Power equipment remote monitoring method based on Beidou technology

Through Beidou technology, the block processing and optical flow analysis of the video frame of power equipment are extracted, and the feature vectors are performed, differential block detection and encoding are solved, and the problem of low monitoring efficiency of traditional power equipment is realized, and the efficient remote monitoring and abnormal behavior of power equipment is realized. Inspection improves the safety and reliability of the system.

CN118942040APending Publication Date: 2024-11-12STATE GRID BEIJING ELECTRIC POWER CO +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411079532.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-07
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Traditional power equipment monitoring relies on manual inspection and simple monitoring systems, and there are problems such as untimely acquisition of information, limited monitoring range, and inefficient efficiency, and it is difficult to effectively detect abnormal behaviors.

Method used

Using the remote monitoring method of power equipment based on Beidou technology, the video frame is blocked and optical flow analysis is performed, feature vectors are extracted, differential block detection and encoding are performed, and the Beidou short message is generated for transmission is realized to reduce the dimension and abnormality of video data. Behavior detection.

Benefits of technology

It improves the efficiency and accuracy of power equipment monitoring, can detect abnormal behaviors in real time, reduce transmission resource losses, and enhances the safety and reliability of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118942040A_ABST
    Figure CN118942040A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of power equipment monitoring, and discloses a power equipment remote monitoring method based on the Beidou technology, and the method comprises the steps: carrying out the blocking processing and feature extraction of all video frames of a power equipment remote monitoring video; performing difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the power equipment; encoding the image blocks marked as the differential blocks in the video frame differential block sequence twice to obtain a Beidou short message; and transmitting the Beidou short message to a background server for remote monitoring by using a Beidou satellite. According to the method, image blocks marked as differential blocks are extracted, an abnormal image area sequence of the power equipment is extracted, dimensionality reduction is performed on video data, semantic information of image content is coded by adopting a spatial sampling and residual calculation mode, and secondary coding is performed based on carrier representation of a Beidou satellite transmitted signal. A Beidou short message is formed to realize remote monitoring.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of remote monitoring, and in particular to a method for remote monitoring of electric power equipment based on Beidou technology. Background Art

[0002] With the development of modern society, the safe and stable operation of the power system as the lifeline of the national economy and important infrastructure is particularly important. Traditional power equipment monitoring mainly relies on manual inspections and simple monitoring systems, which have problems such as untimely information acquisition, limited monitoring range, and low monitoring efficiency. In recent years, with the development of domestic information technology and communication technology, remote monitoring technology has gradually been applied to power systems, greatly improving the efficiency and accuracy of power equipment monitoring. The Beidou satellite navigation system not only provides accurate positioning, navigation and timing services, but also shows great application potential in the fields of communication and monitoring. Using Beidou technology for remote monitoring of core key power equipment can realize real-time monitoring of equipment status, accurate data transmission and intelligent management of equipment, while ensuring data link security and improving the safety and reliability of the power system. Summary of the invention

[0003] In view of this, the present invention provides a method for remote monitoring of electric power equipment based on Beidou technology, the purpose of which is to: 1) perform block processing on video frames in the remote monitoring video of electric power equipment, and perform optical flow analysis on each image block to obtain the optical flow value of each image block, combine the temporal perception information of the same image block position, fuse the optical flow value of the image block in the previous video frame with the optical flow value of the current image block, and use the fused optical flow value to form a feature vector of the image block, so as to realize the fusion processing of the feature information of the video frame, and characterize the image features by using the optical flow information of the pixel, wherein the richer the optical flow information, the faster the speed of the object moving in the image, and the greater the probability of abnormal behavior; 2) perform difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of electric power equipment, wherein the greater the difference, the greater the probability of abnormal behavior at the image block position, Or the greater the difference in semantic information between adjacent video frames, the greater the probability of abnormal behavior between adjacent video frames, and the image blocks marked as difference blocks are extracted to form a video frame difference block sequence, so as to realize the extraction of abnormal image area sequence in the remote monitoring video of power equipment, reduce the dimension of video data, adopt different spatial sampling point arrays for image blocks at different positions, and perform deformation control on the convolution results of image blocks to improve the encoding effect of image content in the preset power equipment position area, and encode the semantic information of image content by spatial sampling and residual calculation, and perform secondary encoding on the encoding result based on the carrier representation of the Beidou satellite transmission signal to form a Beidou short message with lower data dimension, reduce the loss of transmission resources, and use Beidou satellite to transmit Beidou short message to the background server, and the background server decodes the Beidou short message to obtain the video frame difference block sequence to realize remote monitoring.

[0004] To achieve the above object, the present invention provides a method for remote monitoring of electric power equipment based on Beidou technology, comprising the following steps:

[0005] S1: Collecting remote monitoring video of the power equipment to be monitored, performing block processing and feature extraction on all video frames of the remote monitoring video of the power equipment, and obtaining a video frame block feature vector sequence of the monitoring video of the power equipment;

[0006] S2: Perform difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the power equipment, and mark the difference blocks of the next video frame to form a video frame difference block sequence;

[0007] S3: Encode the image blocks marked as difference blocks in the video frame difference block sequence to obtain encoded video frame difference block encoding data, and form a video frame difference block encoding data sequence, wherein deep content-aware view encoding is the main implementation method of image block encoding;

[0008] S4: re-encode the video frame difference block coded data sequence into a Beidou short message, and transmit the Beidou short message to the background server by using the Beidou satellite. The background server decodes the Beidou short message to obtain the video frame difference block sequence for remote monitoring.

[0009] As a further improvement method of the present invention:

[0010] Optionally, in step S1, collecting the electric power equipment remote monitoring video of the electric power equipment to be monitored, and performing block processing on all video frames of the electric power equipment remote monitoring video, includes:

[0011] Collect the remote monitoring video of the power equipment to be monitored, where the collected remote monitoring video of the power equipment is in the form of:

[0012] I=(I1,I2,...,I n ,...,I N )

[0013] in:

[0014] I represents the electric equipment remote monitoring video of the electric equipment to be monitored;

[0015] I1,I2,...,I n ,...,I N represents the frame sequence in the remote monitoring video I of the power equipment, I n represents the nth video frame in the remote monitoring video I of the electric power equipment, and N represents the number of video frames of the remote monitoring video I of the electric power equipment;

[0016] All video frames of the remote monitoring video of the power equipment are processed in blocks, wherein the video frame I n The block processing flow is:

[0017] The video frame I n The image is divided into 9 equal-sized blocks, where the segmentation results are:

[0018]

[0019] in:

[0020] Represents the video frame I n The 9 image blocks obtained by segmentation are: represents the mth image block obtained by segmentation, m∈[1,9];

[0021] Feature extraction is performed on the video frame after block processing to obtain a video frame block feature vector corresponding to the video frame.

[0022] Optionally, the step of performing feature extraction on the video frame after the block processing to obtain a video frame block feature vector corresponding to the video frame includes:

[0023] The feature extraction is performed on the video frame after the block processing to obtain the video frame block feature vector corresponding to the video frame, where the video frame I n The feature extraction process is:

[0024] S11: Video frame I n Each image block in is grayed out, where the image block The grayscale processing formula is:

[0025]

[0026] in:

[0027] Represents an image block The grayscale processing result is Represents an image block The pixel at row x and column y Grayscale value, x∈[1,X], y∈[1,Y], X represents the number of pixel rows in the image block, and Y represents the number of pixel columns in the image block;

[0028] Represents image blocks in sequence The color value of the pixel in the xth row and yth column in the RGB color channel;

[0029] S12: filtering the grayscaled image blocks. The filtering formula is:

[0030]

[0031] in:

[0032] Represents an image block The filtering result is Represents grayscale value The filtering result of

[0033] Represents an image block China-Israel The standard deviation of the grayscale value of the 3×3 neighborhood centered at is, Represents an image block China-Israel The mean gray value of the 3×3 neighborhood centered at ;

[0034] exp(·) represents an exponential function with a natural constant as the base;

[0035] S13: Calculate the gradient value of each pixel, where pixel The gradient value of is:

[0036]

[0037] in:

[0038] Represents pixels The gradient value of

[0039] S14: Calculate the optical flow value of each pixel, where the pixel The optical flow value is:

[0040]

[0041] in:

[0042] Represents pixels The optical flow value of , T represents transposition;

[0043] In pixels The gradient value of the eight neighboring pixels centered at Represents pixels Neighborhood gradient sequence of ;

[0044] In pixels The eight neighboring pixels and pixels centered on The Euclidean distance between them, diag(·) represents the diagonal matrix, Represents pixels Neighborhood distance matrix of ;

[0045] In pixels The gradient change of the eight neighboring pixels centered at Represents pixels The sequence of neighborhood gradient changes;

[0046] S15: The optical flow value of the pixel is fused with the temporal perception information to obtain the fused optical flow value of the pixel, where the optical flow value The fusion processing formula is:

[0047]

[0048] in:

[0049] Represents pixels The fusion optical flow value of

[0050] S16: Constructing video frame I n The corresponding video frame block feature vector f n :

[0051]

[0052] in:

[0053] Represents video frame I n Middle image block The corresponding feature vector.

[0054] Optionally, in step S2, performing difference block detection on video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the electric power equipment includes:

[0055] The video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of power equipment are detected by difference block, where adjacent video frame I n ,I n+1 The corresponding video frame block feature vector f n ,f n+1 The difference block detection process is as follows:

[0056] S21: Divide the video frame into blocks with feature vector f n ,f n+1 The feature vectors of different image blocks in the image are used to extract semantic information, where the feature vector The semantic information extraction formula is:

[0057]

[0058] in:

[0059] Represents the feature vector Semantic information in

[0060] * indicates convolution calculation;

[0061] w1 represents the convolution weight matrix;

[0062] Represents the feature vector The attention weight of

[0063] S22: Calculate the video frame block feature vector f n ,f n+1 The difference in semantic information corresponding to the image block at the same position in The differences between are:

[0064]

[0065] in:

[0066] Representing semantic information The difference between

[0067] ||·||2 represents the L2 norm;

[0068] According to the difference calculation result, the difference blocks of the next video frame are marked to form a video frame difference block sequence.

[0069] Optionally, the difference blocks of the next video frame after marking constitute a video frame difference block sequence, including:

[0070] like If the difference is greater than the preset threshold, the video frame I n+1 The mth image block in Mark as difference blocks; according to the difference block marking results of all video frames, extract the image blocks marked as difference blocks of the video frames to form a video frame difference block sequence:

[0071] Ω=(Ω2,...,Ω n ,...,Ω N )

[0072] Ω n ={Ω n (u)|u∈[1,Sum n ]}

[0073] in:

[0074] Ω represents the video frame difference block sequence;

[0075] Ω n Represents video frame I n A set of image patches marked as difference patches in ;

[0076] Ω n (u) represents the image block set Ω n The u-th image block in n Represents video frame I n The total number of image patches marked as difference patches in .

[0077] Optionally, in step S3, encoding the image blocks marked as difference blocks in the video frame difference block sequence to form a video frame difference block encoding data sequence includes:

[0078] The image blocks marked as difference blocks in the video frame difference block sequence are encoded to obtain the encoded video frame difference block encoded data, where the image block Ω n The encoding process of (u) is:

[0079] S31: Using deformable convolution to image block Ω n (u) Perform convolution processing:

[0080]

[0081] in:

[0082] loc(Ω n (u)) represents the image block Ω n (u) is the position in the video frame, loc(Ω n (u))∈[1,9];

[0083] Indicates that it is at the spatial position loc(Ω n (u)) is a spatial sampling matrix of an image block, wherein the spatial sampling matrix is ​​in the form of a matrix with 3 rows and 3 columns, and the values ​​of the elements in the matrix are {-1, 0, 1};

[0084] w2 represents the convolution weight matrix;

[0085] Q n (u) represents the image block Ω n The convolution result of (u);

[0086] S32: Use a multi-scale deformable downsampling module to process the convolution result Q n (u) Perform spatial sampling, where the spatial sampling process is: use a 3×3 pixel convolution kernel with a convolution step size of 2 to Q n (u) is convolved to obtain Q n (u) Downsampling result in spatial dimension

[0087] S33: Perform residual calculation connection on the downsampling result to obtain the image block Ω n The encoding result of (u) is:

[0088]

[0089] in:

[0090] H n (u) represents the image block Ω n The encoding result of (u), w 1×1 represents the convolution weight matrix of 1×1 pixels, w 3×3 represents a convolution weight matrix of 3×3 pixels, and δ(·) represents an activation function. In the embodiment of the present invention, the selected activation function is a ReLU function.

[0091] Optionally, the video frame difference block encoding data sequence is:

[0092] H=(H2,...,Hn ,...,H N )

[0093] H n ={H n (u)|u∈[1,Sum n ]}

[0094] in:

[0095] H is the video frame difference block coding data sequence;

[0096] H n Represents the image block set Ω n The set of encoding results of all image blocks in .

[0097] Optionally, in the step S4, the video frame difference block coded data sequence is secondary encoded into a Beidou short message, and the Beidou short message is transmitted to the background server by using the Beidou satellite, including:

[0098] The video frame difference block encoding data sequence is re-encoded into Beidou short message, and the re-encoding process is as follows:

[0099] S41: Generate a random code number, a timestamp of the message generation time, and the number of the monitored power equipment;

[0100] S42: The encoding results of the image blocks in the video frame difference block encoding data sequence are sequentially re-encoded, wherein H n The secondary encoding formula of (u) is:

[0101]

[0102] in:

[0103] ω represents the carrier modulation frequency;

[0104] H n (u,v) represents the encoding result H n The vth data value in (u), V represents the length of the encoding result;

[0105] h n (u) indicates H n (u) secondary encoding result;

[0106] S43: The generated random code number, the message generation time stamp, the number of the monitored power equipment and the secondary encoding result of the encoding result in the video frame difference block encoding data sequence constitute a Beidou short message;

[0107] Beidou satellites are used to transmit Beidou short messages to the backend server, which decodes the Beidou short messages to obtain video frame difference block sequences for remote monitoring.

[0108] In order to solve the above problem, the present invention provides an electronic device, the electronic device comprising:

[0109] A memory storing at least one instruction;

[0110] Communication interface, enabling electronic equipment to communicate; and

[0111] The processor executes the instructions stored in the memory to implement the above-mentioned Beidou technology-based remote monitoring method for power equipment.

[0112] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned Beidou technology-based remote monitoring method for power equipment.

[0113] Compared with the prior art, the present invention proposes a remote monitoring method for power equipment based on Beidou technology, which has the following advantages:

[0114] First, this scheme proposes a feature vector extraction method based on optical flow information and a difference analysis method, which fuses the optical flow value of the pixel with the temporal perception information to obtain the fused optical flow value of the pixel, where the optical flow value The fusion processing formula is:

[0115]

[0116] in: Represents pixels The fused optical flow value constitutes the video frame I n The corresponding video frame block feature vector f n :

[0117]

[0118] in: Represents video frame I n Middle image block The video frame block feature vectors corresponding to the adjacent video frames in the remote monitoring video of the power equipment are subjected to difference block detection, where the adjacent video frame I n ,I n+1 The corresponding video frame block feature vector f n ,f n+1 The difference block detection process is as follows: the video frame block feature vector f n ,f n+1 The feature vectors of different image blocks in the image are used to extract semantic information, where the feature vector The semantic information extraction formula is:

[0119]

[0120] in: Represents the feature vector The semantic information in; * indicates convolution calculation; w1 indicates the convolution weight matrix;

[0121] Represents the feature vector Attention weight; Calculate the video frame block feature vector f n ,f n+1 The difference in semantic information corresponding to the image block at the same position in The differences between are:

[0122]

[0123] in: Representing semantic information The difference between them; ||·||2 represents the L2 norm; according to the difference calculation result, the difference blocks of the next video frame are marked to form a video frame difference block sequence. This scheme processes the video frames in the remote monitoring video of the power equipment by dividing them into blocks, and performs optical flow analysis on each image block to obtain the optical flow value of each image block. Combined with the temporal perception information of the same image block position, the optical flow value of the image block in the previous video frame is fused with the optical flow value of the current image block, and the fused optical flow value constitutes the feature vector of the image block to realize the fusion processing of the video frame feature information, and uses the optical flow information of the pixel to characterize the image features, wherein the richer the optical flow information, the faster the object moves in the image, and the greater the probability of abnormal behavior. The video frame block feature vectors corresponding to the adjacent video frames in the remote monitoring video of the power equipment are detected by difference block, wherein the greater the difference, the greater the probability of abnormal behavior at the position of the image block, or the greater the difference in semantic information of adjacent video frames, the greater the probability of abnormal behavior between adjacent video frames, and the image blocks marked as difference blocks are extracted to form a video frame difference block sequence, so as to realize the extraction of abnormal image area sequence in the remote monitoring video of the power equipment and reduce the dimension of the video data.

[0124] At the same time, this scheme proposes an image block encoding method, which encodes the image blocks marked as difference blocks in the video frame difference block sequence to obtain the encoded video frame difference block encoding data, where the image block Ω n The encoding process of (u) is as follows: deformable convolution is used to encode the image block Ω n (u) Perform convolution processing:

[0125]

[0126] Where: loc(Ω n (u)) represents the image block Ω n (u) is the position in the video frame, loc(Ω n (u))∈[1,9];

[0127] Indicates that it is at the spatial position loc(Ω n (u)), where the spatial sampling matrix is ​​a matrix of 3 rows and 3 columns, and the values ​​of the elements in the matrix are {-1, 0, 1}; w2 represents the convolution weight matrix; Q n (u) represents the image block Ω n (u) is the convolution result; a multi-scale deformable downsampling module is used to convolution the result Q n (u) Perform spatial sampling, where the spatial sampling process is: use a 3×3 pixel convolution kernel with a convolution step size of 2 to Q n (u) is convolved to obtain Q n (u) Downsampling result in spatial dimension Perform residual calculation on the downsampling result to obtain the image block Ω n The encoding result of (u) is:

[0128]

[0129] Where: H n (u) represents the image block Ω n The encoding result of (u), w 1×1 represents the convolution weight matrix of 1×1 pixels, w 3×3 represents the convolution weight matrix of 3×3 pixels, and δ(·) represents the activation function.

[0130] Compared with the prior art, the present invention has the following beneficial effects:

[0131] This scheme adopts different spatial sampling point arrays for image blocks at different positions, performs deformation control on the convolution results of image blocks, improves the encoding effect of image content in the preset power equipment location area, and encodes the semantic information of the image content by spatial sampling and residual calculation. Based on the carrier representation of the Beidou satellite transmission signal, the encoding result is secondary encoded to form a Beidou short message with lower data dimension, reducing the loss of transmission resources. The Beidou short message is transmitted to the background server by Beidou satellite. The background server decodes the Beidou short message to obtain the video frame difference block sequence to realize remote monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0132] Figure 1 A flowchart of a method for remote monitoring of electric power equipment based on Beidou technology provided by an embodiment of the present invention;

[0133] Figure 2 A schematic diagram of the structure of an electronic device for implementing a method for remote monitoring of power equipment based on Beidou technology provided by an embodiment of the present invention.

[0134] In the figure: 1 electronic device, 10 processor, 11 memory, 12 program, 13 communication interface.

[0135] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0136] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0137] The embodiment of the present application provides a method for remote monitoring of electric power equipment based on Beidou technology. The execution subject of the method for remote monitoring of electric power equipment based on Beidou technology includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the method for remote monitoring of electric power equipment based on Beidou technology can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0138] Embodiment 1:

[0139] S1: Collect the remote monitoring video of the power equipment to be monitored, perform block processing and feature extraction on all video frames of the remote monitoring video of the power equipment, and obtain a video frame block feature vector sequence of the power equipment monitoring video.

[0140] In the step S1, the remote monitoring video of the electric power equipment to be monitored is collected, and all video frames of the remote monitoring video of the electric power equipment are processed in blocks, including:

[0141] Collect the remote monitoring video of the power equipment to be monitored, where the collected remote monitoring video of the power equipment is in the form of:

[0142] I=(I1,I2,...,I n ,...,I N )

[0143] in:

[0144] I represents the electric equipment remote monitoring video of the electric equipment to be monitored;

[0145] I1,I2,...,I n ,...,I Nrepresents the frame sequence in the remote monitoring video I of the power equipment, I n represents the nth video frame in the remote monitoring video I of the electric power equipment, and N represents the number of video frames of the remote monitoring video I of the electric power equipment;

[0146] All video frames of the remote monitoring video of the power equipment are processed in blocks, wherein the video frame I n The block processing flow is:

[0147] The video frame I n The image is divided into 9 equal-sized blocks, where the segmentation results are:

[0148]

[0149] in:

[0150] Represents the video frame I n The 9 image blocks obtained by segmentation are: represents the mth image block obtained by segmentation, m∈[1,9];

[0151] Feature extraction is performed on the video frame after block processing to obtain a video frame block feature vector corresponding to the video frame.

[0152] The step of extracting features from the video frame after the block processing to obtain a video frame block feature vector corresponding to the video frame includes:

[0153] The feature extraction is performed on the video frame after the block processing to obtain the video frame block feature vector corresponding to the video frame, where the video frame I n The feature extraction process is:

[0154] S11: Video frame I n Each image block in is grayed out, where the image block The grayscale processing formula is:

[0155]

[0156] in:

[0157] Represents an image block The grayscale processing result is Represents an image block The pixel at row x and column y Grayscale value, x∈[1,X], y∈[1,Y], X represents the number of pixel rows in the image block, and Y represents the number of pixel columns in the image block;

[0158] Represents image blocks in sequence The color value of the pixel in the xth row and yth column in the RGB color channel;

[0159] S12: filtering the grayscaled image blocks. The filtering formula is:

[0160]

[0161]

[0162] in:

[0163] Represents an image block The filtering result is Represents grayscale value The filtering result of

[0164] Represents an image block China-Israel The standard deviation of the grayscale value of the 3×3 neighborhood centered at is, Represents an image block China-Israel The mean gray value of the 3×3 neighborhood centered at ;

[0165] exp(·) represents an exponential function with a natural constant as the base;

[0166] S13: Calculate the gradient value of each pixel, where pixel The gradient value of is:

[0167]

[0168] in:

[0169] Represents pixels The gradient value of

[0170] S14: Calculate the optical flow value of each pixel, where the pixel The optical flow value is:

[0171]

[0172] in:

[0173] Represents pixels The optical flow value of , T represents transposition;

[0174] In pixels The gradient value of the eight neighboring pixels centered at Represents pixels Neighborhood gradient sequence of ;

[0175] In pixels The eight neighboring pixels and pixels centered on The Euclidean distance between them, diag(·) represents the diagonal matrix, Represents pixels Neighborhood distance matrix of ;

[0176] In pixels The gradient change of the eight neighboring pixels centered at Represents pixels The sequence of neighborhood gradient changes;

[0177] S15: The optical flow value of the pixel is fused with the temporal perception information to obtain the fused optical flow value of the pixel, where the optical flow value The fusion processing formula is:

[0178]

[0179] in:

[0180] Represents pixels The fusion optical flow value of

[0181] S16: Constructing video frame I n The corresponding video frame block feature vector f n :

[0182]

[0183] in:

[0184] Represents video frame I n Middle image block The corresponding feature vector.

[0185] S2: Perform difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the power equipment, and mark the difference blocks of the subsequent video frame to form a video frame difference block sequence.

[0186] In the step S2, the difference block detection is performed on the video frame block feature vectors corresponding to the adjacent video frames in the remote monitoring video of the power equipment, including:

[0187] The video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of power equipment are detected by difference block, where adjacent video frame I n ,I n+1 The corresponding video frame block feature vector fn ,f n+1 The difference block detection process is as follows:

[0188] S21: Divide the video frame into blocks with feature vector f n ,f n+1 The feature vectors of different image blocks in the image are used to extract semantic information, where the feature vector The semantic information extraction formula is:

[0189]

[0190] in:

[0191] Represents the feature vector Semantic information in

[0192] * indicates convolution calculation;

[0193] w1 represents the convolution weight matrix;

[0194] Represents the feature vector The attention weight of

[0195] S22: Calculate the video frame block feature vector f n ,f n+1 The difference in semantic information corresponding to the image block at the same position in The differences between are:

[0196]

[0197] in:

[0198] Representing semantic information The difference between

[0199] ||·||2 represents the L2 norm;

[0200] According to the difference calculation result, the difference blocks of the next video frame are marked to form a video frame difference block sequence.

[0201] The difference blocks of the next video frame after the marking constitute a video frame difference block sequence, including:

[0202] like If the difference is greater than the preset threshold, the video frame I n+1 The mth image block in Mark as difference blocks; according to the difference block marking results of all video frames, extract the image blocks marked as difference blocks of the video frames to form a video frame difference block sequence:

[0203] Ω=(Ω2,...,Ω n ,...,Ω N )

[0204] Ω n ={Ω n (u)|u∈[1,Sum n ]}

[0205] in:

[0206] Ω represents the video frame difference block sequence;

[0207] Ω n Represents video frame I n A set of image patches marked as difference patches in ;

[0208] Ω n (u) represents the image block set Ω n The u-th image block in n Represents video frame I n The total number of image patches marked as difference patches in .

[0209] S3: Encode the image blocks marked as difference blocks in the video frame difference block sequence to obtain encoded video frame difference block encoding data, and form a video frame difference block encoding data sequence.

[0210] In the step S3, the image blocks marked as difference blocks in the video frame difference block sequence are encoded to form a video frame difference block encoding data sequence, including:

[0211] The image blocks marked as difference blocks in the video frame difference block sequence are encoded to obtain the encoded video frame difference block encoded data, where the image block Ω n The encoding process of (u) is:

[0212] S31: Using deformable convolution to image block Ω n (u) Perform convolution processing:

[0213]

[0214] in:

[0215] loc(Ω n (u)) represents the image block Ω n (u) is the position in the video frame, loc(Ω n (u))∈[1,9];

[0216] Indicates that it is at the spatial position loc(Ω n(u)) is a spatial sampling matrix of an image block, wherein the spatial sampling matrix is ​​in the form of a matrix with 3 rows and 3 columns, and the values ​​of the elements in the matrix are {-1, 0, 1};

[0217] w2 represents the convolution weight matrix;

[0218] Q n (u) represents the image block Ω n The convolution result of (u);

[0219] S32: Use a multi-scale deformable downsampling module to process the convolution result Q n (u) Perform spatial sampling, where the spatial sampling process is: use a 3×3 pixel convolution kernel with a convolution step size of 2 to Q n (u) is convolved to obtain Q n (u) Downsampling result in spatial dimension

[0220] S33: Perform residual calculation connection on the downsampling result to obtain the image block Ω n The encoding result of (u) is:

[0221]

[0222] in:

[0223] H n (u) represents the image block Ω n The encoding result of (u), w 1×1 represents the convolution weight matrix of 1×1 pixels, w 3×3 represents a convolution weight matrix of 3×3 pixels, and δ(·) represents an activation function. In the embodiment of the present invention, the selected activation function is a ReLU function.

[0224] The video frame difference block encoding data sequence is:

[0225] H=(H2,...,H n ,...,H N )

[0226] H n ={H n (u)|u∈[1,Sum n ]}

[0227] in:

[0228] H is the video frame difference block coding data sequence;

[0229] H n Represents the image block set Ω n The set of encoding results of all image blocks in .

[0230] S4: re-encode the video frame difference block coded data sequence into a Beidou short message, and transmit the Beidou short message to the background server by using the Beidou satellite. The background server decodes the Beidou short message to obtain the video frame difference block sequence for remote monitoring.

[0231] In the step S4, the video frame difference block encoding data sequence is secondary encoded into a Beidou short message, and the Beidou short message is transmitted to the background server by using the Beidou satellite, including:

[0232] The video frame difference block encoding data sequence is re-encoded into Beidou short message, and the re-encoding process is as follows:

[0233] S41: Generate a random code number, a timestamp of the message generation time, and the number of the monitored power equipment;

[0234] S42: The encoding results of the image blocks in the video frame difference block encoding data sequence are sequentially re-encoded, wherein H n The secondary encoding formula of (u) is:

[0235]

[0236] in:

[0237] ω represents the carrier modulation frequency;

[0238] H n (u,v) represents the encoding result H n The vth data value in (u), V represents the length of the encoding result;

[0239] h n (u) indicates H n (u) secondary encoding result;

[0240] S43: The generated random code number, the message generation time stamp, the number of the monitored power equipment and the secondary encoding result of the encoding result in the video frame difference block encoding data sequence constitute a Beidou short message;

[0241] Beidou satellites are used to transmit Beidou short messages to the backend server, which decodes the Beidou short messages to obtain video frame difference block sequences for remote monitoring.

[0242] Embodiment 2:

[0243] like Figure 2 , is a schematic diagram of the structure of an electronic device for implementing a method for remote monitoring of power equipment based on Beidou technology provided by an embodiment of the present invention.

[0244] The electronic device 1 may include a processor 10 , a memory 11 , a communication interface 13 and a bus, and may also include a computer program stored in the memory 11 and executable on the processor 10 , such as a program 12 .

[0245] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 may be used not only to store application software and various types of data installed in the electronic device 1, such as the code of the program 12, etc., but also to temporarily store data that has been output or is to be output.

[0246] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect the various components of the entire electronic device, and executes or executes the programs or modules stored in the memory 11 (such as the program 12 for realizing remote monitoring of power equipment based on Beidou technology), and calls the data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

[0247] The communication interface 13 may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices, and to achieve connection and communication between internal components of the electronic device.

[0248] The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0249] Figure 2 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 2 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0250] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, etc. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0251] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0252] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0253] The program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0254] Collecting remote monitoring videos of electric power equipment to be monitored, performing block processing and feature extraction on all video frames of the remote monitoring videos of electric power equipment, and obtaining a video frame block feature vector sequence of the monitoring videos of electric power equipment;

[0255] Performing difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the power equipment, and marking the difference block of the next video frame to form a video frame difference block sequence;

[0256] Encoding the image blocks marked as difference blocks in the video frame difference block sequence to obtain encoded video frame difference block encoded data and form a video frame difference block encoded data sequence;

[0257] The video frame difference block coded data sequence is re-encoded into Beidou short message, and the Beidou short message is transmitted to the background server by Beidou satellite. The background server decodes the Beidou short message to obtain the video frame difference block sequence for remote monitoring.

[0258] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figure 1 to Figure 2 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0259] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. And the terms "including", "comprising" or any other variants thereof in this article are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0260] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0261] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A remote monitoring method for power equipment based on Beidou technology, characterized in that: The method comprises: S1: Collecting remote monitoring video of the power equipment to be monitored, performing block processing and feature extraction on all video frames of the remote monitoring video of the power equipment, and obtaining a video frame block feature vector sequence of the monitoring video of the power equipment; S2: Perform difference block detection on the video frame block feature vectors corresponding to adjacent video frames in the remote monitoring video of the power equipment, and mark the difference blocks of the next video frame to form a video frame difference block sequence; S3: Encode the image blocks marked as difference blocks in the video frame difference block sequence to obtain encoded video frame difference block encoded data, and form a video frame difference block encoded data sequence; S4: re-encode the video frame difference block coded data sequence into a Beidou short message, and transmit the Beidou short message to the background server by using the Beidou satellite. The background server decodes the Beidou short message to obtain the video frame difference block sequence for remote monitoring.

2. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 1, characterized in that: In the step S1, the remote monitoring video of the electric power equipment to be monitored is collected, and all video frames of the remote monitoring video of the electric power equipment are processed in blocks, including: Collect the remote monitoring video of the power equipment to be monitored, where the collected remote monitoring video of the power equipment is in the form of: l=(I1,I2,...,l n ...,I N ) in: I represents the electric equipment remote monitoring video of the electric equipment to be monitored; I1,I2,...,I n ,...,I N represents the frame sequence in the remote monitoring video I of the power equipment, I n represents the nth video frame in the remote monitoring video I of the electric power equipment, and N represents the number of video frames of the remote monitoring video I of the electric power equipment; All video frames of the remote monitoring video of the power equipment are processed in blocks, wherein the video frame I n The block processing flow is: The video frame I n The image is divided into 9 equal-sized blocks, where the segmentation results are: in: Represents the video frame I n The 9 image blocks obtained by segmentation are: represents the mth image block obtained by segmentation, m∈[1,9]; Feature extraction is performed on the video frame after block processing to obtain a video frame block feature vector corresponding to the video frame.

3. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 2, characterized in that: The step of extracting features from the video frame after the block processing to obtain a video frame block feature vector corresponding to the video frame includes: Video Frame I n The feature extraction process is: S11: Video frame I n Each image block in is grayed out, where the image block The grayscale processing formula is: in: Represents an image block The grayscale processing result is Represents an image block The pixel at row x and column y Grayscale value, x∈[1,X], y∈[1,Y], X represents the number of pixel rows in the image block, and Y represents the number of pixel columns in the image block; Represents image blocks in sequence The color value of the pixel in the xth row and yth column in the RGB color channel; S12: filtering the grayscaled image blocks. The filtering formula is: in: Represents an image block The filtering result is Represents grayscale value The filtering result of Represents an image block China-Israel The standard deviation of the grayscale value of the 3×3 neighborhood centered at is, Represents an image block China-Israel The mean gray value of the 3×3 neighborhood centered at ; exp(·) represents an exponential function with a natural constant as the base; S13: Calculate the gradient value of each pixel, where pixel The gradient value of is: in: Represents pixels The gradient value of S14: Calculate the optical flow value of each pixel, where the pixel The optical flow value is: in: Represents pixels The optical flow value of , T represents transposition; In pixels The gradient value of the eight neighboring pixels centered at Represents pixels Neighborhood gradient sequence of ; In pixels The eight neighboring pixels centered on The Euclidean distance between them, diag(·) represents the diagonal matrix, Represents pixels Neighborhood distance matrix of ; In pixels The gradient change of the eight neighboring pixels centered at Represents pixels The sequence of neighborhood gradient changes; S15: The optical flow value of the pixel is fused with the temporal perception information to obtain the fused optical flow value of the pixel, where the optical flow value The fusion processing formula is: in: Represents pixels The fusion optical flow value of S16: Constructing video frame I n The corresponding video frame block feature vector f n : in: Represents video frame I n Middle image block The corresponding feature vector.

4. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 3, characterized in that: In the step S2, the difference block detection is performed on the video frame block feature vectors corresponding to the adjacent video frames in the remote monitoring video of the power equipment, including: Adjacent video frame I n ,I n+1 The corresponding video frame block feature vector f n ,f n+1 The difference block detection process is as follows: S21: Divide the video frame into blocks with feature vector f n ,f n+1 The feature vectors of different image blocks in the image are used to extract semantic information, where the feature vector The semantic information extraction formula is: in: Represents the feature vector Semantic information in * indicates convolution calculation; w1 represents the convolution weight matrix; Represents the feature vector The attention weight of S22: Calculate the video frame block feature vector f n ,f n+1 The difference in semantic information corresponding to the image block at the same position in The differences between are: in: Representing semantic information The difference between ||·||2 represents the L2 norm; According to the difference calculation result, the difference blocks of the next video frame are marked to form a video frame difference block sequence.

5. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 4, characterized in that: The difference blocks of the next video frame after the marking constitute a video frame difference block sequence, including: like If the difference is greater than the preset threshold, the video frame I n+1 The mth image block in Mark as difference blocks; according to the difference block marking results of all video frames, extract the image blocks marked as difference blocks of the video frames to form a video frame difference block sequence: Ω=(Ω2,...,Ω n ,...,Oh N ) Oh n ={Ω n (u)|u∈[1,Sum n ]} in: Ω represents the video frame difference block sequence; Ω n Represents video frame I n A set of image patches marked as difference patches in ; Ω n (u) represents the image block set Ω n The u-th image block in n Represents video frame I n The total number of image patches marked as difference patches in .

6. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 1, characterized in that: In the step S3, the image blocks marked as difference blocks in the video frame difference block sequence are encoded to form a video frame difference block encoding data sequence, including: Image block Ω n The encoding process of (u) is: S31: Using deformable convolution to image block Ω n (u) Perform convolution processing: in: loc(Ω n (u)) represents the image block Ω n (u) is the position in the video frame, loc(Ω n (u))∈[1,9]; Indicates that it is at the spatial position loc(Ω n (u)) is a spatial sampling matrix of an image block, wherein the spatial sampling matrix is ​​in the form of a matrix with 3 rows and 3 columns, and the values ​​of the elements in the matrix are {-1, 0, 1}; w2 represents the convolution weight matrix; Q n (u) represents the image block Ω n (u) The convolution result; S32: Use a multi-scale deformable downsampling module to process the convolution result Q n (u) Perform spatial sampling, where the spatial sampling process is: use a 3×3 pixel convolution kernel with a convolution step size of 2 to Q n (u) is convolved to obtain Q n (u) Downsampling result in spatial dimension S33: Perform residual calculation connection on the downsampling result to obtain the image block Ω n The encoding result of (u) is: in: H n (u) represents the image block Ω n The encoding result of (u), w 1×1 represents the convolution weight matrix of 1×1 pixels, w 3×3 represents the convolution weight matrix of 3×3 pixels, and δ(·) represents the activation function.

7. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 6, characterized in that: The video frame difference block encoding data sequence is: H=(H2,...,H n ,...,H N ) H n ={H n (u)|u∈[1,Sum n ]} in: H is the video frame difference block coding data sequence; H n Represents the image block set Ω n The set of encoding results of all image blocks in .

8. A method for remote monitoring of electric power equipment based on Beidou technology as claimed in claim 7, characterized in that: In the step S4, the video frame difference block encoding data sequence is secondary encoded into a Beidou short message, and the Beidou short message is transmitted to the background server by using the Beidou satellite, including: The secondary encoding process is: S41: Generate a random code number, a timestamp of the message generation time, and the number of the monitored power equipment; S42: The encoding results of the image blocks in the video frame difference block encoding data sequence are sequentially re-encoded, wherein H n The secondary encoding formula of (u) is: in: ω represents the carrier modulation frequency; H n (u,v) represents the encoding result H n The vth data value in (u), V represents the length of the encoding result; h n (u) indicates H n (u) secondary encoding result; S43: The generated random code number, the message generation time stamp, the number of the monitored power equipment and the secondary encoding result of the encoding result in the video frame difference block encoding data sequence constitute a Beidou short message; Beidou satellites are used to transmit Beidou short messages to the backend server, which decodes the Beidou short messages to obtain video frame difference block sequences for remote monitoring.