the image to be detected meets the preset quality standard includes: Read the quality analysis result of the image to be detected, and determine whether the image analysis result includes all the information; If the image analysis result does not include all the image information, the image to be detected is acquired again; If the image analysis result includes all image information, it is determined that the image information meets the preset standard; If the image information does not meet the preset standard, the image to be detected is acquired again; If the image information meets the preset standards, the image to be detected will be input into the cloud server for fault detection of low-voltage metering equipment in the power industry, and the fault analysis results will be sent to the mobile terminal.

5. An image verification system for low-voltage metering equipment in the power industry, characterized in that: The system comprises: An image acquisition module is used to acquire the image to be tested of the metering equipment in the low-voltage area of ​​the power industry; An image analysis module is used to input the image to be tested of the metering equipment in the low-voltage area of ​​the power industry into the power image quality analysis model for quality analysis and output the image quality analysis results; A result detection module, used to determine whether the image to be detected meets the preset quality standard based on the image quality analysis result; If the preset quality standards are met, the image to be inspected will be identified in the cloud.

6. A terminal, characterized in that: The terminal includes a processor and a memory, wherein at least one program code is stored in the memory, and the at least one program code is loaded and executed by the processor to implement an image verification method for low-voltage substation metering equipment in the power industry as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that , the storage medium stores at least one program code, and the at least one program code is loaded and executed by the processor to implement an image verification method for low-voltage metering equipment in the power industry as described in any one of claims 1 to 4.

Images