CN116452594B - Visualized monitoring and early warning method and system for power transmission line state - Google Patents
Visualized monitoring and early warning method and system for power transmission line state Download PDFInfo
- Publication number
- CN116452594B CN116452594B CN202310720414.0A CN202310720414A CN116452594B CN 116452594 B CN116452594 B CN 116452594B CN 202310720414 A CN202310720414 A CN 202310720414A CN 116452594 B CN116452594 B CN 116452594B
- Authority
- CN
- China
- Prior art keywords
- pixel point
- blank
- image
- target
- acquiring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000005540 biological transmission Effects 0.000 title claims abstract description 93
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012544 monitoring process Methods 0.000 title claims abstract description 51
- 238000005070 sampling Methods 0.000 claims abstract description 60
- 238000001914 filtration Methods 0.000 claims abstract description 47
- 230000008859 change Effects 0.000 claims abstract description 19
- 230000000007 visual effect Effects 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 7
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 10
- 230000011218 segmentation Effects 0.000 claims description 9
- 230000001154 acute effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 230000000875 corresponding effect Effects 0.000 description 13
- 238000007689 inspection Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 11
- 238000010606 normalization Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000009499 grossing Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007306 functionalization reaction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration by the use of histogram techniques
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention relates to the technical field of image data processing, and provides a transmission line state visual monitoring and early warning method and system, comprising the following steps: collecting an infrared image of a power transmission line; acquiring the optimal Gaussian kernel size according to the gray variance and the local variance of each super pixel region; acquiring a reference range and a plurality of first weight parameters of each blank, and acquiring a plurality of second weight parameters according to gray value and position distribution of each pixel point in the reference range; acquiring a guide image of the infrared image, and acquiring a filling value of each blank lattice according to the change of gray variance in the guide image and the infrared image and combining weight parameters to obtain an up-sampling image; obtaining a guide filtering intensity parameter according to the difference of the local variance statistical histogram of the up-sampling image and the infrared image; and guiding and filtering to obtain a clear image, and completing visual monitoring of the state of the power transmission line. The invention aims to solve the problem that the number of pixels of the power transmission line is small, so that the guiding filtering enhancement effect is inaccurate.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to a transmission line state visual monitoring and early warning method and system.
Background
The power transmission line is provided with important tasks in the process of electric energy transmission, the running state of the power transmission line is closely related to life of people, and regular inspection of the power transmission line is an important measure for ensuring safe and stable running of a power grid; at present, unmanned aerial vehicle inspection operation is used as a main mode of power transmission line inspection operation, an inspection process is carried out on an inspected power transmission line body and power equipment by a thermal infrared imager carried by the unmanned aerial vehicle, monitoring image data of the power transmission line can be rapidly obtained in real time, and potential fault hidden danger is found according to the temperature distribution condition of the surface of the line.
Due to the fact that the topography is complex and the power transmission line planning problem is solved, the flying height cannot be close to the power transmission line due to the fact that the problem of obstacles is likely to occur in the unmanned aerial vehicle inspection process, more interference data can appear under the condition of complex inspection environment, meanwhile, the inspection operation is also limited by climatic factors, inspection images of the power transmission line are poor in resolution ratio and definition in many times, and therefore a strong image processing function is needed to conduct preprocessing and enhancement on monitoring images so as to guarantee accuracy of monitoring data of the power transmission line. The existing literature, "an infrared image detail enhancement algorithm based on parameter self-adaptive guided filtering", proposes to acquire the guided filtering strength by using a local variance histogram of an original infrared image, but for a power transmission line, the number of pixel points belonging to the power transmission line part is small, so that the local variance contribution rate is low, the guided filtering strength self-adaptive according to the local variance may not be suitable, and the guided filtering cannot well realize denoising and enhancement on the power transmission line part.
Disclosure of Invention
The invention provides a visualized monitoring and early warning method and a visualized monitoring and early warning system for a power transmission line state, which aim to solve the problem that the existing power transmission line has less pixel number and is easy to cause inaccurate guiding filtering enhancement effect, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a method for visually monitoring and early warning a power transmission line status, including the steps of:
collecting an infrared image of a power transmission line;
performing super-pixel segmentation on the infrared image, and acquiring the optimal Gaussian kernel size according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes;
obtaining a plurality of blank grids by up-sampling interpolation of an infrared image, obtaining a reference range of each blank grid according to the optimal Gaussian kernel size, and generating a first weight parameter of each pixel point in the reference range according to the Gaussian kernel;
acquiring the extension direction of the power transmission line, and acquiring a second weight parameter of each pixel point in each blank reference range according to the gray value and the position distribution of each pixel point in each blank reference range;
acquiring a guide image of an infrared image, acquiring a filling value of each blank cell according to gray variances of each pixel point in a first sliding window in the infrared image and the guide image in a reference range of each blank cell, and combining a first weight parameter and a second weight parameter to finish up-sampling interpolation to obtain an up-sampling image;
Acquiring a guide filtering intensity parameter according to the difference of two local variance statistical histograms of the infrared image and the up-sampling image;
and performing guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, and performing anomaly monitoring on the clear image to complete visual monitoring of the state of the power transmission line.
Optionally, the method for obtaining the optimal gaussian kernel size includes the following specific steps:
acquiring the gray variance of each super pixel region, and acquiring the local variance of each pixel point under different Gaussian kernel sizes;
taking any one Gaussian kernel size as a target Gaussian kernel size, the expression of a first objective function of the target Gaussian kernel size is as follows:
wherein ,the number of pixels representing local variance at target kernel size, +.>Indicate->Local variance of each pixel point under the target Gaussian kernel size, < >>Indicate->Gray variance of super pixel area to which each pixel belongs, < ->Representing absolute value;
and iterating the Gaussian kernel sizes from the initial value according to the increase size, calculating the output value of the first objective function under each Gaussian kernel size, and taking the Gaussian kernel size with the minimum output value as the optimal Gaussian kernel size.
Optionally, the obtaining the gray variance of each super pixel region and obtaining the local variance of each pixel point under different gaussian kernel sizes includes the following specific methods:
calculating the variance of the gray value of the pixel point in the corresponding region of each super pixel, and marking the variance as the gray variance of each super pixel region;
taking any pixel point in the infrared image as a target pixel point and taking the target pixel point as the center to obtainOther pixels in the local area, wherein +.>The Gaussian kernel size is represented, gray value variances are calculated for the target pixel points and the pixel points in the local range, the gray value variances are recorded as local variances of the target pixel points under the corresponding Gaussian kernel size, and the local variances of the target pixel points under different Gaussian kernel sizes are obtained; and obtaining the local variance of each pixel point under different Gaussian kernel sizes.
Optionally, the obtaining the second weight parameter of each pixel point in each blank reference range includes the following specific methods:
the gray change characteristics and the deviation included angles of each pixel point in each blank reference range are obtained, any blank is taken as a target blank, and the first blank is in the target blank reference rangeSecond weight parameter of each pixel point +. >The calculation method of (1) is as follows:
wherein ,indicating the%>Deviation included angle of each pixel point to target blank lattice, < >>Indicating the%>Gray scale variation characteristic of each pixel point to target blank lattice,>minimum value representing gray scale variation characteristic in target blank reference range, < >>To avoid hyper-parameters with denominator 0, < ->Representing absolute value;
and acquiring a second weight parameter of each pixel point in the reference range of each blank lattice.
Optionally, the method for obtaining the gray scale change characteristic and the offset included angle of each pixel point in the reference range of each blank comprises the following specific steps:
taking any one blank as a target blank, taking the average value of the pixel points in the reference range as a planned filling value of the target blank, acquiring any one pixel point in the reference range as a target pixel point, calculating the absolute value of the difference between the gray value of the target pixel point and the planned filling value, and recording the ratio of the absolute value of the difference to the Euclidean distance between the target pixel point and the target blank as the gray change characteristic of the target pixel point to the target blank;
acquiring a connection line of the target pixel point and the target blank, and marking an acute angle in an included angle formed by the connection line and the extending direction of the power transmission line as a deviated included angle of the target pixel point to the target blank;
And acquiring gray scale change characteristics and deviation included angles of each pixel point in each blank lattice reference range.
Optionally, the method for obtaining the filling value of each blank cell includes the following specific steps:
obtaining local variance contribution rates of each pixel point in the guide image and the infrared image respectively, and taking any one blank as a target blank within a reference range of the target blankThe method for calculating the comprehensive weight parameters of each pixel point comprises the following steps:
wherein ,represents the +.o within the target blank reference range>Reference degree of individual pixels, +.>Represents the +.o within the target blank reference range>First weight parameter of each pixel, < ->Representing target blank reference rangesInterior (I)>Second weight parameter of each pixel, < ->Represents the +.o within the target blank reference range>Local variance contribution rate of each pixel point in guide image,/for the guide image>Represents the +.o within the target blank reference range>Local variance contribution rate of each pixel point in infrared image, < >>To avoid hyper-parameters with denominator 0, < ->Representing absolute value;
obtaining the reference degree of each pixel point in the reference range of the target blank lattice, normalizing all the reference degrees, marking the obtained result as the comprehensive weight parameter of each pixel point, and marking the result obtained by weighting and summing as the filling value of the target blank lattice according to the comprehensive weight parameter and the gray value of each pixel point; and acquiring the filling value of each blank cell.
Optionally, the method for obtaining the local variance contribution rate of each pixel point in the guide image and the infrared image includes the following specific steps:
constructing by taking any one pixel point as a target pixel point and taking the target pixel point as the centerThe first sliding windows with the sizes are used for respectively calculating gray variance of the target pixel points in the first sliding windows in the guide image and the infrared image; acquiring infrared images and guiding imagesGray variance in the first sliding window of each pixel point;
normalizing the gray variance in the first sliding window of all the pixel points in the guide image, and marking the obtained result as the local variance contribution rate of each pixel point in the guide image;
and normalizing the gray variance in the first sliding window of all the pixel points in the infrared image, and marking the obtained result as the local variance contribution rate of each pixel point in the infrared image.
Optionally, the specific method for obtaining the guide filtering strength parameter includes:
acquiring gray variance in all first sliding windows in the infrared image, and generating a local variance statistical histogram of the infrared image; acquiring gray variance in all second sliding windows in the up-sampling image, generating a local variance statistical histogram of the up-sampling image, and guiding filtering strength parameters The calculation method of (1) is as follows:
wherein ,maximum value representing the number of items with vertical axis values in both histograms, < >>A third part of the statistical histogram of local variances representing an infrared image>Longitudinal axis value of each item with longitudinal axis value, < ->A local variance statistical histogram representing an up-sampled image +.>Longitudinal axis value of each item with longitudinal axis value, < ->An exponential function based on a natural constant is represented.
Optionally, the specific method for acquiring gray variance in all second sliding windows in the up-sampled image includes:
taking any one first sliding window in an infrared image as a target first sliding window, acquiring each row and each column of the target first sliding window in the infrared image, extracting each row and each column corresponding to each row in an up-sampling image, extracting each row or each column which is already acquired from adjacent downward rows or adjacent rightward columns in the up-sampling image, and jointly forming a second sliding window corresponding to the target first sliding window in the up-sampling image by a plurality of rows and a plurality of columns extracted from the up-sampling image, wherein the second sliding window is recorded as a target second sliding window;
acquiring gray values of all pixel points in a second sliding window of the target, obtaining gray variance of all pixel points in the second sliding window of the target, and marking the gray variance as the gray variance in the second sliding window of the target; and acquiring gray level variances in all the second sliding windows.
In a second aspect, another embodiment of the present invention provides a transmission line status visual monitoring and early warning system, including:
the infrared image acquisition module is used for acquiring infrared images of the transmission line;
an infrared image processing module: performing super-pixel segmentation on the infrared image, and acquiring the optimal Gaussian kernel size according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes;
obtaining a plurality of blank grids by up-sampling interpolation of an infrared image, obtaining a reference range of each blank grid according to the optimal Gaussian kernel size, and generating a first weight parameter of each pixel point in the reference range according to the Gaussian kernel;
acquiring the extension direction of the power transmission line, and acquiring a second weight parameter of each pixel point in each blank reference range according to the gray value and the position distribution of each pixel point in each blank reference range;
acquiring a guide image of an infrared image, acquiring a filling value of each blank cell according to gray variances of each pixel point in a first sliding window in the infrared image and the guide image in a reference range of each blank cell, and combining a first weight parameter and a second weight parameter to finish up-sampling interpolation to obtain an up-sampling image;
Acquiring a guide filtering intensity parameter according to the difference of two local variance statistical histograms of the infrared image and the up-sampling image;
and the power transmission line monitoring module is used for conducting guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, and performing abnormal monitoring on the clear image to complete visual monitoring of the power transmission line state.
The beneficial effects of the invention are as follows: the invention improves the pixel expression of the power transmission line region by an up-sampling method; firstly, according to the iterative Gaussian kernel size, analyzing the difference between the local variance and the gray variance of the super pixel region, obtaining the optimal Gaussian kernel size and obtaining the reference range and the first weight parameter of the blank; then, the effective information of the power transmission line in the up-sampling process is kept complete by weighting interpolation of the neighborhood pixel points in the extension direction of the power transmission line and avoiding the problem of block effect, so as to obtain a second weight parameter; finally, according to the local variance contribution rate change of the guide image and the infrared, compensating the weight of the area with higher noise intensity, so that the interference information of the high noise part is reserved and not smoothed by up-sampling, and only weaker noise and background area are smoothed to obtain an up-sampling image; calculating the difference of the local variance statistical histogram between the infrared image and the up-sampling image, wherein the up-sampling image is equivalent to the gray variance of the background area, and the larger the difference of the histogram is, the larger the integral noise interference is, and the smaller the difference is, the accurate guide filtering intensity parameter can be obtained through front-back comparison; because the effective information of the power transmission line area is reserved as much as possible in the up-sampling process, the filtering strength is obtained by the difference before and after up-sampling on the premise of ensuring the completeness of the effective information, and the high-strength noise in the image can be effectively restrained and the completeness of the effective information is reserved after the filtering is smooth, so that the noise interference is avoided in the monitoring of the power transmission line state, the accuracy of the abnormal monitoring of the power transmission line is improved, and the efficiency of the visual monitoring of the power transmission line state is further improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
Fig. 1 is a schematic flow chart of a visual monitoring and early warning method for a power transmission line state according to an embodiment of the present invention;
fig. 2 is a block diagram of a transmission line status visualization monitoring and early warning system according to another embodiment of the present invention;
fig. 3 is a schematic diagram of an infrared image of a transmission line.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flow chart of a method for visualized monitoring and early warning of a power transmission line state according to an embodiment of the invention is shown, and the method comprises the following steps:
and S001, collecting an infrared image of the transmission line through unmanned aerial vehicle inspection operation.
The purpose of the embodiment is to acquire an infrared image of the transmission line through unmanned aerial vehicle inspection operation, and perform visual monitoring of the transmission line according to the infrared image of the transmission line, so that the infrared image of the transmission line needs to be acquired first; the transmission line is acquired by infrared image through unmanned aerial vehicle inspection operation, which is not described in detail in this embodiment, but the infrared image may be represented as a gray image, the gray value of each pixel point reflects the temperature of the corresponding position, and the embodiment uses the infrared gray image to describe, and the pixel value of each pixel point is described as a gray value; referring to fig. 3, an infrared image of a transmission line is shown.
Thus, an infrared image of the transmission line is obtained.
It should be noted that, the outdoor operation interference source is complex, more noise interference exists in the infrared image, the number of pixel points of the power transmission line part is small, and when the guided filtering is utilized for smoothing, the local variance information of the pixel points at the power transmission line part is insufficient to express the filtering strength at the position, which may cause the smoothing result to be over-smooth and under-smooth; meanwhile, the pixel points occupy less pixels and are easy to lose detail information when noise is filtered, and the detail information loss is further aggravated when the filtering intensity setting is deviated, so that the adaptive guiding filtering intensity parameter setting is required for the power transmission line part.
And step S002, super-pixel segmentation is carried out on the infrared image, and the optimal Gaussian kernel size is obtained according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes.
It should be noted that, because the pixel expression of the power transmission line part is less, the image pyramid is required to be used for up-sampling interpolation, but the interpolation process is not only interfered by noise, but also noise information on the infrared image is inhibited to a certain extent, and even if the pixel expression of the power transmission line part is improved, accurate guide filtering intensity parameters may not be obtained; then interpolation is needed under the condition of noise existence to improve the pixel expression of the power transmission line part, the interpolation process keeps the effective information of the target power transmission line area, meanwhile, the influence of local noise is kept, and then the state change of the local variance histogram before and after the interpolation is analyzed to obtain more accurate guide filtering intensity parameters.
It should be further noted that, the up-sampling interpolation is to insert a blank row and a blank column of pixel grids between adjacent rows and adjacent columns of the original image, and perform weighted interpolation by using pixel points in a blank neighborhood to obtain a filling value of the blank grid; the pixel points closer to the blank lattice should have larger weight, so the Gaussian kernel is used as a rule for giving interpolation weight; meanwhile, in order to ensure that the interpolated image is not distorted, super-pixel segmentation is performed on the infrared image, and local variances of all neighborhood pixel points in a reference range of the blank lattice need to be ensured to approach gray variances of the super-pixel region, so that the optimal Gaussian kernel size is obtained to construct the reference range of the blank lattice.
Specifically, the infrared image is first subjected to super-pixel segmentation to obtain a plurality of super-pixels, wherein the super-pixel segmentation is known in the art, and the embodiment is not repeated; calculating the variance of the gray value of the pixel point in the corresponding region of each super pixel, and marking the variance as the gray variance of each super pixel region; for any pixel point in the infrared image, taking the pixel point as the center to obtainOther pixels in the local area, wherein +.>Indicates Gaussian kernel size, < >>The initial value of (2) is 3, and then iterative increment is carried out by taking 2 as the increment size each time; to->For example, calculating gray value variance of the pixel point and the pixel points in the local range, marking the gray value variance as the local variance of the pixel point under the Gaussian kernel size, and obtaining the local variance of the pixel point under different Gaussian kernel sizes; obtaining local variances of each pixel point under different Gaussian kernel sizes according to the method; it should be noted that, if the pixel point of the infrared image is insufficient to form a local range of a certain pixel point under a certain gaussian kernel size, that is, the local range of the pixel point exceeds the infrared image boundary, the local range of the pixel point is not obtained, and the sub-local variance is not calculated any more 。
Further, according to the gray variance of the super pixel area and the local variance of each pixel point under different gaussian kernel sizes, constructing a first objective function for the gaussian kernel sizes, wherein the expression of the first objective function of any one gaussian kernel size is as follows:
wherein ,the number of pixels representing the local variance at the gaussian kernel size, +.>Indicate->Local variance of each pixel point under the Gaussian kernel size, < >>Indicate->Gray variance of super pixel area to which each pixel belongs, < ->Representing absolute value; the smaller the difference between the local variance of each pixel point and the gray variance of the super-pixel area, the closer to 0, which indicates that the information in the local range of the pixel point under the size of the Gaussian kernel is more approximate to the information of the super-pixel area, and the size of the Gaussian kernel is more proper; and iterating the Gaussian kernel sizes from the initial value, simultaneously calculating the output value of the first objective function under each Gaussian kernel size, and taking the Gaussian kernel size with the minimum output value as the optimal Gaussian kernel size.
So far, the optimal Gaussian kernel size is obtained and used for subsequently constructing the reference range of the blank lattice in the up-sampling interpolation.
Step S003, a plurality of blank grids are obtained through up-sampling interpolation on the infrared image, a reference range of each blank grid and a first weight parameter of each pixel point in the reference range are obtained according to the optimal Gaussian kernel size, and a second weight parameter of each pixel point in the reference range of each blank grid is obtained according to the gray value and the position distribution of each pixel point in the reference range.
It should be noted that, the extension direction of the power transmission line is known, so that in order to ensure the integrity of the information on the power transmission line in the interpolation process, the interpolation weight needs to be focused on the extension direction of the power transmission line; when the adjacent pixel points are selected for weighting, the direction of the connecting line of the adjacent pixel points is closer to the extending direction of the power transmission line, the weight of the adjacent pixel points is larger, and the weight is smaller when the adjacent pixel points deviate; meanwhile, if the whole infrared image is given a larger weight according to the extending direction of the transmission line, obvious blocking effect can appear on the non-transmission line area, so that the interpolation weight is adjusted according to the characteristic of minimum gray level change in the interpolation process.
Specifically, firstly, a plurality of blank grids are obtained through up-sampling interpolation, for any blank grid, the blank grid is taken as the center, the local range of the size of the optimal Gaussian kernel is taken as the reference range of the blank grid, gaussian kernel weights in a group of reference ranges are generated according to Gaussian kernels, and the Gaussian kernel weights are recorded as first weight parameters of each pixel point; it should be noted that, the reference range includes other blank cells, the first weight parameter is only for the pixel point, and is invalid for other blank cells, i.e. other blank cells do not participate in the calculation of the subsequent weight parameter; acquiring a reference range of each blank cell and a first weight parameter of each pixel point in the reference range according to the method; the blank lattice near the boundary of the near infrared image has incomplete reference range, and the pixel points in the reference range are not required to be filled at this time, and the calculation is performed on the pixel points in the reference range.
Further, taking any blank as an example, taking the average value of the pixel points in the reference range as the planned filling value of the blank, obtaining the absolute value of the difference value between the gray value of any pixel point in the reference range and the planned filling value,the ratio of the absolute value of the difference value to the Euclidean distance between the pixel point and the blank grid is recorded as the gray level change characteristic of the pixel point to the blank grid; acquiring a connecting line of the pixel point and the blank grid, and recording an acute angle in an included angle formed by the connecting line and the extending direction of the power transmission line as a deviated included angle of the pixel point to the blank grid; according to the method, the gray change characteristics and the deviation included angles of each pixel point in the blank reference range are obtained, and then the first pixel point in the blank reference rangeSecond weight parameter of each pixel point +.>The calculation method of (1) is as follows:
wherein ,indicating the%>Deviation included angle of each pixel point to the blank lattice, < >>Indicating the%>Gray scale variation characteristic of each pixel point to the blank lattice,>minimum value of gray scale variation characteristic in reference range of the blank lattice, < >>To avoid hyper-parameters with denominator 0, this embodiment uses +.>To make a description of->Representing absolute value; the smaller the deviation included angle is, the larger the interpolation weight is, and the larger the second weight parameter is; the closer the gray level change feature is to the minimum gray level change feature, the smaller the possibility of blocking effect is caused, the larger the interpolation weight is, the larger the second weight parameter is, and the offset included angle is related to the influence of the gray level change feature through the Euclidean norm to obtain the second weight parameter; and acquiring a second weight parameter of each pixel point in the reference range of each blank according to the method.
Thus, a reference range of each blank cell in the up-sampling interpolation process and a first weight parameter and a second weight parameter of each pixel point in the reference range are obtained.
Step S004, obtaining a guiding image of the infrared image, obtaining gray variance of each pixel point in a first sliding window in the infrared image and the guiding image, and obtaining a filling value of each blank cell according to the gray variance, the first weight parameter and the second weight parameter, and finishing up-sampling interpolation to obtain an up-sampling image.
It should be noted that, the unmanned aerial vehicle route of patrolling and examining is fixed, and in frequent operation of patrolling and examining, there is a large amount of historical image data, only need the coordinate position that input unmanned aerial vehicle is located, can retrieve corresponding historical image from the historical database, and one image that the signal-to-noise ratio is minimum is selected as the guide image of guide filtering from the historical database, assists unmanned aerial vehicle infrared monitoring system to accomplish the processing of real-time monitoring image.
It should be further noted that, by acquiring the guiding image of the infrared image, the local noise contribution rate is represented according to the difference of the gray variance of the pixel points at the same position on the infrared image and the guiding image in the first sliding window, and the larger the noise contribution rate is, the larger the deviation of the pixel points caused to the filling value is, the compensation should be performed, and the influence of the pixel points to the filling value is reduced.
Specifically, according to the coordinate position of the corresponding unmanned aerial vehicle, the infrared image is subjected to historical shadow calling of a large number of identical coordinate positions from the historical databaseFor example, the historical image with the smallest signal-to-noise ratio is used as the guiding image of the infrared image, the signal-to-noise ratio is calculated as a known technology, and the embodiment is not repeated; constructing by taking any pixel point as the centerFirst sliding window of size, this embodiment uses +.>The gray values of other pixels in the first sliding window of the pixel are respectively obtained from the infrared image and the guide image, and the gray variance of the pixel in the first sliding window of the two images is obtained by combining the gray values of the pixel in the two images; acquiring gray variance in a first sliding window of each pixel point in the infrared image and the guide image according to the method; if a pixel point cannot obtain a complete first sliding window in the image, that is, the first sliding window corresponding to the pixel point close to the image boundary exceeds the image boundary, the pixel point filling is not needed for the first sliding window of the pixel points, and only the gray variance of the pixel points in the first sliding window is calculated.
Further, carrying out linear normalization on gray variance in a first sliding window of all pixel points in the guide image, and marking an obtained result as a local variance contribution rate of each pixel point in the guide image; carrying out linear normalization on gray variance in a first sliding window of all pixel points in the infrared image, and marking an obtained result as a local variance contribution rate of each pixel point in the infrared image; take any one blank as an example, the blank is within the reference range The method for calculating the comprehensive weight parameters of each pixel point comprises the following steps:
wherein ,indicating +.>Reference degree of individual pixels, +.>Indicating +.>First weight parameter of each pixel, < ->Indicating +.>Second weight parameter of each pixel, < ->Indicating +.>Local variance contribution rate of each pixel point in guide image,/for the guide image>Indicating +.>Local variance contribution rate of each pixel point in infrared image, < >>To avoid hyper-parameters with denominator 0, this embodiment uses +.>To make a description of->Representing absolute value; the reference degree is positively correlated with the first weight parameter and the second weight parameter, and the larger the difference of the local variance contribution rates is,the larger the noise contribution rate is, the reference degree needs to be reduced, and the smaller the reference degree is; obtaining the reference degree of each pixel point in the blank reference range according to the method, carrying out softmax normalization on all the reference degrees, marking the obtained result as a comprehensive weight parameter of each pixel point, and marking the result obtained by weighting and summing as a filling value of the blank according to the comprehensive weight parameter and the gray value of each pixel point; and acquiring the filling value of each blank according to the method, and finishing the up-sampling interpolation of the infrared head portrait, wherein the obtained image is recorded as an up-sampling image.
So far, the filling value of each blank is obtained, up-sampling interpolation is completed, and up-sampling images are obtained.
And S005, constructing second sliding windows on the up-sampled image, acquiring gray variance in each second sliding window, respectively generating a local variance statistical histogram by combining the gray variance in the first sliding window, and acquiring a guiding filtering intensity parameter according to the histogram difference.
It should be noted that, because of the up-sampling interpolation, the up-sampling image is expanded four times in size compared with the infrared image, that is, the number of rows and columns is respectively expanded to be twice as large as the original number; and obtaining second sliding windows by expanding the first sliding windows by four times, calculating gray variance in each second sliding window in the up-sampled image, respectively generating a local variance statistical histogram by the gray variance in the first sliding window and the gray variance in the second sliding window, and obtaining a guiding filtering intensity parameter according to the histogram difference.
Specifically, in step S003, a plurality of first sliding windows in the infrared image are obtained, where the first sliding windows have a size ofThe second sliding window size is set to +.>The second sliding window is aimed at the up-sampled image, then for any one first sliding window, the corresponding row and column of the first sliding window in the up-sampled image is obtained, because +. >Three rows and three columns are obtained, and one row or one column next to each other is obtained, and three rows and three columns are obtained, wherein six rows and six columns form a second sliding window corresponding to the up-sampled image of the first sliding window, the second sliding window is added with adjacent interpolation parts compared with the corresponding first sliding window, and the size of the second sliding window in the embodiment is->The method comprises the steps of carrying out a first treatment on the surface of the According to the method, the corresponding second sliding window of each first sliding window in the up-sampling image is obtained, and it is noted that the first sliding windows close to the boundary part may not obtain complete second sliding windows, the second sliding windows do not need to be filled, and only existing pixel points can be calculated in the subsequent gray variance calculation.
Further, for any one second sliding window in the up-sampling image, gray values of all pixel points in the second sliding window are obtained, gray variance of all pixel points in the second sliding window is obtained, and the gray variance is recorded as the gray variance in the second sliding window; acquiring gray variances in all second sliding windows according to the method, generating a local variance statistical histogram of the up-sampled image according to the gray variances, and acquiring the number of items with vertical axis values in the local variance statistical histogram, wherein the vertical axis values are the number of the second sliding windows; recording gray variance of all pixel points in each first sliding window in the infrared image as gray variance in the first sliding window, generating a local variance statistical histogram of the infrared image, equally acquiring the number of items with vertical axis values in the local variance statistical histogram, and guiding the filtering intensity parameter if the vertical axis value is the number of the first sliding windows The calculation method of (1) is as follows:
wherein ,maximum value representing the number of items with vertical axis values in both histograms, < >>A third part of the statistical histogram of local variances representing an infrared image>The vertical axis value of a term having a vertical axis value, i.e., the number of first sliding windows; />A local variance statistical histogram representing an up-sampled image +.>The vertical axis value of a term having a vertical axis value, i.e., the number of second sliding windows; it should be noted that, since there may be a difference in the number of items having vertical axis values in the two histograms, for a histogram having a smaller number, the last several items may have no vertical axis value, i.e., the vertical axis value is 0, and the vertical axis value is 0; />An exponential function that is based on a natural constant; the mean square error of the histogram is obtained according to the change of the item with the vertical axis value by comparing the difference of the histogram, and as the up-sampling contains a certain information smoothing function, the smaller the up-sampling loss is, the weaker the noise intensity is, and the larger the guiding filtering intensity parameter is needed; the larger the upsampling loss, the larger the noise intensity, the smaller the pilot filter intensity parameter needs to be; the present embodiment employs +.>To present inverse proportion relation and normalization process, and the implementer can select inverse proportion function and normalization function according to actual situation.
Thus, the pilot filter strength parameter is obtained.
And step S006, conducting guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, and conducting abnormal monitoring on the clear image to complete visual monitoring of the state of the power transmission line.
The infrared image is guided and filtered according to the guiding and filtering intensity parameters, noise reduction treatment is achieved, the obtained image is recorded as a clear image, the guiding and filtering is the prior art, and the embodiment is not repeated; the adaptive guide filtering intensity parameter is used for eliminating noise, and meanwhile, the detail information of the power transmission line part is reserved, so that the accuracy of the subsequent abnormal monitoring result is ensured.
Furthermore, abnormal temperature of the clear image is monitored, an LOF abnormal point detection method can be adopted to obtain the power transmission line in the temperature abnormal region, and then the temperature abnormal alarm device is triggered, so that abnormal monitoring visualization of the power transmission line is realized, and visual monitoring of the state of the power transmission line is completed.
So far, clear images are obtained through guided filtering and anomaly monitoring is carried out, and visual monitoring of the state of the power transmission line is completed.
Referring to fig. 2, a structural block diagram of a transmission line state visual monitoring and early warning system according to another embodiment of the present invention is shown, where the system includes:
And the infrared image acquisition module S101 acquires infrared images of the transmission line through unmanned aerial vehicle inspection operation.
The infrared image processing module S102:
(1) Performing super-pixel segmentation on the infrared image, and acquiring the optimal Gaussian kernel size according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes;
(2) Obtaining a plurality of blank grids by up-sampling interpolation of an infrared image, obtaining a reference range of each blank grid and a first weight parameter of each pixel point in the reference range according to the optimal Gaussian kernel size, and obtaining a second weight parameter of each pixel point in the reference range of each blank grid according to the gray value and the position distribution of each pixel point in the reference range;
(3) Acquiring a guide image of the infrared image, acquiring gray variance of each pixel point in a first sliding window in the infrared image and the guide image, acquiring a filling value of each blank according to the gray variance, the first weight parameter and the second weight parameter, and finishing up-sampling interpolation to obtain an up-sampling image;
(4) And constructing a second sliding window on the up-sampling image, acquiring gray variance in each second sliding window, respectively generating a local variance statistical histogram by combining the gray variance in the first sliding window, and acquiring a guiding filtering intensity parameter according to the histogram difference.
And the power transmission line monitoring module S103 performs guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, performs abnormal monitoring on the clear image and completes visual monitoring of the power transmission line state.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (9)
1. The visualized monitoring and early warning method for the state of the power transmission line is characterized by comprising the following steps of:
collecting an infrared image of a power transmission line;
performing super-pixel segmentation on the infrared image, and acquiring the optimal Gaussian kernel size according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes;
obtaining a plurality of blank grids by up-sampling interpolation of an infrared image, obtaining a reference range of each blank grid according to the optimal Gaussian kernel size, and generating a first weight parameter of each pixel point in the reference range according to the Gaussian kernel;
acquiring the extension direction of the power transmission line, and acquiring a second weight parameter of each pixel point in each blank reference range according to the gray value and the position distribution of each pixel point in each blank reference range;
Acquiring a guide image of an infrared image, acquiring a filling value of each blank cell according to gray variances of each pixel point in a first sliding window in the infrared image and the guide image in a reference range of each blank cell, and combining a first weight parameter and a second weight parameter to finish up-sampling interpolation to obtain an up-sampling image;
acquiring a guide filtering intensity parameter according to the difference of two local variance statistical histograms of the infrared image and the up-sampling image;
conducting guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, and conducting abnormal monitoring on the clear image to complete visual monitoring of the state of the power transmission line;
the method for acquiring the second weight parameter of each pixel point in each blank reference range comprises the following specific steps:
the gray change characteristics and the deviation included angles of each pixel point in each blank reference range are obtained, any blank is taken as a target blank, and the first blank is in the target blank reference rangeSecond weight parameter of each pixel point +.>The calculation method of (1) is as follows:
wherein ,indicating the%>Deviation included angle of each pixel point to target blank lattice, < >>Indicating the%>Gray scale variation characteristic of each pixel point to target blank lattice, >Minimum value representing gray scale variation characteristic in target blank reference range, < >>To avoid hyper-parameters with denominator 0, < ->Representing absolute value;
and acquiring a second weight parameter of each pixel point in the reference range of each blank lattice.
2. The method for visually monitoring and early warning the state of the power transmission line according to claim 1, wherein the method for obtaining the optimal gaussian kernel size comprises the following specific steps:
acquiring the gray variance of each super pixel region, and acquiring the local variance of each pixel point under different Gaussian kernel sizes;
taking any one Gaussian kernel size as a target Gaussian kernel size, the expression of a first objective function of the target Gaussian kernel size is as follows:
wherein ,the number of pixels representing local variance at target kernel size, +.>Indicate->Local variance of each pixel point under the target Gaussian kernel size, < >>Indicate->Gray variance of super pixel area to which each pixel belongs, < ->Representing absolute value;
and iterating the Gaussian kernel sizes from the initial value according to the increase size, calculating the output value of the first objective function under each Gaussian kernel size, and taking the Gaussian kernel size with the minimum output value as the optimal Gaussian kernel size.
3. The visualized monitoring and early warning method for the state of the power transmission line according to claim 2, wherein the method for obtaining the gray variance of each super-pixel region and obtaining the local variance of each pixel point under different gaussian kernel sizes comprises the following specific steps:
calculating the variance of the gray value of the pixel point in the corresponding region of each super pixel, and marking the variance as the gray variance of each super pixel region;
taking any pixel point in the infrared image as a target pixel point and taking the target pixel point as the center to obtainOther pixels in the local area, wherein +.>The Gaussian kernel size is represented, gray value variances are calculated for the target pixel points and the pixel points in the local range, the gray value variances are recorded as local variances of the target pixel points under the corresponding Gaussian kernel size, and the local variances of the target pixel points under different Gaussian kernel sizes are obtained; and obtaining the local variance of each pixel point under different Gaussian kernel sizes.
4. The method for visualized monitoring and early warning of the state of the power transmission line according to claim 1, wherein the method for obtaining the gray scale change characteristics and the deviation included angles of each pixel point in the reference range of each blank lattice comprises the following specific steps:
taking any one blank as a target blank, taking the average value of the pixel points in the reference range as a planned filling value of the target blank, acquiring any one pixel point in the reference range as a target pixel point, calculating the absolute value of the difference between the gray value of the target pixel point and the planned filling value, and recording the ratio of the absolute value of the difference to the Euclidean distance between the target pixel point and the target blank as the gray change characteristic of the target pixel point to the target blank;
Acquiring a connection line of the target pixel point and the target blank, and marking an acute angle in an included angle formed by the connection line and the extending direction of the power transmission line as a deviated included angle of the target pixel point to the target blank;
and acquiring gray scale change characteristics and deviation included angles of each pixel point in each blank lattice reference range.
5. The method for visualized monitoring and early warning of the state of the power transmission line according to claim 1, wherein the step of obtaining the filling value of each blank cell comprises the following specific steps:
obtaining local variance contribution rates of each pixel point in the guide image and the infrared image respectively, and taking any one blank as a target blank within a reference range of the target blankThe method for calculating the comprehensive weight parameters of each pixel point comprises the following steps:
wherein ,represents the +.o within the target blank reference range>Reference degree of individual pixels, +.>Represents the +.o within the target blank reference range>First weight parameter of each pixel, < ->Represents the +.o within the target blank reference range>Second weight parameter of each pixel, < ->Represents the +.o within the target blank reference range>Local variance contribution rate of each pixel point in guide image,/for the guide image>Represents the +.o within the target blank reference range >Local variance contribution rate of each pixel point in infrared image, < >>To avoid hyper-parameters with denominator 0, < ->Representing absolute value;
obtaining the reference degree of each pixel point in the reference range of the target blank lattice, normalizing all the reference degrees, marking the obtained result as the comprehensive weight parameter of each pixel point, and marking the result obtained by weighting and summing as the filling value of the target blank lattice according to the comprehensive weight parameter and the gray value of each pixel point; and acquiring the filling value of each blank cell.
6. The visualized monitoring and early warning method for the state of the power transmission line according to claim 5, wherein the method for obtaining the local variance contribution rate of each pixel point in the guiding image and the infrared image respectively comprises the following specific steps:
constructing by taking any one pixel point as a target pixel point and taking the target pixel point as the centerThe first sliding windows with the sizes are used for respectively calculating gray variance of the target pixel points in the first sliding windows in the guide image and the infrared image; acquiring gray variance in a first sliding window of each pixel point in the infrared image and the guide image;
normalizing the gray variance in the first sliding window of all the pixel points in the guide image, and marking the obtained result as the local variance contribution rate of each pixel point in the guide image;
And normalizing the gray variance in the first sliding window of all the pixel points in the infrared image, and marking the obtained result as the local variance contribution rate of each pixel point in the infrared image.
7. The method for visually monitoring and early warning the state of a power transmission line according to claim 1, wherein the step of obtaining the guide filtering strength parameter comprises the following specific steps:
acquiring gray variance in all first sliding windows in the infrared image, and generating a local variance statistical histogram of the infrared image; acquiring gray variance in all second sliding windows in the up-sampling image, generating a local variance statistical histogram of the up-sampling image, and guiding filtering strength parametersThe calculation method of (1) is as follows:
wherein ,maximum value representing the number of items with vertical axis values in both histograms, < >>A third part of the statistical histogram of local variances representing an infrared image>Longitudinal axis value of each item with longitudinal axis value, < ->A local variance statistical histogram representing an up-sampled image +.>Longitudinal axis value of each item with longitudinal axis value, < ->An exponential function based on a natural constant is represented.
8. The visualized monitoring and early warning method for the state of the power transmission line according to claim 7, wherein the specific method for acquiring the gray variance in all the second sliding windows in the up-sampled image comprises the following steps:
Taking any one first sliding window in an infrared image as a target first sliding window, acquiring each row and each column of the target first sliding window in the infrared image, extracting each row and each column corresponding to each row in an up-sampling image, extracting each row or each column which is already acquired from adjacent downward rows or adjacent rightward columns in the up-sampling image, and jointly forming a second sliding window corresponding to the target first sliding window in the up-sampling image by a plurality of rows and a plurality of columns extracted from the up-sampling image, wherein the second sliding window is recorded as a target second sliding window;
acquiring gray values of all pixel points in a second sliding window of the target, obtaining gray variance of all pixel points in the second sliding window of the target, and marking the gray variance as the gray variance in the second sliding window of the target; and acquiring gray level variances in all the second sliding windows.
9. The utility model provides a transmission line state visual monitoring early warning system which characterized in that, this system includes:
the infrared image acquisition module is used for acquiring infrared images of the transmission line;
an infrared image processing module: performing super-pixel segmentation on the infrared image, and acquiring the optimal Gaussian kernel size according to the gray variance of each super-pixel region and the local variance of each pixel point under different Gaussian kernel sizes;
obtaining a plurality of blank grids by up-sampling interpolation of an infrared image, obtaining a reference range of each blank grid according to the optimal Gaussian kernel size, and generating a first weight parameter of each pixel point in the reference range according to the Gaussian kernel;
Acquiring the extension direction of the power transmission line, and acquiring a second weight parameter of each pixel point in each blank reference range according to the gray value and the position distribution of each pixel point in each blank reference range;
acquiring a guide image of an infrared image, acquiring a filling value of each blank cell according to gray variances of each pixel point in a first sliding window in the infrared image and the guide image in a reference range of each blank cell, and combining a first weight parameter and a second weight parameter to finish up-sampling interpolation to obtain an up-sampling image;
acquiring a guide filtering intensity parameter according to the difference of two local variance statistical histograms of the infrared image and the up-sampling image;
the power transmission line monitoring module conducts guided filtering on the infrared image according to the guided filtering intensity parameters to obtain a clear image, and conducts abnormal monitoring on the clear image to complete visual monitoring of the power transmission line state;
the method for acquiring the second weight parameter of each pixel point in each blank reference range comprises the following specific steps:
acquiring each blank reference rangeGray scale change characteristics and offset included angles of pixel points, taking any one blank as a target blank, and enabling the target blank to be within a reference range of Second weight parameter of each pixel point +.>The calculation method of (1) is as follows:
wherein ,indicating the%>Deviation included angle of each pixel point to target blank lattice, < >>Indicating the%>Gray scale variation characteristic of each pixel point to target blank lattice,>minimum value representing gray scale variation characteristic in target blank reference range, < >>To avoid hyper-parameters with denominator 0, < ->Representing absolute value;
and acquiring a second weight parameter of each pixel point in the reference range of each blank lattice.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310720414.0A CN116452594B (en) | 2023-06-19 | 2023-06-19 | Visualized monitoring and early warning method and system for power transmission line state |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310720414.0A CN116452594B (en) | 2023-06-19 | 2023-06-19 | Visualized monitoring and early warning method and system for power transmission line state |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116452594A CN116452594A (en) | 2023-07-18 |
| CN116452594B true CN116452594B (en) | 2023-08-29 |
Family
ID=87122352
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310720414.0A Active CN116452594B (en) | 2023-06-19 | 2023-06-19 | Visualized monitoring and early warning method and system for power transmission line state |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116452594B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116912096B (en) * | 2023-09-14 | 2024-01-09 | 惠州市耀盈精密技术有限公司 | IMOLD system-based precision mold template image generation method |
| CN116993643B (en) * | 2023-09-27 | 2023-12-12 | 山东建筑大学 | Unmanned aerial vehicle photogrammetry image correction method based on artificial intelligence |
| CN117522863B (en) * | 2023-12-29 | 2024-03-29 | 临沂天耀箱包有限公司 | Integrated box body quality detection method based on image features |
| CN117522758B (en) * | 2024-01-04 | 2024-03-26 | 深圳对对科技有限公司 | Smart community resource management method and system based on big data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784637A (en) * | 2017-09-30 | 2018-03-09 | 烟台艾睿光电科技有限公司 | A kind of method of infrared image enhancement |
| CN108376391A (en) * | 2018-01-29 | 2018-08-07 | 江苏宇特光电科技股份有限公司 | A kind of intelligence infrared image scene Enhancement Method |
| CN111369458A (en) * | 2020-02-28 | 2020-07-03 | 中国人民解放军空军工程大学 | Infrared dim target background suppression method based on multi-scale rolling guide filtering smoothing |
| WO2021196822A1 (en) * | 2020-03-31 | 2021-10-07 | 电子科技大学 | Adaptive self-guided filtering-based loop filtering method |
| CN116012346A (en) * | 2023-02-01 | 2023-04-25 | 江苏天南电力股份有限公司 | Power transmission line monitoring method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10223772B2 (en) * | 2016-03-22 | 2019-03-05 | Algolux Inc. | Method and system for denoising and demosaicing artifact suppression in digital images |
-
2023
- 2023-06-19 CN CN202310720414.0A patent/CN116452594B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784637A (en) * | 2017-09-30 | 2018-03-09 | 烟台艾睿光电科技有限公司 | A kind of method of infrared image enhancement |
| CN108376391A (en) * | 2018-01-29 | 2018-08-07 | 江苏宇特光电科技股份有限公司 | A kind of intelligence infrared image scene Enhancement Method |
| CN111369458A (en) * | 2020-02-28 | 2020-07-03 | 中国人民解放军空军工程大学 | Infrared dim target background suppression method based on multi-scale rolling guide filtering smoothing |
| WO2021196822A1 (en) * | 2020-03-31 | 2021-10-07 | 电子科技大学 | Adaptive self-guided filtering-based loop filtering method |
| CN116012346A (en) * | 2023-02-01 | 2023-04-25 | 江苏天南电力股份有限公司 | Power transmission line monitoring method and system |
Non-Patent Citations (1)
| Title |
|---|
| Gradient Domain Weighted Guided Image Filtering;Wang Bo等;《arXiv:2211.16796 [cs.CV]》;第1-29页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116452594A (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116452594B (en) | Visualized monitoring and early warning method and system for power transmission line state | |
| CN108961235B (en) | Defective insulator identification method based on YOLOv3 network and particle filter algorithm | |
| CN110598851A (en) | Time series data abnormity detection method fusing LSTM and GAN | |
| CN102609904B (en) | Bivariate nonlocal average filtering de-noising method for X-ray image | |
| CN108596087B (en) | Driving fatigue degree detection regression model based on double-network result | |
| CN115797354B (en) | Method for detecting appearance defects of laser welding seam | |
| CN114820401B (en) | Method for enhancing marine backlight infrared image by combining histogram transformation and edge information | |
| CN107749058A (en) | A kind of machine vision detection method and system of boiler tubing surface defect | |
| CN116502478B (en) | Submarine topography monitoring-based pile-off auxiliary decision-making method for self-lifting platform | |
| CN117633696B (en) | Hydraulic engineering construction monitoring method and system based on BIM model | |
| CN111402249A (en) | Image evolution analysis method based on deep learning | |
| CN107944497A (en) | Image block method for measuring similarity based on principal component analysis | |
| CN113378672A (en) | Multi-target detection method for defects of power transmission line based on improved YOLOv3 | |
| CN117152203A (en) | Target tracking method and device based on multi-core correlation filter and electronic equipment | |
| CN116399592A (en) | Bearing fault diagnosis method based on channel attention dual-path feature extraction | |
| WO2022206149A1 (en) | Three-dimensional spectrum situation completion method and apparatus based on generative adversarial network | |
| CN110517226A (en) | The offal method for extracting region of multiple features texture image fusion based on bilateral filtering | |
| CN108764311A (en) | A kind of shelter target detection method, electronic equipment, storage medium and system | |
| CN114092841A (en) | Substation unmanned aerial vehicle inspection image edge identification method and system and storage medium | |
| CN113919396A (en) | Vibration signal and image characteristic machine tool cutter wear state monitoring method based on semi-supervised learning | |
| CN117409376B (en) | Infrared online monitoring method and system for high-voltage sleeve | |
| CN117237902B (en) | Robot character recognition system based on deep learning | |
| CN113496159B (en) | Multi-scale convolution and dynamic weight cost function smoke target segmentation method | |
| CN112459970B (en) | Wind turbine generator anomaly detection method using data counterstudy | |
| CN113379666B (en) | Image fusion method based on interval type fuzzy set and CHMM model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |