Edge video analysis intelligent service method and system for power grid edge side

Abstract

This invention relates to the field of power technology and provides an intelligent service method for edge video analysis at the power grid edge. This method addresses the problem that existing network monitoring and analysis terminals cannot intelligently locate and mark edge scenes at power grid nodes, preventing backend servers from accurately acquiring the corresponding original video data. The intelligent service method for edge video analysis at the power grid edge includes: acquiring edge video data of power grid nodes to obtain a converted image set; decomposing the converted image set based on an image decomposition model, extracting the decomposed edge video feature points, and encoding the edge video feature points; and analyzing and calculating the edge video feature points using a preset edge video analysis model. This invention, by encoding the edge video feature points, ensures that the original video data can be accurately located, transmitted, and analyzed, thus improving the processing efficiency of the edge-side processor.

Description

Technical Field

[0001] This invention belongs to the field of power technology, and in particular relates to an intelligent service method and system for edge video analysis at the edge of the power grid. Background Technology

[0002] With the advent of the Internet of Things era and the popularization of wireless networks, the number of devices at the network edge and the amount of data they generate have increased dramatically. In order to ensure the normal operation of the power grid, it is necessary to set up network monitoring and analysis terminals in various scenarios such as transmission lines, substations, distribution lines and safety monitoring nodes in the power grid. Through the interaction and connection between the network monitoring and analysis terminals and the back-end server, real-time monitoring and analysis of various nodes in the power grid can be realized to ensure the normal operation of the power grid.

[0003] In recent years, with the widespread application of FPGAs, GPUs, and TPUs, edge computing capabilities have been greatly enhanced, and intelligent image recognition technology at the edge has also been widely used. Image recognition technology mainly fulfills the functional requirements of object / scene recognition, classification, localization, detection, and image segmentation within images or videos in industrial applications. However, existing network monitoring and analysis terminals cannot achieve intelligent localization and labeling of edge scenes at power grid nodes, preventing back-end servers from accurately obtaining the corresponding raw video data and affecting the judgment and analysis of the power grid edge. Based on this, we propose an intelligent service method and system for edge video analysis oriented towards the power grid edge. Summary of the Invention

[0004] This invention provides an intelligent service method and system for edge video analysis for the edge side of the power grid, aiming to solve the problem that existing network monitoring and analysis terminals cannot achieve intelligent positioning and marking of power grid node edge side scenes, making it impossible for back-end servers to accurately obtain the corresponding original video data.

[0005] This invention is implemented as follows: an intelligent service method for edge video analysis oriented towards the edge of the power grid, used for edge video analysis of power grid nodes, the intelligent service method for edge video analysis oriented towards the edge of the power grid includes:

[0006] Acquire edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor;

[0007] Obtain the preprocessing results of edge video data, and transform the preprocessed edge video data to obtain a transformed image set;

[0008] Based on the image decomposition model, the converted image set is decomposed, the edge video feature points after decomposition are extracted, and the edge video feature points are encoded.

[0009] The edge video feature points after encoding are obtained. Using the encoded edge video feature points as input, the edge video feature points are analyzed and calculated through a preset edge video analysis model to obtain the edge video analysis results of the power grid node.

[0010] Preferably, the method for preprocessing edge video data based on an edge-side processor specifically includes:

[0011] Extract edge video data of power grid nodes, determine whether the edge video data is encrypted, if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm, if the edge video data is not encrypted, classify the edge video data as data to be processed;

[0012] The decrypted edge video data is obtained as the edge video data to be processed. The decrypted edge video data is then integrated to obtain a video stream integration set. The video stream integration set is then filtered by the edge-side processor to obtain a high-definition video stream integration set.

[0013] Extract high-definition video streams and compress and convert them to 5G protocols.

[0014] Preferably, the method for converting the preprocessed edge video data to obtain a converted image set specifically includes:

[0015] Receive preprocessed edge video data and obtain the total number of frames in the edge video data;

[0016] The total number of frames in the edge video data is uniformly divided to obtain a uniform set of frame intervals;

[0017] Obtain the frame set corresponding to each set of frame number intervals;

[0018] Based on a preset threshold for a range greater than 1450 pixels x 1450 pixels, the frame sets corresponding to each set of frame ranges are filtered.

[0019] Preferably, the method for converting the preprocessed edge video data to obtain a converted image set further includes:

[0020] The filtered frame set is obtained, and the quality of the frames in the obtained frame set is measured and judged. Specifically, the THRESH BINARY threshold or the inverse binary threshold THRESH BINARY INV is used to measure and judge the quality of the frames.

[0021] Preferably, the compression and 5G protocol conversion of the integrated high-definition video stream specifically includes:

[0022] Based on HEVC ultra-low latency compression mode, read and integrate high-definition video streams;

[0023] A hybrid compression framework is established based on HEVC ultra-low latency compression, and predictive compression of the integrated video stream set is obtained based on the hybrid compression framework.

[0024] Eliminate the correlation between the time and spatial domains of the video stream aggregation set, and perform 5G protocol conversion on the HEVC-compressed video stream aggregation set.

[0025] Preferably, the method for establishing the image decomposition model specifically includes:

[0026] Establish a first disassembly dataset and a second disassembly dataset for disassembling image objects, wherein the first disassembly dataset contains multiple first frame images labeled with disassembly codes, and the second disassembly dataset contains multiple second frame images without disassembly labels;

[0027] The first disassembly dataset is used to predict the disassembly label of the second frame in the second disassembly dataset to obtain the disassembly label corresponding to the second frame. Then, the third frame labeled with the disassembly label is obtained based on the second frame and its corresponding disassembly label to establish the third disassembly dataset.

[0028] Repeat the above steps n times to obtain the nth decomposition dataset. Use the first decomposition dataset and the nth decomposition dataset to train a preset deep neural network to learn an image decomposition model.

[0029] Preferably, the method for extracting and encoding the decomposed edge video feature points specifically includes:

[0030] Determine the energy response results of pixels in the decomposed image at multiple preset response orientations;

[0031] The direction selection result of the feature point is determined based on the energy response result and the preset feature point selection rule;

[0032] The energy response results and the direction selection results of the feature points selected by the preset feature point selection rules are integrated to encode the edge video feature points.

[0033] Preferably, the analysis method of the edge video analysis model specifically includes:

[0034] Obtain edge video feature points after feature point encoding processing;

[0035] The edge video analysis model is used to buffer and load the edge video data of the power grid nodes corresponding to the edge video feature points;

[0036] By merging the edge video data of the power grid node with the center video data of the power grid node, the integrated edge video analysis results of the power grid node are obtained.

[0037] An edge video analytics intelligent service system for the power grid edge, based on the aforementioned edge video analytics intelligent service method for the power grid edge, comprises:

[0038] The edge video data acquisition module is used to acquire edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor.

[0039] The image set acquisition module acquires the preprocessing results of edge video data, converts the preprocessed edge video data, and obtains the converted image set.

[0040] The feature point coding module, based on the image decomposition model, decomposes the converted image set, extracts the edge video feature points after decomposition, and assigns codes to the edge video feature points.

[0041] The edge video analysis result acquisition module is used to acquire the coded edge video feature points. Using the coded edge video feature points as input, the module analyzes and calculates the edge video feature points through a preset edge video analysis model to obtain the edge video analysis results of the power grid nodes.

[0042] Preferably, the edge video data acquisition module includes:

[0043] The video data encryption judgment unit is used to extract edge video data of power grid nodes, determine whether the edge video data is encrypted, and if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm; if the edge video data is not encrypted, the edge video data is classified as data to be processed.

[0044] The video stream integration set acquisition unit acquires the decrypted edge video data to be processed, integrates the decrypted edge video data to be processed, and obtains the video stream integration set. The edge-side processor filters the video stream integration set to obtain a high-definition video stream integration set.

[0045] The integrated compression unit is used to extract high-definition video stream integrated sets, compress the high-definition video stream integrated sets, and perform 5G protocol conversion.

[0046] Compared with the prior art, the embodiments of this application have the following main advantages:

[0047] This invention, through the encoding of edge video feature points, ensures that the original video data can be accurately located, transmitted, analyzed, and calculated, thereby guaranteeing the accuracy of judgment and analysis of the power grid edge and improving the processing efficiency of the edge processor. Attached Figure Description

[0048] Figure 1 This is a schematic diagram illustrating the implementation process of the edge video analysis intelligent service method for the power grid edge side provided by the present invention.

[0049] Figure 2 This is a schematic diagram of the implementation process of preprocessing edge video data based on an edge-side processor provided by the present invention.

[0050] Figure 3 This is a schematic diagram illustrating the implementation process of compressing and converting high-definition video streams using the 5G protocol provided by the present invention.

[0051] Figure 4 This is a schematic diagram illustrating the implementation process of converting preprocessed edge video data to obtain a converted image set, as provided by the present invention.

[0052] Figure 5 This is a schematic diagram illustrating the implementation process of establishing the image decomposition model provided by the present invention.

[0053] Figure 6 This is a schematic diagram illustrating the implementation process of extracting and encoding edge video feature points after decomposition, as provided by the present invention.

[0054] Figure 7 This is a schematic diagram illustrating the implementation process of the analysis method of the edge video analysis model provided by the present invention.

[0055] Figure 8 This is a schematic diagram of the edge video analysis intelligent service system for the power grid edge side provided by the present invention.

[0056] Figure 9 This is a schematic diagram of the edge video data acquisition module provided by the present invention.

[0057] Figure 10 This is a schematic diagram of the structure of the image set acquisition module provided by the present invention.

[0058] In the diagram: 100 - Edge video data acquisition module, 110 - Video data encryption judgment unit, 120 - Video stream integration set acquisition unit, 130 - Integration set compression unit, 200 - Converted image set acquisition module, 210 - Total frame acquisition unit, 220 - Video data cutting unit, 230 - Frame set acquisition unit, 240 - Frame set filtering unit, 250 - Frame quality judgment unit, 300 - Feature point coding module, 400 - Edge video analysis result acquisition module. Detailed Implementation

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0060] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0061] Existing network monitoring and analysis terminals cannot achieve intelligent positioning and marking of power grid edge scenes, preventing back-end servers from accurately acquiring the corresponding raw video data and affecting the judgment and analysis of the power grid edge. Based on this, we propose an intelligent edge video analysis service method for the power grid edge. This method includes acquiring edge video data of power grid nodes, preprocessing the edge video data using an edge-side processor, obtaining the preprocessed edge video data results, converting the preprocessed edge video data to obtain a converted image set, decomposing the converted image set based on an image decomposition model, extracting the decomposed edge video feature points, encoding the edge video feature points, and finally obtaining the encoded edge video feature points. Using the encoded edge video feature points as input, a preset edge video analysis model is used to analyze and calculate the edge video feature points to obtain the power grid node edge video analysis results. This embodiment of the invention, through encoding the edge video feature points, ensures that the raw video data can be accurately positioned, transmitted, and analyzed, thereby guaranteeing the accuracy of the judgment and analysis of the power grid edge and improving the processing efficiency of the edge-side processor.

[0062] This invention provides an intelligent service method for edge video analytics oriented towards the power grid edge side, such as... Figure 1 As shown, the edge video analytics intelligent service method for the power grid edge side specifically includes:

[0063] Step S10: Obtain edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor.

[0064] Step S20: Obtain the preprocessing results of the edge video data, and convert the preprocessed edge video data to obtain a converted image set.

[0065] Step S30: Based on the image decomposition model, the converted image set is decomposed, the decomposed edge video feature points are extracted, and the edge video feature points are encoded.

[0066] Step S40: Obtain the coded edge video feature points. Using the coded edge video feature points as input, analyze and calculate the edge video feature points through a preset edge video analysis model to obtain the edge video analysis results of the power grid node.

[0067] In this embodiment, the edge processor is a wireless data transmission camera with a built-in front-end CPU. At least three sets of edge processors are set for each power grid node, thereby ensuring accurate analysis of the edge side of the node and improving the efficiency of edge side analysis for power grid node monitoring.

[0068] Meanwhile, in this embodiment, the edge-side processor not only sends the edge video data of the power grid node, but also records at least one set of center video data of the power grid node, thereby realizing full computational analysis of the power grid node. For example, the edge video data format of the power grid node includes, but is not limited to, avi, mp4, MPEG-7, MPEG-21, and WMV formats.

[0069] For example, the edge processor uses Python to acquire video streams via OpenCV; when the monitoring target is a power grid node, the power grid node is marked in the video with a red rectangle; then the edge processor extracts the feature information, uploads the feature information, and uses an edge video analysis model to analyze and calculate the edge video feature points to obtain the edge video analysis results of the power grid node.

[0070] For example, the edge processor terminal also has communication connections to a camera, RF (Radio Frequency) circuit, sensors, audio circuit, WiFi module, etc., including sensors such as light sensors, motion sensors, and other sensors.

[0071] This invention, through the encoding of edge video feature points, ensures that the original video data can be accurately located, transmitted, analyzed, and calculated, thereby guaranteeing the accuracy of judgment and analysis of the power grid edge and improving the processing efficiency of the edge processor.

[0072] This invention provides a method for preprocessing edge video data based on an edge-side processor, such as... Figure 2 As shown, the method for preprocessing edge video data based on an edge-side processor specifically includes:

[0073] Step S101: Extract edge video data of power grid nodes, determine whether the edge video data is encrypted, if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm, if the edge video data is not encrypted, classify the edge video data as data to be processed.

[0074] For example, based on the decryption algorithm, edge video data is decrypted using the TF32A09 security chip. At the same time, after all edge video data has been decrypted, the TF32A09 security chip is turned off, thereby ensuring the security and privacy of edge video data transmission.

[0075] Step S102: Obtain the decrypted edge video data to be processed, integrate the decrypted edge video data to be processed to obtain a video stream integration set, and filter the video stream integration set through the edge-side processor to obtain a high-definition video stream integration set.

[0076] Step S103: Extract the high-definition video stream aggregation set, compress the high-definition video stream aggregation set, and perform 5G protocol conversion.

[0077] In step S103, as Figure 3 As shown, the compression and 5G protocol conversion of the integrated high-definition video stream specifically includes:

[0078] Step S1031: Based on HEVC ultra-low latency compression mode, read the high-definition video stream integration set;

[0079] Step S1032: Establish a hybrid compression framework based on HEVC ultra-low latency compression, and obtain predictive compression of the integrated video stream set based on the hybrid compression framework;

[0080] Step S1033: Eliminate the correlation between the time domain and spatial domain of the video stream aggregation set, and perform 5G protocol conversion on the HEVC compressed video stream aggregation set.

[0081] For example, by using HEVC ultra-low latency compression to establish a hybrid compression framework, video streams can be compressed within a limited time. The compression records and compressed files can also be retained, so that the compressed files can be retrieved after the compression task is completed, avoiding file loss. It also eliminates the need to upload the complete video stream that consumes a lot of memory, thereby reducing the system load and improving the efficiency of video stream analysis and calculation.

[0082] This invention provides a method for converting preprocessed edge video data to obtain a converted image set, such as... Figure 4 As shown, the method for converting preprocessed edge video data to obtain a converted image set specifically includes:

[0083] Step S201: Receive preprocessed edge video data and obtain the total number of frames of the edge video data;

[0084] Step S202: The total number of frames in the edge video data is uniformly divided to obtain a uniform set of frame intervals;

[0085] Step S203: Obtain the frame set corresponding to each set of frame number intervals;

[0086] Step S204: Based on a preset threshold for a range greater than 1450 pixels x 1450 pixels, filter the frame sets corresponding to each set of frame number ranges.

[0087] Step S205: Obtain the filtered frame set, and measure and judge the quality of the frames in the obtained frame set. Specifically, the THRESH BINARY threshold or the inverse binary threshold THRESH BINARY INV is used to measure and judge the quality of the frames.

[0088] In this embodiment, the total number of frames of the edge video data is uniformly divided, and the number of division intervals is set to t. The total number of frames of the edge video data is M. Then, the number of frames S corresponding to each group of division intervals is:

[0089] (1).

[0090] For example, a preset threshold for the frame range greater than 1450 pixels x 1450 pixels is used to filter the frame set corresponding to each frame range set. The filtering rule is to traverse all frames in the frame set and then determine whether the resolution of the frame is greater than 1450 pixels x 1450 pixels. If it is less than 1450 pixels x 1450 pixels, the frame is discarded, and the frames greater than 1450 pixels x 1450 pixels are retained to form a new set. This filtering rule can effectively filter out invalid and blurry frames, reducing the load on the system's operation.

[0091] This invention provides a method for establishing an image decomposition model, such as... Figure 5 As shown, the method for establishing the image decomposition model specifically includes:

[0092] Step S301: Establish a first disassembly dataset and a second disassembly dataset for disassembling image objects, wherein the first disassembly dataset contains multiple first frame images labeled with disassembly codes, and the second disassembly dataset contains multiple second frame images without disassembly labels.

[0093] Step S302: Predict the disassembly label of the second frame in the second disassembly dataset using the first disassembly dataset to obtain the disassembly label corresponding to the second frame, and obtain the third frame labeled with the disassembly label based on the second frame and its corresponding disassembly label to establish the third disassembly dataset.

[0094] Step S303: Repeat the above steps n times to obtain the nth decomposition dataset. Use the first decomposition dataset and the nth decomposition dataset to train a preset deep neural network to learn an image decomposition model.

[0095] In this embodiment, a deep neural network (NN) is a complex neural network system formed by extensive interconnection of a large number of simple processing units (called neurons). It reflects many fundamental characteristics of human brain function and is a highly complex nonlinear dynamic learning system. Simply put, it is a mathematical model. In this embodiment, a preset deep neural network is trained using the first decomposition dataset and the nth decomposition dataset to learn an image decomposition model.

[0096] It is important to note that the objects to be disassembled here can be various physical objects, such as transmission lines, substations, distribution lines, safety monitoring nodes, power grid nodes with different attributes, distribution cabinets with different shapes, and distribution networks with different voltage attributes. These objects can be captured as digital images by the camera device in any category and can be identified and disassembled using an image disassembly model.

[0097] This invention provides a method for extracting and encoding the decomposed edge video feature points, such as... Figure 6 As shown, the method for extracting and encoding the decomposed edge video feature points specifically includes:

[0098] Step S401: Determine the energy response results of pixels in the decomposed image at multiple preset response orientations;

[0099] Step S402: Determine the direction selection result of the feature point based on the energy response result and the preset feature point selection rules;

[0100] Step S403: Integrate the energy response results and the direction selection results of the feature points selected by the preset feature point selection rules, and assign codes to the edge video feature points.

[0101] In this embodiment, the edge video feature point coding process specifically involves coding the significant feature points in the decomposed image frames. By coding, image frames containing significant feature points can be marked, which facilitates the reverse extraction of power grid node edge video data through edge video feature point analysis.

[0102] This invention provides an analysis method for edge video analysis models, such as... Figure 7 As shown, the analysis method of the edge video analysis model specifically includes:

[0103] Step S501: Obtain the edge video feature points after feature point encoding processing;

[0104] Step S502: Load the edge video data of the power grid node corresponding to the edge video feature points into a buffer based on the edge video analysis model;

[0105] Step S503: Merge the edge video data of the power grid node with the center video data of the power grid node to obtain the integrated edge video analysis results of the power grid node.

[0106] This invention provides an intelligent service system for edge video analytics oriented towards the power grid edge, such as... Figure 8 As shown, the edge video analytics intelligent service system for the power grid edge side specifically includes:

[0107] The edge video data acquisition module 100 is used to acquire edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor;

[0108] The image set acquisition module 200 acquires the preprocessing results of edge video data, converts the preprocessed edge video data, and obtains the converted image set.

[0109] The feature point coding module 300, based on the image decomposition model, decomposes the converted image set, extracts the edge video feature points after decomposition, and performs coding processing on the edge video feature points.

[0110] The edge video analysis result acquisition module 400 is used to acquire the coded edge video feature points. Using the coded edge video feature points as input, the module analyzes and calculates the edge video feature points through a preset edge video analysis model to obtain the edge video analysis results of the power grid node.

[0111] In this embodiment, the edge processor is a wireless data transmission camera with a built-in front-end CPU. At least three sets of edge processors are set for each power grid node, thereby ensuring accurate analysis of the edge side of the node and improving the efficiency of edge side analysis for power grid node monitoring.

[0112] Meanwhile, in this embodiment, the edge-side processor not only sends the edge video data of the power grid node, but also records at least one set of center video data of the power grid node, thereby realizing full computational analysis of the power grid node. For example, the edge video data format of the power grid node includes, but is not limited to, avi, mp4, MPEG-7, MPEG-21, and WMV formats.

[0113] For example, the edge processor uses Python to acquire video streams via OpenCV; when the monitoring target is a power grid node, the power grid node is marked in the video with a red rectangle; then the edge processor extracts the feature information, uploads the feature information, and uses an edge video analysis model to analyze and calculate the edge video feature points to obtain the edge video analysis results of the power grid node.

[0114] For example, the edge processor terminal also has communication connections to a camera, RF (Radio Frequency) circuit, sensors, audio circuit, WiFi module, etc., including sensors such as light sensors, motion sensors, and other sensors.

[0115] This invention, through the encoding of edge video feature points, ensures that the original video data can be accurately located, transmitted, analyzed, and calculated, thereby guaranteeing the accuracy of judgment and analysis of the power grid edge and improving the processing efficiency of the edge processor.

[0116] This invention provides a converted image set acquisition module 200, such as... Figure 9 As shown, the edge video data acquisition module 100 specifically includes:

[0117] The video data encryption judgment unit 110 is used to extract edge video data of power grid nodes, determine whether the edge video data is encrypted, and if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm; if the edge video data is not encrypted, classify the edge video data as data to be processed.

[0118] The video stream integration set acquisition unit 120 acquires the decrypted edge video data to be processed, integrates the decrypted edge video data to be processed, and obtains a video stream integration set. The edge-side processor filters the video stream integration set to obtain a high-definition video stream integration set.

[0119] The integration and compression unit 130 is used to extract high-definition video stream integration sets, compress the high-definition video stream integration sets, and perform 5G protocol conversion.

[0120] This invention provides a converted image set acquisition module 200, such as... Figure 10 As shown, the image set acquisition module 200 specifically includes:

[0121] The total frame count acquisition unit 210 is used to receive preprocessed edge video data and acquire the total frame count of the edge video data.

[0122] The video data segmentation unit 220 is used to uniformly segment the total number of frames of the edge video data to obtain a uniform set of frame number intervals.

[0123] The frame set acquisition unit 230 is used to acquire the frame set corresponding to each set of frame number intervals.

[0124] The frame set filtering unit 240 filters the frame sets corresponding to each set of frame number intervals based on a preset threshold greater than 1450 pixels x 1450 pixels.

[0125] The frame quality discrimination unit 250 acquires the filtered frame set and measures and judges the quality of the frames in the acquired frame set. The frame quality is measured and judged by using the THRESH BINARY threshold or the inverse binary threshold THRESHBINARY INV.

[0126] The present invention provides a schematic diagram of the structure of a computer device, which includes a display screen 21, a memory 22, a processor 24, and a computer program 25. The memory 22 stores the computer program 25, and when the computer program 25 is executed by the processor 24, the processor 24 performs the steps of the face anti-recognition method.

[0127] It is understood that, in the preferred embodiments provided by the present invention, the computer device may also be a laptop computer, a personal digital assistant (PDA), a mobile phone, or other devices capable of communication.

[0128] The present invention also provides a schematic diagram of a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the edge video analysis intelligent service method for the power grid edge side.

[0129] For example, a computer program can be divided into one or more modules, one or more of which are stored in memory and executed by a processor to complete the present invention. One or more modules can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a terminal device. For example, the aforementioned computer program can be divided into units or modules of the edge video analytics intelligent service system for the power grid edge side provided in the various system embodiments described above.

[0130] Those skilled in the art will understand that the above description of the terminal device is merely an example and does not constitute a limitation on the terminal device. It may include more or fewer components than described above, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0131] The processor referred to can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the terminal device, connecting various parts of the user terminal via various interfaces and lines.

[0132] The aforementioned memory can be used to store computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned terminal device by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as information collection template display function, product information publishing function, etc.); the data storage area may store data created based on the use of the edge video analysis intelligent service system for the power grid edge side (such as product information collection templates corresponding to different product types, product information that different product providers need to publish, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0133] If the modules / units integrated into the terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the modules / units in the systems of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the functions of the various system embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0134] In summary, this invention provides an intelligent service method for edge video analysis at the power grid edge. By encoding edge video feature points, this embodiment ensures that the original video data can be accurately located, transmitted, and analyzed, thereby guaranteeing the accuracy of judgment and analysis at the power grid edge and improving the processing efficiency of the edge processor.

[0135] It should be noted that, for the sake of simplicity, the foregoing embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to the present invention. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0136] It should be understood that the disclosed apparatus can be implemented in other ways, given the several embodiments provided in this application. For example, the apparatus embodiments described above are merely illustrative; the division of units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or communication connections shown or discussed may be through some interfaces; the indirect coupling or communication connections between devices or units may be telecommunications or other forms.

[0137] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on these embodiments, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still combine, add, delete, or otherwise adjust the features of the various embodiments of the present invention according to the circumstances without conflict or creative effort, thereby obtaining different technical solutions that do not fundamentally depart from the concept of the present invention. These technical solutions also fall within the scope of protection of the present invention.

Claims

1. An edge video analysis intelligent service method for the power grid edge side, for power grid node edge video analysis, characterized in that, The edge video analytics intelligent service method for the power grid edge side includes: Acquire edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor; Obtain the preprocessing results of edge video data, and transform the preprocessed edge video data to obtain a transformed image set; Based on the image decomposition model, the converted image set is decomposed, the edge video feature points after decomposition are extracted, and the edge video feature points are encoded. The edge video feature points after encoding are obtained. Using the encoded edge video feature points as input, the edge video feature points are analyzed and calculated through a preset edge video analysis model to obtain the edge video analysis results of the power grid node. The method for establishing the image decomposition model specifically includes: Establish a first disassembly dataset and a second disassembly dataset for disassembling image objects, wherein the first disassembly dataset contains multiple first frame images labeled with disassembly codes, and the second disassembly dataset contains multiple second frame images without disassembly labels; The first disassembly dataset is used to predict the disassembly label of the second frame in the second disassembly dataset to obtain the disassembly label corresponding to the second frame. Then, the third frame labeled with the disassembly label is obtained based on the second frame and its corresponding disassembly label to establish the third disassembly dataset. Repeat the above steps n times to obtain the nth decomposition dataset. Use the first decomposition dataset and the nth decomposition dataset to train a preset deep neural network and learn an image decomposition model. The method for extracting and encoding the edge video feature points after decomposition specifically includes: Determine the energy response results of pixels in the decomposed image at multiple preset response orientations; The direction selection result of the feature point is determined based on the energy response result and the preset feature point selection rule; The energy response results and the direction selection results of the feature points selected by the preset feature point selection rules are integrated to encode the edge video feature points. 2.The grid edge side-oriented edge video analytics intelligent service method of claim 1, wherein, The method for preprocessing edge video data based on an edge-side processor specifically includes: Extract edge video data of power grid nodes, determine whether the edge video data is encrypted, if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm, if the edge video data is not encrypted, classify the edge video data as data to be processed; The decrypted edge video data is obtained as the edge video data to be processed. The decrypted edge video data is then integrated to obtain a video stream integration set. The video stream integration set is then filtered by the edge-side processor to obtain a high-definition video stream integration set. Extract high-definition video streams and compress and convert them to 5G protocols. 3.The grid edge side-oriented edge video analytics intelligent service method of claim 2, wherein, The method for converting preprocessed edge video data to obtain a converted image set specifically includes: Receive preprocessed edge video data and obtain the total number of frames in the edge video data; The total number of frames in the edge video data is uniformly divided to obtain a uniform set of frame intervals; Obtain the frame set corresponding to each set of frame number intervals; Based on a preset threshold for a range greater than 1450 pixels x 1450 pixels, the frame sets corresponding to each set of frame ranges are filtered. 4.The grid edge side-oriented edge video analytics intelligent service method of claim 3, wherein, The method for converting preprocessed edge video data to obtain a converted image set further includes: The filtered frame set is obtained, and the quality of the frames in the obtained frame set is measured and judged. Specifically, the THRESH BINARY threshold or the inverse binary threshold THRESH BINARY INV is used to measure and judge the quality of the frames.

5. The edge video analytics intelligent service method for the power grid edge side as described in claim 3, characterized in that, The compression and 5G protocol conversion of the integrated high-definition video streams specifically include: Based on HEVC ultra-low latency compression mode, read and integrate high-definition video streams; A hybrid compression framework is established based on HEVC ultra-low latency compression, and predictive compression of the integrated video stream set is obtained based on the hybrid compression framework. Eliminate the correlation between the time and spatial domains of the video stream aggregation set, and perform 5G protocol conversion on the HEVC-compressed video stream aggregation set.

6. The edge video analytics intelligent service method for the power grid edge side as described in any one of claims 1-5, characterized in that, The analysis method of the edge video analysis model specifically includes: Obtain edge video feature points after feature point encoding processing; The edge video analysis model is used to buffer and load the edge video data of the power grid nodes corresponding to the edge video feature points; By merging the edge video data of the power grid node with the center video data of the power grid node, the integrated edge video analysis results of the power grid node are obtained.

7. An edge video analysis intelligent service system for the power grid edge side based on the edge video analysis intelligent service method for the power grid edge side according to any one of claims 1-6, characterized in that, The edge video analytics intelligent service system for the power grid edge side includes: The edge video data acquisition module is used to acquire edge video data of power grid nodes and preprocess the edge video data based on the edge-side processor. The image set acquisition module acquires the preprocessing results of edge video data, converts the preprocessed edge video data, and obtains the converted image set. The feature point coding module, based on the image decomposition model, decomposes the converted image set, extracts the edge video feature points after decomposition, and assigns codes to the edge video feature points. The edge video analysis result acquisition module is used to acquire the coded edge video feature points. Using the coded edge video feature points as input, the module analyzes and calculates the edge video feature points through a preset edge video analysis model to obtain the edge video analysis results of the power grid nodes.

8. The edge video analytics intelligent service system for the power grid edge side as described in claim 7, characterized in that, The edge video data acquisition module includes: The video data encryption judgment unit is used to extract edge video data of power grid nodes, determine whether the edge video data is encrypted, and if the edge video data is encrypted, decrypt the edge video data based on the decryption algorithm; if the edge video data is not encrypted, the edge video data is classified as data to be processed. The video stream integration set acquisition unit acquires the decrypted edge video data to be processed, integrates the decrypted edge video data to be processed, and obtains the video stream integration set. The edge-side processor filters the video stream integration set to obtain a high-definition video stream integration set. The integrated compression unit is used to extract high-definition video stream integrated sets, compress the high-definition video stream integrated sets, and perform 5G protocol conversion.

Images