A digital power transmission channel intelligent safety inspection method

Abstract

The application discloses a kind of digital power transmission channel intelligent safety inspection methods, comprising: collecting and statistics power transmission channel is destroyed factor, in the monitoring point of power transmission channel, the image picture of the factor of destroying power transmission channel is collected, the image picture of the factor of destroying power transmission channel collected is handled, extract the overall image feature of the factor of destroying power transmission channel in image picture, this digital power transmission channel intelligent safety inspection method, by collecting the different types and different states of the image picture of the external environmental factor causing damage to power transmission channel, and the image picture is digitally processed, to be used as the model data of power transmission channel monitoring inspection identification and judgment, can realize the intelligent identification and pre-judgment of the possible safety risk of the environment around power transmission channel, early warning to the factor of existing safety risk.The application reduces the safety risk of power transmission channel being destroyed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission channel inspection, in particular to a digital power transmission channel intelligent safety inspection method. BACKGROUND

[0002] Power transmission channel inspection is to reduce the risk of power grid for the normal operation of power grid line. With the rapid development of power grid, the per capita line maintenance length is increasing, and the work efficiency and quality of manual transmission line maintenance cannot be guaranteed. With the development of Internet technology, remote intelligent inspection of power transmission channels can be realized through remote monitoring equipment.

[0003] The existing intelligent inspection method of power transmission channel is to use monitoring equipment to conduct real-time inspection operation on each monitoring point of the power transmission channel. When the power transmission channel has a safety risk, emergency maintenance and repair operation is carried out on the power transmission channel with safety risk. According to statistics, most of the factors causing safety risk of existing power transmission channel are external environmental factors. The existing patent: a method for preventing external damage to the monitoring and shooting power line protection area (application number: 201510282837.4) discloses a power line remote monitoring and inspection method. However, the inspection method can only monitor the environment around the power transmission line, and cannot intelligently identify and predict the safety risks that may exist in the environment around the power transmission channel. The effect of power transmission channel safety inspection is not good. SUMMARY

[0004] In order to solve the problem that the existing power transmission channel inspection method can only monitor the environment around the power transmission line, cannot intelligently identify and predict the safety risks that may exist in the environment around the power transmission channel, and the effect of power transmission channel safety inspection is not good, the purpose of the present application is to provide a digital power transmission channel intelligent safety inspection method, which can intelligently identify and predict the safety risks that may exist in the environment around the power transmission channel, and can give early warning to the factors that exist safety risks, thereby reducing the safety risks of power transmission channel damage.

[0005] The purpose of the present application is achieved by the following technical solutions:

[0006] A digital power transmission channel intelligent safety inspection method comprises the following specific steps:

[0007] S1, collecting and counting the factors causing damage to the power transmission channel;

[0008] S2, collecting the image of the factor causing damage to the power transmission channel at the monitoring point of the power transmission channel;

[0009] S3, processing the collected image of the factor causing damage to the power transmission channel, and extracting the overall image features of the factor causing damage to the power transmission channel in the image;

[0010] S4, establish an image contrast data set, an image contrast sub-database and an image contrast database using the overall image features of the factors extracted in S3;

[0011] S5, establish a feature recognition database using the overall image features of the factors extracted in S3;

[0012] S6, perform similarity contrast analysis on the overall image features of the factors collected to damage the power transmission channel, and establish a similarity contrast analysis sub-database and a similarity contrast analysis database;

[0013] S7, deploy monitoring equipment in the power transmission channel, acquire image pictures of the power transmission channel and its surrounding environment in real time, and identify the image picture information in real time;

[0014] S8, compare and analyze the identified image picture information with the same features of different factors in the feature recognition database, and combine the feature recognition database, the image contrast database and the similarity contrast analysis database to determine whether the image picture information has a security risk of damaging the power transmission channel.

[0015]

[0016] Further, the statistics in S1 also include the proportion of factors that damage the power transmission channel and the seasons in which the damage occurs.

[0017] Further, the image pictures of the factors that damage the power transmission channel in S2 specifically include:

[0018] S201, collect image pictures of all factors that damage the power transmission channel as counted in S1;

[0019] S202, collect different types of image pictures of the same factor, collect different state pictures of the same type of factor, or collect different state pictures of the same factor.

[0020] Further, the specific steps of extracting the overall image features of the factors that damage the power transmission channel in S3 include:

[0021] S301, adjust the size of the image pictures of the factors that damage the power transmission channel based on the HOG algorithm, and perform grayscale, normalization and image segmentation preprocessing;

[0022] S302, perform convolution on each block after image segmentation, and obtain the gradient direction and amplitude of each pixel point in the image:

[0023] ​Wherein, (n, m) is the pixel position, Gn(n, m) = I(n+1, m) - I(n-1, m), Gm(n, m) = I(n, m+1) - I(n, m-1), Gn and Gm represent the gradient values in horizontal and vertical directions, M(n, m) represents the amplitude of the gradient, and θ(n, m) represents the gradient direction;

[0024] S303, according to the gradient direction and amplitude at each pixel point, a histogram is established, the histogram features extracted in each block are connected in a loop to form a one-dimensional vector, and an overall image feature of the factor damaging the power transmission channel is formed.

[0025] Further, the specific steps of establishing the image contrast data set, the image contrast sub-database and the image contrast database in S4 include:

[0026] S401, different types of overall image features of the same factor damaging the power transmission channel in the same state are combined to establish an image contrast data set;

[0027] S402, multiple groups of image contrast data sets established in different states are combined to establish an image contrast sub-database;

[0028] S403, multiple groups of image contrast sub-databases established by different factors are combined to establish an image contrast database.

[0029] Further, the specific steps of establishing the feature recognition database in S5 include:

[0030] S501, according to the overall image features of the factors extracted in S3, different types of overall image features of the same factor damaging the power transmission channel are compared and recognized to obtain the same features of different types of the same factor;

[0031] S502, the same features of different types of each factor damaging the power transmission channel are obtained, all the same features are combined, and a feature recognition database is established.

[0032] Further, the specific steps of establishing the similarity contrast analysis sub-database and the similarity contrast analysis database in S6 include:

[0033] S601, the overall image features of the factors damaging the power transmission channel collected are subjected to similarity contrast analysis processing, different types of the same factor damaging the power transmission channel in the same state are subjected to similarity contrast analysis, the similarity of the overall image features between two different types is obtained, and the minimum similarity between different types of the same factor damaging the power transmission channel in the same state is obtained;

[0034] S602, combine the minimum similarities corresponding to different states of the same factor damaging the power transmission channel to establish a similarity comparison and analysis sub-database;

[0035] S603, combine the similarity comparison and analysis sub-databases corresponding to different factors damaging the power transmission channel to establish a similarity comparison and analysis database.

[0036] Further, the specific method for obtaining the minimum similarity in S601 includes:

[0037] S6011, extracting the overall image features of the relevant factors in the image frame of the factor damaging the power transmission channel;

[0038] S6012, performing scaling processing and grayscale processing on the extracted overall image features of the factor damaging the power transmission channel;

[0039] S6013, determining the pixel points of the overall image features based on the size of the overall image features after scaling processing, calculating the average value of each row of pixel points in the overall image features using the meanStdDev function based on OpenCV, and recording:

[0040] Wherein, mean is the average value of each row of pixel points, N is the number of each row of pixel points, and a is the grayscale value of the pixel point.

[0041] S6014, calculating the variance of the average value of each row of pixel points in all the overall image features to obtain the feature value of the overall image features, and the variance is:

[0042] Wherein, M is the number of rows of the overall image features, and mean0 is the average value of the pixel points of M rows of the overall image features.

[0043] S6015, determining the similarity between two groups of overall image features according to the variance difference between the two groups of overall image features, and the smaller the variance difference, the greater the similarity.

[0044] Further, the specific steps of identifying whether the image frame information identified in S8 exists a security risk of damaging the power transmission channel include:

[0045] S801, comparing and analyzing the identified image frame information with the same features of different factors in the feature recognition database, and when the same features appear in the image frame, extracting the corresponding features in the image frame and the image information associated therewith;

[0046] S802, according to the same feature in the feature recognition database obtained by contrast analysis, the image contrast data group corresponding to the same feature is called from the image contrast database;

[0047] S803, the image information extracted in S801 is compared with a plurality of random overall image features in the image contrast data group called in S802, and a plurality of similarities of the extracted image information and the overall image features in the image contrast data group are obtained;

[0048] S804, the plurality of similarities obtained in S803 are compared with the minimum similarity corresponding in the similarity contrast analysis sub-database in the similarity contrast analysis database, and whether the image picture information identified in S801 exists the security risk of damaging the power transmission channel is judged.

[0049] Compared with the prior art, the present application has the following beneficial effects:

[0050] The digital power transmission channel intelligent safety inspection method, by collecting different types and different states of image pictures of external environmental factors damaging the power transmission channel, and digitally processing the image pictures, forming image contrast data group, image contrast sub-database, image contrast database, feature recognition database, similarity contrast analysis sub-database and similarity contrast analysis database, which are used as model data for power transmission channel monitoring and inspection identification and judgment, using the above database and data group, the safety risk possibly existing in the environment around the power transmission channel can be intelligently identified and predicted, and the safety risk of the power transmission channel being damaged is reduced BRIEF DESCRIPTION OF DRAWINGS

[0051] Fig. 1 The flow chart of the digital power transmission channel intelligent safety inspection method of the present application;

[0052] Fig. 2 The flow chart of the present application for judging whether the identified image picture information exists the security risk of damaging the power transmission channel. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0054] The implementation of the digital power transmission channel intelligent safety inspection method is as follows:

[0055] Please refer toFigs. 1-2 The application discloses a digital power transmission channel intelligent safety inspection method, which comprises the following specific steps.

[0056] S1, collecting and counting factors damaging the power transmission channel and proportions of the factors damaging the power transmission channel and seasons in which the factors damaging the power transmission channel occur;

[0057] Specifically, the factors include large machines used in a protection area of the power transmission channel, construction sites in the protection area, operation equipment such as tower erection, wire stringing, erection or removal of crossing frames in the protection area, trees not meeting a safety distance, buildings, facilities and sites not meeting the safety distance, kites flying near the protection area, fishing areas near the protection area, fire-prone areas near the protection area, sites near the protection area in which floating objects are likely to be formed to endanger safe operation of the power transmission line, defects of foreign objects hung on the power transmission line, bird nests in the power transmission channel and icing phenomena on the power transmission channel in winter.

[0058] S2, collecting image pictures of the factors damaging the power transmission channel at monitoring points of the power transmission channel;

[0059] S201, collecting image pictures of all the factors damaging the power transmission channel counted in S1;

[0060] S202, collecting different types of image pictures of the same factor, different state pictures of the same type of factor or different state pictures of the same factor.

[0061] S3, processing the collected image pictures of the factors damaging the power transmission channel and extracting overall image features of the factors damaging the power transmission channel in the image pictures;

[0062] S301, adjusting the collected image pictures of the factors damaging the power transmission channel based on the size of the image pictures, and performing gray-scale processing, normalization and image segmentation preprocessing;

[0063] S302, performing convolution on each block after image segmentation, and obtaining a gradient direction and a gradient value of each pixel point in the image;

[0064] wherein (n, m) is a pixel point position, Gn(n, m)=I(n+1, m)-I(n-1, m), Gm(n, m)=I(n, m+1)-I(n, m-1), Gn and Gm represent gradient values in horizontal and vertical directions, M(n, m) represents a gradient value, and θ(n, m) represents a gradient direction;

[0065] S303, establishing a histogram according to the gradient direction and the gradient value of each pixel point, connecting the extracted histogram features in each block in a head-to-tail mode to form a one-dimensional vector, and forming overall image features of the factors damaging the power transmission channel.

[0066] S4, establish an image contrast data set, an image contrast sub-database and an image contrast database using the overall image features of the factors extracted in S3:

[0067] S401, combine different types of overall image features of the same factor in the same state that disrupts the power transmission channel to establish an image contrast data set;

[0068] S402, combine multiple groups of image contrast data sets established in different states to establish an image contrast sub-database;

[0069] S403, combine multiple groups of image contrast sub-databases established by different factors to establish an image contrast database.

[0070] S5, establish a feature recognition database using the overall image features of the factors extracted in S3:

[0071] S501, according to the overall image features of the factors extracted in S3, compare and identify different types of overall image features of the same factor that disrupts the power transmission channel to obtain the same features of different types of the same factor;

[0072] S502, obtain the same features of different types of each factor that disrupts the power transmission channel, combine all the same features to establish a feature recognition database.

[0073] S6, conduct similarity contrast analysis on the overall image features of the factors that disrupt the power transmission channel collected to establish a similarity contrast analysis sub-database and a similarity contrast analysis database:

[0074] S601, conduct similarity contrast analysis on the overall image features of the factors that disrupt the power transmission channel collected, conduct similarity contrast analysis on different types of the same factor in the same state that disrupts the power transmission channel, obtain the similarity of the overall image features between each two different types, and obtain the minimum similarity between different types of the same factor in the same state that disrupts the power transmission channel:

[0075] S6011, extract the overall image features of the relevant factors in the image frames of the factors that disrupt the power transmission channel;

[0076] S6012, scale and grayscale the extracted overall image features of the factors that disrupt the power transmission channel;

[0077] S6013, determine the pixel points of the overall image features based on the size of the overall image features after scaling, calculate the mean value of each row of pixel points in the overall image features based on OpenCV using the meanStdDev function, and record:

[0078] Wherein, mean is the average value of each row of pixel points, N is the number of each row of pixel points, and a is the gray value of the pixel point;

[0079] S6014, variance calculation is performed on the average value of each row of pixel points in all overall image features, and the obtained variance is the eigenvalue of the overall image feature, and the variance is:

[0080] Wherein, M is the number of rows of the overall image feature, and mean0 is the average value of the M rows of overall image feature pixel points;

[0081] S6015, according to the variance difference between the two groups of different overall image features, the similarity between the two groups of overall image features is determined, and the smaller the variance difference is, the greater the similarity is;

[0082] S602, the minimum similarity corresponding to the same factor in different states of all damaged power transmission channels is combined to establish a similarity comparison and analysis sub-database;

[0083] S603, the similarity comparison and analysis sub-databases corresponding to different factors of all damaged power transmission channels are combined to establish a similarity comparison and analysis database.

[0084] S7, the monitoring equipment is deployed in the power transmission channel, the image picture of the power transmission channel and its surrounding environment is acquired in real time, and the image picture information is recognized in real time;

[0085] S8, the recognized image picture information is compared and analyzed with the same features of different factors in the feature recognition database, and combined with the feature recognition database, the image comparison database and the similarity comparison and analysis database, whether the recognized image picture information exists security hidden danger of damaging the power transmission channel is judged:

[0086] S801, the recognized image picture information is compared and analyzed with the same features of different factors in the feature recognition database, when the same features as in the feature recognition database appear in the image picture, the corresponding features in the image picture and the image information associated with them are extracted;

[0087] S802, according to the same features in the feature recognition database obtained by comparison and analysis, the image comparison data group corresponding to the same features is called from the image comparison database;

[0088] S803, the image information extracted in S801 is compared with the random multiple overall image features in the image comparison data group called in S802, and the similarity between the multiple extracted image information and the overall image features in the image comparison data group is obtained;

[0089] S804, compare the plurality of similarities obtained in S803 with the corresponding minimum similarity in the similarity comparison sub-database in the similarity comparison analysis database, and determine whether the image picture information identified in S801 has a security risk of causing damage to the power transmission channel.

[0090] While embodiments of the present application have been shown and described, it is to be understood that the embodiments described are merely divergences of the principles and specific embodiments of the present application and that numerous modifications, substitutions, changes, and alterations can be made thereto without departing from the spirit and scope of the present application, which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent safety inspection of digital power transmission channels, characterized in that: The method comprises the following specific steps: S1, collecting and counting factors damaging the power transmission channel; S2, collecting image pictures of factors damaging the power transmission channel at monitoring points of the power transmission channel; S3, processing the collected image pictures of factors damaging the power transmission channel, and extracting overall image features of the factors damaging the power transmission channel in the image pictures; S4, establishing an image comparison data group, an image comparison sub-database and an image comparison database by using the overall image features of the factors extracted in S3; S5, establishing a feature recognition database by using the overall image features of the factors extracted in S3; S6, performing similarity comparison analysis on the overall image features of the collected factors damaging the power transmission channel, and establishing a similarity comparison analysis sub-database and a similarity comparison analysis database; S7, deploying monitoring equipment at the power transmission channel, acquiring image pictures of the power transmission channel and its surrounding environment in real time, and identifying image picture information in real time; S8, comparing and analyzing the identified image picture information with the same features of different factors in the feature recognition database, and combining the feature recognition database, the image comparison database and the similarity comparison analysis database to determine whether the identified image picture information has a security risk of damaging the power transmission channel; The specific steps of extracting the overall image features of the factors damaging the power transmission channel in S3 comprise: S301, adjusting the size of the collected image pictures of the factors damaging the power transmission channel based on the HOG algorithm, and performing grayscale, normalization and image segmentation preprocessing; S302, performing convolution on each block after image segmentation, and acquiring the gradient direction and amplitude value of each pixel point in the image: Wherein, (n, m) is the pixel point position, Gn(n, m) = I(n+1, m)-I(n-1, m), Gm(n, m) = I(n, m+1)-I(n, m-1), Gn and Gm represent the gradient values in the horizontal and vertical directions, M(n, m) represents the amplitude of the gradient, and θ(n, m) represents the gradient direction; S303, establishing a histogram according to the gradient direction and amplitude value of each pixel point, connecting the extracted histogram features in each block head to tail, forming a one-dimensional vector, and forming the overall image features of the factors damaging the power transmission channel; The specific steps of establishing the image comparison data group, the image comparison sub-database and the image comparison database in S4 comprise: S401, combining different types of overall image features of the same factor damaging the power transmission channel in the same state to establish an image comparison data group; S402, combining multiple groups of image comparison data groups established under different states to establish an image comparison sub-database; S403, combining multiple groups of image comparison sub-databases established by different factors to establish an image comparison database; The specific steps of establishing the feature recognition database in S5 comprise: S501, comparing and identifying different types of overall image features of the same factor damaging the power transmission channel according to the overall image features of the factors extracted in S3, and acquiring the same features of different types of the same factor; S502, acquire different types of same features of each factor of the damaged power transmission channel, combine all the same features to establish a feature recognition database; The specific steps of establishing the similarity comparison and analysis sub-database and the similarity comparison and analysis database in S6 include: S601, perform similarity comparison and analysis processing on the overall image features of the collected factors of the damaged power transmission channel, perform similarity comparison and analysis on different types of the same factor of the damaged power transmission channel in the same state, acquire the similarity of the overall image features between each two different types, and obtain the minimum similarity between different types of the same factor of the damaged power transmission channel in the same state; S602, combine the corresponding minimum similarities of the same factor of the damaged power transmission channel in different states to establish a similarity comparison and analysis sub-database; S603, combine the similarity comparison and analysis sub-databases corresponding to different factors of the damaged power transmission channel to establish a similarity comparison and analysis database; The specific steps of identifying whether the image frame information recognized in S8 exists a security risk of damaging the power transmission channel include: S801, compare the recognized image frame information with the same features of different factors in the feature recognition database, and when the same features as in the feature recognition database appear in the image frame, extract the corresponding features in the image frame and the image information associated therewith; S802, according to the same features in the feature recognition database obtained through comparison and analysis, retrieve the image comparison data group corresponding to the same features from the image comparison database; S803, perform similarity comparison between the image information extracted in S801 and the random multiple overall image features in the image comparison data group retrieved in S802, to obtain the similarity between the multiple extracted image information and the overall image features in the image comparison data group; S804, compare the multiple similarities obtained in S803 with the corresponding minimum similarities in the similarity comparison and analysis sub-database in the similarity comparison and analysis database, to determine whether the image frame information recognized in S801 exists a security risk of damaging the power transmission channel.

2. The intelligent security inspection method for digitized power transmission channel according to claim 1, characterized in that: The statistics in S1 also include the proportion of factors of the damaged power transmission channel and the seasons in which the damage occurs.

3. The method of claim 1, wherein the method further comprises: The image frames of the factors of the damaged power transmission channel collected in S2 specifically include: S201, collect the image frames of all the factors of the damaged power transmission channel counted in S1; S202, collect different type image frames of the same factor, collect different state frames of the same type factor, or collect different state frames of the same factor.

4. The intelligent security patrol method for digitized power transmission channel according to claim 1, characterized in that: The specific method of acquiring the minimum similarity in S601 includes: S6011, extract the overall image features of the relevant factors in the image frames of the factors of the damaged power transmission channel; S6012, perform scaling processing and grayscale processing on the extracted overall image features of the factors of the damaged power transmission channel; S6013. By scaling the overall image feature, determine the number of pixels in the overall image feature. Based on OpenCV, use the meanStdDev function to calculate the average value of each row of pixels in the overall image feature, and record it. Where mean is the average value of each row of pixels, N is the number of pixels in each row, and a is the gray value of the pixel. S6014. Calculate the variance of the average value of each row of pixels in all the overall image features. The resulting variance is the feature value of the overall image feature. The variance is: Where M is the number of rows of the overall image features, and mean0 is the average value of the pixels in the M rows of the overall image features; S6015. Determine the similarity between the two sets of overall image features based on the variance difference between the two sets of different overall image features. The smaller the variance difference, the greater the similarity.

Images