Power equipment nameplate identification method and device and storage medium

Abstract

The invention discloses a power equipment nameplate identification method and device and a storage medium. The method comprises the following steps: acquiring at least one original image containing a target power equipment nameplate; the original image is input into an image denoising model, a target to-be-recognized image is determined based on a model output result, and the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing a power equipment nameplate and an expected denoising image corresponding to the sample image; and determining a target nameplate area in the target to-be-identified image based on an edge detection method and a contour search method, and extracting equipment nameplate data in the target nameplate area. And the identification efficiency and accuracy of the equipment nameplate data are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method, a device and a storage medium for recognizing a nameplate of electric equipment. Background Art

[0002] Nameplate recognition of power equipment plays a vital role in modern power systems. Nameplates usually contain key information about the equipment and are essential for equipment maintenance, management, and fault diagnosis. With the development of intelligent recognition technology, more and more intelligent recognition systems are being applied to nameplate recognition of power equipment.

[0003] Traditional manual recognition methods are not only inefficient but also prone to errors. The accuracy of existing automated recognition systems in complex environments still needs to be improved. Existing recognition methods perform poorly in situations where lighting conditions vary greatly, especially in high noise levels, which can seriously affect recognition results. Summary of the invention

[0004] The present invention provides a method, a device and a storage medium for identifying a nameplate of electric equipment, so as to solve the problem of low recognition accuracy of equipment nameplate data.

[0005] According to one aspect of the present invention, a method for identifying a nameplate of an electric power device is provided, the method comprising:

[0006] Acquire at least one original image including a nameplate of a target electric power device;

[0007] Input the original image into an image denoising model, and determine a target image to be identified based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image including a nameplate of an electric power device and a desired denoised image corresponding to the sample image;

[0008] Based on the edge detection method and the contour search method, the target nameplate area in the target image to be identified is determined, and the equipment nameplate data in the target nameplate area is extracted.

[0009] According to another aspect of the present invention, a device for identifying a nameplate of an electric equipment is provided, the device comprising:

[0010] An original image acquisition module, used to acquire at least one original image containing a nameplate of a target electric power device;

[0011] A module for determining an image to be identified, used for inputting the original image into an image denoising model, and determining a target image to be identified based on an output result of the model, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image including a nameplate of an electric equipment and a desired denoised image corresponding to the sample image;

[0012] The device nameplate data extraction module is used to determine the target nameplate area in the target image to be identified based on edge detection method and contour search method, and extract the device nameplate data in the target nameplate area.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying the nameplate of electric equipment described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying the nameplate of an electric power device described in any embodiment of the present invention when executed.

[0018] The technical solution of the embodiment of the present invention is to obtain at least one original image containing the nameplate of the target power equipment; then, input the original image into the image denoising model, and determine the target image to be identified based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing the nameplate of the power equipment and an expected denoised image corresponding to the sample image; improve the image quality of the target image to be identified, and improve the accuracy of subsequent data identification based on the target image to be identified; finally, determine the target nameplate area in the target image to be identified based on the edge detection method and the contour search method, extract the equipment nameplate data in the target nameplate area, and accurately extract the equipment nameplate data in the target nameplate area. The problem of low recognition accuracy of equipment nameplate data is solved, and the beneficial effect of improving the recognition efficiency and accuracy of equipment nameplate data is achieved.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0021] Figure 1 is a flow chart of a method for identifying a nameplate of electric equipment provided according to Embodiment 1 of the present invention;

[0022] Figure 2 is a flow chart of a method for identifying a nameplate of electric equipment provided according to Embodiment 2 of the present invention;

[0023] Figure 3 This is a schematic diagram of the structure of a device for identifying a nameplate of electric equipment provided according to a third embodiment of the present invention;

[0024] Figure 4 The present invention is a schematic diagram of the structure of an electronic device for implementing the method for identifying the nameplate of electric equipment according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] Embodiment 1

[0028] Figure 1A flowchart of a method for identifying a nameplate of an electric device is provided for the first embodiment of the present invention. This embodiment is applicable to the case of identifying a nameplate of an electric device. The method can be performed by an identification device for the nameplate of an electric device. The identification device for the nameplate of an electric device can be implemented in the form of hardware and / or software. The identification device for the nameplate of an electric device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0029] S110: Acquire at least one original image containing a nameplate of a target electric power device.

[0030] Among them, the nameplate of the power equipment can be understood as a sign on the power equipment marked with data related to the power equipment.

[0031] Specifically, at least one original image including the nameplate of the target electric equipment is acquired through a camera, and the original image may include images of the nameplate of the target electric equipment taken at different shooting angles or different shooting brightnesses.

[0032] Optionally, obtaining at least one original image containing the nameplate of the target power equipment includes: obtaining a nameplate image of the target power equipment nameplate based on multiple sets of shooting parameters, and fusing the multiple nameplate images to obtain the original image containing the nameplate of the target power equipment.

[0033] Specifically, the nameplate images are taken from multiple angles or at different light source intensities, and the multiple nameplate images are fused to reduce the impact of reflections and shadows. Use a multi-light source imaging system to take images of the nameplate from multiple angles or at different light source intensities; fuse the multiple nameplate images using a preset image fusion algorithm, and perform weighted averaging of the grayscale values ​​(or color values) of the corresponding pixels of different images to obtain the original image containing the target power equipment nameplate; or, decompose the image into different scales and different resolutions, and then fuse the low-frequency energy information and high-frequency detail information separately to obtain the original image containing the target power equipment nameplate. Among them, the preset image fusion algorithm includes a weighted average image fusion algorithm or a multi-resolution decomposition and fusion algorithm.

[0034] Preferably, a multimodal fusion framework can also be designed to combine visual information with sensor data to enhance the understanding and positioning capabilities of non-standard nameplate styles. Through visual sensors (e.g., cameras): obtain two-dimensional image data. Through lidar, obtain three-dimensional point cloud data; align the two-dimensional image with the three-dimensional point cloud data to establish a corresponding relationship between the two. Use three-dimensional point cloud data to assist in the analysis of two-dimensional images; combine visual features and point cloud features to perform feature-level fusion, and based on the fused multimodal data, perform subsequent nameplate data recognition (e.g., character recognition, pattern recognition) and matching (e.g., template matching, feature matching). Improve the robustness and accuracy of the system.

[0035] Optionally, after obtaining at least one original image of the nameplate of the target power equipment, it also includes: preprocessing the original image, and updating the original nameplate image based on the preprocessed original image; wherein the preprocessing includes: at least one of grayscale processing, gradient transformation processing, smoothing filtering processing and thresholding processing.

[0036] Specifically, an original image containing the nameplate of the target power equipment is read, and at least one or a combination of grayscale processing, gradient transformation processing, smoothing filtering processing and threshold processing is selected as needed to pre-process the original nameplate image, and the pre-processed image is updated to a new original nameplate image.

[0037] S120: Input the original image into an image denoising model, and determine a target image to be identified based on the model output result.

[0038] The image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing a nameplate of an electric power device and a desired denoised image corresponding to the sample image. The target image to be identified can be understood as an image of the nameplate data to be identified.

[0039] Specifically, the original image is input into a pre-trained generative model, and denoising is performed on the original image based on the generative model to output the denoised original image, and the denoised original image is used as the target image to be identified.

[0040] Optionally, inputting the original nameplate image into an image denoising model to obtain a target nameplate image to be identified includes: filtering the original image to obtain a first denoised image; inputting the first denoised image into the image denoising model to obtain a second denoised image; and determining the target image to be identified based on the first denoised image and the second denoised image.

[0041] The first denoised image can be understood as an image that is initially filtered by a filtering algorithm, and the second denoised image can be understood as a model output image.

[0042] Specifically, a preset filtering algorithm (such as median filtering or bilateral filtering) is first used to perform preliminary denoising on the image to obtain a first denoised image. The first denoised image is input into the image denoising model to obtain a second denoised image. Finally, the target image to be identified is determined based on the first denoised image and the second denoised image.

[0043] Optionally, determining the target image to be identified based on the first denoised image and the second denoised image includes: fusing the first denoised image and the second denoised image to obtain an initial image to be identified; and filling image gaps in the initial image to be identified to obtain the target image to be identified.

[0044] Specifically, the first denoised image and the second denoised image are fused to obtain an initial image to be identified. The initial image to be identified is processed by a preset morphological algorithm to obtain a target image to be identified. The preset morphological algorithm includes corrosion, expansion, opening and closing.

[0045] Exemplarily, using a composite morphological algorithm to fill the graphic gaps of the initial image to be identified includes: first performing a closing operation, and then performing a morphological operation again in the order of first corrosion and then expansion, so that the processing result of the morphological filter and the target in the original image are finally optimally matched. The initial image to be identified is closed to fill the small gaps inside the object. The isolated points and small structures in the initial image to be identified are further removed by the corrosion operation to make the image smoother. Subsequently, the boundaries of the objects in the initial image to be identified are restored by the expansion operation, and new small holes or gaps that may be generated by the corrosion operation are filled. By adjusting the size and shape of the structural element, as well as the number of corrosion and expansion operations, the fit between the processing result of the morphological filter and the target in the original image can be optimized, so that the fit between the processing result of the morphological filter and the target in the original image is finally optimally matched.

[0046] Optionally, before inputting the original image into the image denoising model, it also includes: obtaining multiple training samples, wherein the training samples include multiple noisy sample original images and expected denoising images corresponding to the sample original images; inputting the sample original images into a pre-established generative model of a generative adversarial network to obtain a model denoising image corresponding to the sample loss image; determining a first image denoising loss between the model denoising image and the expected denoising image corresponding to the model denoising image through a preset loss function; determining a second image denoising loss between the model denoising image and the expected denoising image corresponding to the model denoising image through a discriminant model of the generative adversarial network; and adjusting model parameters of the generative model based on the first image denoising loss and the second image denoising loss to obtain the image denoising model.

[0047] Specifically, multiple pairs of original images with noise and corresponding desired denoised images are collected. These image pairs constitute a training data set for training a generative adversarial network. The original image of the sample with noise is input into the generative model. The generative model generates a denoised image, i.e., a model denoised image, based on the input noisy image. The difference between the model denoised image and the desired denoised image is calculated using a preset loss function (such as mean square error, peak signal-to-noise ratio, etc.) to obtain a first image denoising loss. The discriminant model discriminates the model denoised image and the desired denoised image, and outputs a discrimination result, i.e., a second image denoising loss. Based on the first image denoising loss and the second image denoising loss, the parameters of the generative model are adjusted using an optimization algorithm to reduce the two loss values. At the same time, the parameters of the discriminant model are also adjusted according to the discrimination result to improve its discrimination ability. Repeat the above training process until the generative model can generate a model denoised image that is sufficiently close to the desired denoised image, and the discriminant model is difficult to distinguish whether the generated image is a model denoised image or a real image, then the model is considered to have converged, the model training is completed, and the image denoising model is obtained.

[0048] S130, determining a target nameplate area in the target image to be identified based on an edge detection method and a contour search method, and extracting device nameplate data in the target nameplate area.

[0049] The target nameplate area can be understood as the rectangular area of ​​the target nameplate in the target image to be recognized. The equipment nameplate data can be understood as the power equipment related data recorded on the target nameplate.

[0050] Specifically, based on the edge detection method and the contour search method, the area in the target image to be identified whose probability is greater than a preset probability threshold is determined as the target nameplate area, and the characters in the nameplate area are identified through the optical character recognition algorithm to extract the equipment nameplate data in the target nameplate area.

[0051] Optionally, the nameplate data includes at least one of equipment model, equipment serial number, rated voltage, rated current, rated frequency and manufacturing date.

[0052] The technical solution of the embodiment of the present invention is to obtain at least one original image containing the nameplate of the target power equipment; then, input the original image into the image denoising model, and determine the target image to be identified based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing the nameplate of the power equipment and an expected denoised image corresponding to the sample image; improve the image quality of the target image to be identified, and improve the accuracy of subsequent data identification based on the target image to be identified; finally, determine the target nameplate area in the target image to be identified based on the edge detection method and the contour search method, extract the equipment nameplate data in the target nameplate area, and accurately extract the equipment nameplate data in the target nameplate area. The problem of low recognition accuracy of equipment nameplate data is solved, and the beneficial effect of improving the recognition efficiency and accuracy of equipment nameplate data is achieved.

[0053] Embodiment 2

[0054] Figure 2 A flow chart of a method for identifying nameplates of electric equipment provided in Embodiment 2 of the present invention, this embodiment is a further optimization of how to determine the target nameplate area in the target image to be identified based on the edge detection method and the contour search method in the above embodiment. Optionally, the method of determining the target nameplate area in the target image to be identified based on the edge detection method and the contour search method includes: determining edge data in the target image to be identified based on the edge detection method, and determining contour data in the target image to be identified based on the contour search algorithm; determining a candidate nameplate area based on the edge data and the contour data, determining multiple corner points in the candidate nameplate area based on the corner point detection algorithm, and performing perspective projection transformation on the multiple corner points to obtain the target nameplate area.

[0055] like Figure 2 As shown, the method includes:

[0056] S210: Acquire at least one original image containing a nameplate of a target electric power device.

[0057] S220: Input the original image into an image denoising model, and determine a target image to be identified based on the model output result.

[0058] S230: determining edge data in the target image to be identified based on an edge detection method, and determining contour data in the target image to be identified based on a contour search algorithm.

[0059] Specifically, the edge detection algorithm is used to detect the edge of the nameplate area in the target image to be identified to obtain edge data. The contour of the nameplate area is identified by the contour search algorithm. The candidate nameplate area is screened out according to the size, shape and other features of the contour.

[0060] S240, determining a candidate nameplate area based on the edge data and the contour data, determining multiple corner points in the candidate nameplate area based on a corner point detection algorithm, performing perspective projection transformation on the multiple corner points to obtain a target nameplate area, and extracting the equipment nameplate data in the target nameplate area.

[0061] Specifically, in the candidate nameplate area, multiple corner points in the candidate nameplate area are determined by a corner point detection algorithm, and four corner points are selected from the multiple corner points detected. These four corner points should represent the four vertices of the candidate nameplate area. The perspective transformation matrix is ​​calculated using the four corner points and their corresponding target positions. Based on the perspective transformation matrix, the candidate nameplate area is transformed by perspective projection to obtain the target nameplate area. Among them, the corner point can be a point in the image where the brightness changes dramatically, such as: an inflection point or an intersection point of an object boundary.

[0062] Preferably, an adaptive illumination adjustment technology can be introduced to dynamically adjust the exposure and contrast according to the collected original image to improve the recognition robustness; specifically, the method includes: analyzing the brightness distribution of the original image and calculating global and local brightness statistics; dynamically adjusting the exposure and contrast of the image according to the brightness statistics to ensure that the nameplate area in the original image has good visibility. Using an illumination invariant feature extractor based on deep learning, the model is trained by a large number of nameplate images under different illumination conditions, so as to learn key features that are not affected by illumination changes. Exemplarily, the overall brightness mean and standard deviation of the original image are calculated to evaluate the global brightness level and contrast of the original image. The original image is divided into multiple grids, and brightness statistics are performed on each grid to identify highlight and shadow areas in the original image. According to the global brightness mean, if the image is too dark, the exposure is increased; if the image is too bright, the exposure is reduced. According to the global brightness standard deviation, if the contrast is low, a contrast enhancement algorithm is applied. The contrast of the local area is improved by adaptive contrast enhancement technology. Collect a large number of sample nameplate images under different lighting conditions, including nameplates, to ensure that the data set covers various lighting scenarios, such as strong light, weak light, backlight, shadow, etc. Train the convolutional neural network based on the sample nameplate images so that it can learn the lighting invariance characteristics of the nameplate under different lighting conditions. Input the image after adaptive lighting adjustment into the trained deep learning model, and the model outputs the lighting invariance characteristics of the nameplate.

[0063] The technical solution of the embodiment of the present invention improves the positioning accuracy of the nameplate area by determining the edge data in the target image to be identified based on the edge detection method, and determining the contour data in the target image to be identified based on the contour search algorithm; determining the candidate nameplate area based on the edge data and the contour data, determining multiple corner points in the candidate nameplate area based on the corner point detection algorithm, and performing perspective projection transformation on the multiple corner points to obtain the target nameplate area. This helps to simplify the subsequent image recognition process and improve data recognition efficiency.

[0064] As an optional example of an embodiment of the present invention, the method for identifying the nameplate of an electric power device in this embodiment specifically includes the following steps:

[0065] Step S1: Acquire at least one original image containing a nameplate of a target electric power device.

[0066] Step S2: preprocessing the original image, and updating the original nameplate image based on the preprocessed original image; wherein the preprocessing includes: at least one of grayscale processing, gradient transformation processing, smoothing filtering processing and thresholding processing.

[0067] Step S3: Adaptively denoising the original nameplate image in step S2 to eliminate isolated noise points in the regional image to obtain a target image to be identified.

[0068] The adaptive denoising process specifically uses the generator part in the generative adversarial network to generate a noise-free image, while the discriminator is used to evaluate the denoising effect and continuously optimize the denoising process.

[0069] Step S4: using a preset morphological algorithm to fill in the graphic gaps of the target image to be identified obtained in step S3, so as to filter out the protruding blocks that affect the target image.

[0070] Step S5: Determine the target nameplate area in the target image to be identified based on the edge detection method and the contour search method, and extract the equipment nameplate data in the target nameplate area.

[0071] In this embodiment, in step S2, the original image is preprocessed in the following manner:

[0072] Step S21: grayscale processing is performed on the original image to obtain a single-channel grayscale image.

[0073] Step S22: performing gradient transformation processing on the grayscale image obtained in step S21 to obtain a grayscale image capable of identifying edges; the gradient transformation processing uses the Scharr operator to construct the gradient amplitude of the grayscale image in the horizontal and vertical directions.

[0074] Step S23: performing smooth filtering processing on the image acquired in step S22 to obtain a filtered image; the smooth filtering processing uses a low-pass smoothing filter, and illustratively, a 5*5 smoothing matrix is ​​used for smooth filtering.

[0075] Step S24: performing threshold processing on the filtered image obtained in step S23 to obtain a regional image with a clear outline.

[0076] The thresholding process specifically includes the following steps:

[0077] Step S241: The threshold value of the foreground and background of the image is determined by using the maximum inter-class variance method. The ratio of the number of background pixels to the total number of image pixels is set as p0, and the ratio of the number of foreground pixels to the total number of image pixels is recorded as p1. The non-average grayscale values ​​of the background and foreground are q0 and q1 respectively. The average grayscale value of the image is:

[0078] average = p0 × q0 + p1 × q1;

[0079] Step S242: Calculate the variance between the image foreground and background, which is:

[0080] variance0=p0× ( q0-average ) 2 +p1× ( q1-average ) 2 ;

[0081] Step S243: Determine the segmentation threshold when the variance is maximum.

[0082] In this embodiment, in step S3, the adaptive denoising process further includes a hybrid filtering technology, combining a traditional filtering algorithm with a method based on deep learning to further refine the denoising result; specifically, the following steps are included:

[0083] Step S31: filtering the original image to obtain a first denoised image;

[0084] Specifically, a preset filtering algorithm (such as median filtering or bilateral filtering) is used to perform preliminary denoising on the image to obtain a first denoised image.

[0085] Step S32: inputting the first denoised image into the image denoising model to obtain a second denoised image;

[0086] Specifically, the image after preliminary denoising is input into a pre-trained deep neural network, which has been trained with a large number of noisy and noise-free image pairs and can automatically learn and remove the best strategy for different types of noise.

[0087] Step S33: determining a target image to be identified based on the first denoised image and the second denoised image.

[0088] Specifically, the results of initial filtering and deep learning denoising are fused to obtain the target image to be recognized.

[0089] In this embodiment, in step S4, the preset morphological algorithm is a composite morphological algorithm including four morphologies of corrosion, expansion, opening and closing; the preset morphological algorithm is used to fill the graphic gap of the target image to be identified obtained in step S3, and the specific method is to first perform a closing operation, and then perform the morphological operation again in the order of corrosion and then expansion, so that the processing result of the morphological filter and the target in the original image are finally optimally fitted; specifically, the following steps are included:

[0090] Step S41: First, a closing operation is performed on the image to fill small holes and break lines.

[0091] Step S42: then performing an erosion reduction operation to remove small interfering structures;

[0092] Step S43: then perform a dilation operation to restore the corroded part and further smooth the edge;

[0093] Step S44: Repeat the above steps until the image fit reaches the best.

[0094] In this embodiment, in step S5, the specific method of contour extraction is to use a contour search algorithm to find the largest rectangular frame surrounding the nameplate and mark the area where the nameplate is located, and then use corner point detection to identify the four edge points of the nameplate in the area, and perform perspective projection operation on the four edge points, and finally convert the irregular quadrilateral into a rectangle; specifically, the following steps are included:

[0095] Step S51: determining edge data in the target image to be identified based on an edge detection method;

[0096] Step S52: determining contour data in the target image to be identified based on a contour search algorithm;

[0097] Step S53: determining a candidate nameplate area based on the edge data and the contour data;

[0098] Step S54: determining a plurality of corner points in the candidate nameplate region based on a corner point detection algorithm;

[0099] Step S55: performing perspective projection transformation on the plurality of corner points to obtain a target nameplate area.

[0100] Preferably, in this embodiment, an adaptive illumination adjustment technology is introduced to dynamically adjust the exposure and contrast according to the collected original image to improve the recognition robustness; specifically, the following steps are included:

[0101] Step S61: Analyze the brightness distribution of the original image and calculate global and local brightness statistics;

[0102] Step S62: dynamically adjusting the exposure and contrast of the original image according to the brightness statistical information to ensure that the nameplate area in the original image has good visibility;

[0103] Step S63: Using a deep learning-based illumination invariant feature extractor, the model is trained with a large number of nameplate images under different illumination conditions, thereby learning key features that are not affected by illumination changes.

[0104] In this embodiment, a multi-light source imaging technology is also included, which obtains a nameplate image of the target power equipment nameplate based on multiple sets of shooting parameters, and fuses the multiple nameplate images to obtain an original image containing the target power equipment nameplate.

[0105] Exemplarily, nameplate images are captured from multiple angles or under different light source intensities, and then these images are fused through an algorithm to reduce the impact of reflections and shadows; specifically, the following steps are included:

[0106] Step S71: using a multi-light source imaging system to obtain a nameplate photographic image of a target electric power equipment nameplate based on multiple sets of photographing parameters;

[0107] Step S72: performing fusion processing on the plurality of nameplate images to obtain an original image including the nameplate of the target electric power equipment.

[0108] Specifically, the original image reduces the impact of reflections and shadows, improving recognition accuracy.

[0109] In this embodiment, a multimodal fusion framework is also designed to combine visual information with other sensor data (such as lidar) to enhance the understanding and positioning capabilities of non-standard nameplate styles; specifically, the following steps are included:

[0110] Step S81: Acquire the three-dimensional point cloud data of the target nameplate through the laser radar, and acquire the two-dimensional image data of the target nameplate through the camera.

[0111] Step S82: registering the two-dimensional image data with the three-dimensional point cloud data to establish a corresponding relationship between the two-dimensional image data and the three-dimensional point cloud data;

[0112] Step S83: using the three-dimensional point cloud data to assist in the analysis of the two-dimensional image data;

[0113] For example: using 3D geometric information to help locate and identify the location and shape of nameplates.

[0114] Step S84: The result of multimodal data fusion is used in subsequent recognition and matching processes to improve the robustness and accuracy of the system.

[0115] The technical solution of the embodiment of the present invention adopts a generative adversarial network for adaptive denoising. It can effectively eliminate isolated noise points and process more complex noise patterns. Adaptive illumination adjustment technology is introduced to dynamically adjust exposure and contrast according to the collected image. This enables the system to maintain good recognition performance under different lighting conditions. By analyzing the brightness distribution of the image and dynamically adjusting the exposure and contrast, good visibility of the nameplate area is ensured. In addition, the illumination invariance feature extractor based on deep learning further enhances the robustness so that it is not affected by illumination changes; through preprocessing, adaptive denoising, morphological processing and contour extraction steps, the accuracy and robustness of nameplate recognition of power equipment are significantly improved, which is particularly suitable for complex environments and diverse nameplate styles.

[0116] Embodiment 3

[0117] Figure 3 This is a schematic diagram of the structure of a device for identifying a nameplate of an electric equipment provided in the third embodiment of the present invention. Figure 3 As shown, the device includes: an original image acquisition module 310, a to-be-recognized image determination module 320 and an equipment nameplate data extraction module 330.

[0118] Among them, the original image acquisition module 310 is used to acquire at least one original image containing the target power equipment nameplate; the image to be identified determination module 320 is used to input the original image into the image denoising model, and determine the target image to be identified based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing the power equipment nameplate and an expected denoised image corresponding to the sample image; the equipment nameplate data extraction module 330 is used to determine the target nameplate area in the target image to be identified based on the edge detection method and the contour search method, and extract the equipment nameplate data in the target nameplate area.

[0119] The technical solution of the embodiment of the present invention is to obtain at least one original image containing the nameplate of the target power equipment through the original image acquisition module; then, through the image determination module to be identified, the original image is input into the image denoising model, and the target image to be identified is determined based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image containing the nameplate of the power equipment and an expected denoised image corresponding to the sample image; improve the image quality of the target image to be identified, and improve the accuracy of subsequent data identification based on the target image to be identified; finally, through the equipment nameplate data extraction module, the target nameplate area in the target image to be identified is determined based on the edge detection method and the contour search method, and the equipment nameplate data in the target nameplate area is extracted, and the equipment nameplate data in the target nameplate area is accurately extracted. The problem of low recognition accuracy of equipment nameplate data is solved, and the beneficial effect of improving the recognition efficiency and accuracy of equipment nameplate data is achieved.

[0120] Optionally, the to-be-recognized image determination module includes:

[0121] A first denoising unit, configured to perform filtering processing on the original image to obtain a first denoised image;

[0122] A second denoising unit, configured to input the first denoised image into the image denoising model to obtain a second denoised image;

[0123] The to-be-recognized image determining unit is configured to determine a target to-be-recognized image based on the first denoised image and the second denoised image.

[0124] Optionally, the to-be-recognized image determination unit includes:

[0125] a fusion processing subunit, configured to fuse the first denoised image and the second denoised image to obtain an initial image to be recognized;

[0126] The filling processing subunit is used to fill the image gaps in the initial image to be recognized to obtain a target image to be recognized.

[0127] Optionally, the device also includes a sample acquisition module, a model training module, a first loss determination module, a second loss determination module, and a model adjustment module.

[0128] The sample acquisition module is used to acquire a plurality of training samples before inputting the original image into the image denoising model, wherein the training samples include a plurality of sample original images with noise and desired denoised images corresponding to the sample original images;

[0129] The model training module is used to input the sample original image into a pre-established generative adversarial network generation model to obtain a model denoised image corresponding to the sample loss image;

[0130] The first loss determination module is used to determine a first image denoising loss between the model denoised image and the expected denoised image corresponding to the model denoised image by using a preset loss function;

[0131] The second loss determination module is used to determine a second image denoising loss between the model denoised image and the expected denoised image corresponding to the model denoised image by generating a discriminant model of the adversarial network;

[0132] The model adjustment module is used to adjust the model parameters of the generation model based on the first image denoising loss and the second image denoising loss to obtain the image denoising model.

[0133] Optionally, the equipment nameplate data extraction module includes:

[0134] A data extraction unit, used to determine edge data in the target image to be identified based on an edge detection method, and to determine contour data in the target image to be identified based on a contour search algorithm;

[0135] The target area determination unit is used to determine a candidate nameplate area based on the edge data and the contour data, determine multiple corner points in the candidate nameplate area based on a corner point detection algorithm, and perform perspective projection transformation on the multiple corner points to obtain a target nameplate area.

[0136] Optionally, the original image acquisition module is specifically used to:

[0137] A nameplate photographed image of a target electric power equipment nameplate is acquired based on multiple sets of photographing parameters, and a fusion process is performed on the multiple nameplate photographed images to obtain an original image containing the nameplate of the target electric power equipment.

[0138] Optionally, the device also includes: a preprocessing module.

[0139] Wherein, the preprocessing module is used to preprocess the original image after obtaining at least one original image of the nameplate of the target power equipment, and update the original nameplate image based on the preprocessed original image; wherein, the preprocessing includes: at least one of grayscale processing, gradient transformation processing, smoothing filtering processing and thresholding processing.

[0140] Optionally, the nameplate data includes at least one of equipment model, equipment serial number, rated voltage, rated current, rated frequency and manufacturing date.

[0141] The device for identifying the nameplate of electric equipment provided in the embodiment of the present invention can execute the method for identifying the nameplate of electric equipment provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0142] Embodiment 4

[0143] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention 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 (such as helmets, glasses, watches, etc.) 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 invention described and / or required herein.

[0144] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0145] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0146] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as the method for identifying the nameplate of the power equipment.

[0147] In some embodiments, the method of identifying the nameplate of the power device may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method of identifying the nameplate of the power device described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the method of identifying the nameplate of the power device by any other suitable means (e.g., by means of firmware).

[0148] 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.

[0149] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program 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.

[0150] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. 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.

[0151] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device 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 a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0152] The systems and techniques described herein may be implemented in a computing system that includes backend 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 frontend 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 backend components, middleware components, or frontend 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), a blockchain network, and the Internet.

[0153] A computing 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 client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0154] 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 described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0155] The above specific implementations do not constitute a limitation on the protection scope of the present invention. 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 invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying a nameplate of an electric power device, characterized in that: include: Acquire at least one original image including a nameplate of a target electric power device; Input the original image into an image denoising model, and determine a target image to be identified based on the model output result, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image including a nameplate of an electric power device and a desired denoised image corresponding to the sample image; Based on the edge detection method and the contour search method, the target nameplate area in the target image to be identified is determined, and the equipment nameplate data in the target nameplate area is extracted.

2. The method according to claim 1, characterized in that The step of inputting the original nameplate image into an image denoising model to obtain a target nameplate image to be identified includes: Performing filtering on the original image to obtain a first denoised image; Inputting the first denoised image into the image denoising model to obtain a second denoised image; A target image to be recognized is determined based on the first denoised image and the second denoised image.

3. The method according to claim 2, characterized in that The step of determining a target image to be identified based on the first denoised image and the second denoised image includes: Fusing the first denoised image and the second denoised image to obtain an initial image to be recognized; The image gaps in the initial image to be identified are filled to obtain a target image to be identified.

4. The method according to claim 1, characterized in that: Before inputting the original image into the image denoising model, the method further includes: Acquire a plurality of training samples, wherein the training samples include a plurality of sample original images with noise and desired denoised images corresponding to the sample original images; Inputting the sample original image into a pre-established generative adversarial network generation model to obtain a model denoised image corresponding to the sample loss image; Determine a first image denoising loss between the model denoised image and the desired denoised image corresponding to the model denoised image by a preset loss function; Determining a second image denoising loss between the model denoised image and the desired denoised image corresponding to the model denoised image by a discriminant model of a generative adversarial network; The model parameters of the generation model are adjusted based on the first image denoising loss and the second image denoising loss to obtain the image denoising model.

5. The method according to claim 1, characterized in that The method of determining the target nameplate area in the target image to be identified based on the edge detection method and the contour search method includes: Determining edge data in the target image to be identified based on an edge detection method, and determining contour data in the target image to be identified based on a contour finding algorithm; A candidate nameplate area is determined based on the edge data and the contour data, a plurality of corner points in the candidate nameplate area is determined based on a corner point detection algorithm, and perspective projection transformation is performed on the plurality of corner points to obtain a target nameplate area.

6. The method according to claim 1, characterized in that The obtaining of at least one original image containing the nameplate of the target electric power equipment comprises: A nameplate photographed image of a target electric power equipment nameplate is acquired based on multiple sets of photographing parameters, and a fusion process is performed on the multiple nameplate photographed images to obtain an original image containing the nameplate of the target electric power equipment.

7. The method according to claim 1, characterized in that After obtaining at least one original image of the target electric equipment nameplate, the method further includes: The original image is preprocessed, and the original nameplate image is updated based on the preprocessed original image; wherein the preprocessing includes: at least one of grayscale processing, gradient transformation processing, smoothing filtering processing and thresholding processing.

8. The method according to claim 1, characterized in that in, The nameplate data includes at least one of a device model, a device serial number, a rated voltage, a rated current, a rated frequency, and a manufacturing date.

9. A device for identifying nameplates of electric equipment, characterized in that: include: An original image acquisition module, used to acquire at least one original image containing a nameplate of a target electric power device; A module for determining an image to be identified, used for inputting the original image into an image denoising model, and determining a target image to be identified based on an output result of the model, wherein the image denoising model is a generative model obtained by training a generative adversarial network based on a sample image including a nameplate of an electric equipment and a desired denoised image corresponding to the sample image; The device nameplate data extraction module is used to determine the target nameplate area in the target image to be identified based on edge detection method and contour search method, and extract the device nameplate data in the target nameplate area.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying the nameplate of an electric power equipment according to any one of claims 1 to 8 when executed.

Images