CN117853962A - Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method - Google Patents
Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method Download PDFInfo
- Publication number
- CN117853962A CN117853962A CN202410256796.0A CN202410256796A CN117853962A CN 117853962 A CN117853962 A CN 117853962A CN 202410256796 A CN202410256796 A CN 202410256796A CN 117853962 A CN117853962 A CN 117853962A
- Authority
- CN
- China
- Prior art keywords
- porcelain insulator
- real
- image
- light
- time
- 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.)
- Granted
Links
- 239000012212 insulator Substances 0.000 title claims abstract description 210
- 229910052573 porcelain Inorganic materials 0.000 title claims abstract description 193
- 238000000034 method Methods 0.000 title claims abstract description 74
- 238000003708 edge detection Methods 0.000 title claims abstract description 27
- 230000004927 fusion Effects 0.000 title claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims abstract description 41
- 230000009466 transformation Effects 0.000 claims abstract description 24
- 238000009792 diffusion process Methods 0.000 claims abstract description 7
- 238000000605 extraction Methods 0.000 claims abstract description 6
- 238000006243 chemical reaction Methods 0.000 claims description 13
- 238000010586 diagram Methods 0.000 claims description 13
- 238000001914 filtration Methods 0.000 claims description 11
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000012634 optical imaging Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010292 electrical insulation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- 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
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection, which comprises the following steps: collecting micro-light and infrared pictures of a porcelain insulator, training a YOLO-V5 model, detecting the position of the porcelain insulator by using the model, rotating a lens to center the porcelain insulator, carrying out graying and feature matching on micro-light and infrared images of the porcelain insulator, calculating a feature matrix, carrying out matrix transformation, binarization and edge contour extraction on the micro-light gray level pictures, and fusing the contour with the infrared images; the S-ORB operator is used for enhancing feature extraction, and meanwhile, a rapid explicit diffusion FED algorithm is used for avoiding sinking into a linear optimal solution, so that a large amount of solving time can be reduced, an optimal approximate solution can be obtained rapidly, and the method has higher technical advancement and wide applicability.
Description
Technical Field
The invention relates to the technical field of insulator detection, in particular to a porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection.
Background
The insulator is an important component for mechanical connection and electrical insulation of the transmission line, and has good mechanical performance and electrical performance; when zero-value insulators appear in the aging of the insulator string, the overall creepage distance is reduced, so that the flashover probability of the insulator string is increased; when lightning overvoltage acts on the zero-value insulator, the zero-value insulator is completely broken down, and strong lightning current and power frequency follow current flow through a gap at the head of the zero-value insulator, so that the safety and stability operation of a power system are not facilitated.
In a power transmission line overhaul site, a common non-contact detection means is used for shooting an infrared image of a target for unmanned aerial vehicle carrying equipment, extracting a temperature diagnosis criterion in the image and determining a defect insulator; however, in the existing insulator shooting technology, the detection is limited by the number of pixels of an infrared camera and the shooting position angle, so that the contour of an infrared image insulator is blurred, and the subsequent infrared image is difficult to detect.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection, which aims to solve the difficult problem in the technical field of fuzzy infrared intelligent detection of the profile of the porcelain insulator and improve the accuracy and efficiency of the subsequent detection of the porcelain insulator.
In order to achieve the above purpose, the present invention provides the following technical solutions: the porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection comprises the following steps:
step S1: constructing an infrared shimmer data set of the porcelain insulator, and training a YOLO-V5 model by adopting the infrared shimmer data set of the porcelain insulator to obtain a trained YOLO-V5 model;
step S2: identifying the porcelain insulator in the real-time porcelain insulator image by adopting a trained YOLO-V5 model, calculating to obtain an included angle according to the pixel difference value between the identified porcelain insulator center and the real-time porcelain insulator image center and an optical imaging principle, adjusting the deflection angle of a camera according to the included angle, and acquiring an infrared image and a low-light image of the real-time porcelain insulator by using the adjusted camera;
step S3: converting the infrared image and the low-light image of the real-time porcelain insulator into an infrared inverse gray level image of the real-time porcelain insulator and a low-light gray level image of the real-time porcelain insulator; calculating characteristic points of a real-time porcelain insulator low-light gray level image and a real-time insulator infrared inverse gray level image by using an S-ORB operator, and solving a nonlinear scale space matrix by using a fast explicit diffusion FED algorithm;
step S4: performing image splicing on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix; performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator and double-neighborhood operator edge detection on the real-time porcelain insulator micro-light gray level map after matrix transformation;
the specific process of performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator edge detection on the real-time porcelain insulator low-light gray level map after matrix transformation is as follows: performing self-adaptive median filtering operation on the real-time porcelain insulator low-light gray level transformation diagram to remove scattered point interference; binarizing a threshold value of the real-time porcelain insulator low-light gray level conversion chart after removing scattered point interference to obtain a real-time porcelain insulator low-light gray level conversion binary chart; searching a first white pixel position of each row of pixels of the real-time porcelain insulator low-light gray level conversion binary image, judging whether a right pixel of a second white pixel is consistent with a pixel value of the second white pixel by using a single neighborhood operator from a second white pixel position of each row of pixels, judging that the pixels are edges if gradient differences occur in the pixel values, stopping executing, continuing executing a double neighborhood operator if the pixel values are consistent, judging that the pixels on the upper side and the lower side of the white pixels are gradient differences, and stopping executing to obtain a low-light porcelain insulator edge contour binary image; wherein the white pixel locations are necessarily edges;
step S5: and fusing the extracted low-light porcelain insulator edge contour binary image with an infrared image of the porcelain insulator.
Further, the specific process of converting the infrared image and the low-light image of the real-time porcelain insulator into the infrared inverse gray level image and the low-light gray level image of the real-time porcelain insulator is as follows: the size of the real-time porcelain insulator micro-light image is adjusted to be equal to the size of the real-time porcelain insulator infrared image;
and carrying out graying and inverse graying on the real-time porcelain insulator infrared image and the real-time porcelain insulator low-light image after the size adjustment by using a weighted average method.
Further, calculating characteristic points of a real-time porcelain insulator low-light gray level image and a real-time insulator infrared inverse gray level image by using an S-ORB operator, respectively selecting a 5X 5 matrix by taking each pixel of the real-time porcelain insulator low-light gray level image and the real-time insulator infrared inverse gray level image as a central point, comparing the central point with 24 pixels around the central point, wherein the difference is 20% of pixel values below the central point and more than 20 are one characteristic point; the nonlinear scale space matrix is solved by using a fast explicit diffusion FED algorithm, and the formula is as follows:
；
；
；
；
；
；
in the method, in the process of the invention,representing the size space of a real-time porcelain insulator micro-light gray scale image;a gaussian function representing a varying scale;representing the gradient magnitude;representing the gradient direction;representing a convolution kernel; m, n represent the width and height of the Gaussian template convolution kernel;representing the position of a real-time porcelain insulator micro-light gray scale image element;representing a scale space factor;representing the evolution time, and making a nonlinear relation with the scale space factor;a scale-space factor square representing the current j time;representing a feature point scale space; h= { h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 And (c) represents a non-linear scale space matrix, where h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 Are all constants;representing the base of the logarithmic function ln; x is x 1 、y 1 Representing the coordinates after matrix transformation; c. c 1 Representing a constant.
Further, the specific process of performing image stitching on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix is as follows: performing matrix transformation on the real-time porcelain insulator low-light gray level image to obtain an image matrix of the real-time porcelain insulator low-light gray level image, multiplying the image matrix of the real-time porcelain insulator low-light gray level image by a nonlinear scale space matrix to obtain a real-time porcelain insulator low-light gray level transformation image, and replacing pixels of the corresponding coordinate positions of the real-time porcelain insulator low-light gray level transformation image with the real-time porcelain insulator infrared inverse gray level image.
Further, the formula for performing self-adaptive median filtering operation on the real-time porcelain insulator micro-light gray level transformation diagram to remove scattered point interference is as follows:
；
in the method, in the process of the invention,representing a gray value;a median value representing gray values within the filter window;minimum value representing gray scale values within a filter window；Representing the maximum value of the gray values within the filter window;is the position of the real-time porcelain insulator micro-light gray scale image elementIs a pixel value of (a).
Further, the formula for obtaining the real-time porcelain insulator micro-light gray level conversion binary image by binarizing the threshold value of the real-time porcelain insulator micro-light gray level conversion image after removing the scattered point interference is as follows:
；
in the method, in the process of the invention,converting the gray scale of the micro light of the real-time porcelain insulator into a binary image pixel value;is the pixel gray value.
Further, the fusion formula of the edge profile binary image of the micro-light porcelain insulator and the infrared image of the porcelain insulator is as follows:
；
in the method, in the process of the invention,representing an edge enhancement diagram of the porcelain insulator after fusion;an infrared image representing the porcelain insulator;and (5) representing a two-value diagram of the edge profile of the micro-light porcelain insulator.
Further, the specific process for constructing the porcelain insulator infrared low-light data set is as follows: the infrared image and the low-light image of the porcelain insulator are acquired through the camera carried by the unmanned aerial vehicle, and the infrared image and the porcelain insulator in the low-light image of the porcelain insulator are marked and stored by using a labelimg third-party tool to manufacture an infrared low-light data set of the porcelain insulator.
Further, the feature extraction network of the YOLO-V5 model adopts a shuffle 2 lightweight network.
A non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions for executing the porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection.
Compared with the prior art, the invention has the following beneficial effects: the invention creatively provides a single-neighborhood and double-neighborhood edge detection method, which can more rapidly traverse and search the image edge; by using nonlinear feature extraction, the method can avoid sinking into a linear optimal solution, can reduce a large amount of solving time, can quickly obtain an optimal approximate solution, and has higher technical advancement and wide applicability.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a schematic diagram of the S-ORB operator of the present invention.
FIG. 3 is a schematic diagram of a single-double neighborhood operator of the present invention.
Detailed Description
As shown in fig. 1, the present invention provides the following technical solutions: the porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection comprises the following steps:
step S1: constructing an infrared shimmer data set of the porcelain insulator, and training a YOLO-V5 model by adopting the infrared shimmer data set of the porcelain insulator to obtain a trained YOLO-V5 model; the feature extraction network of the YOLO-V5 model adopts a shuffle 2 lightweight network;
step S2: identifying the porcelain insulator in the real-time porcelain insulator image by adopting a trained YOLO-V5 model, calculating to obtain an included angle according to the pixel difference value between the identified porcelain insulator center and the real-time porcelain insulator image center and an optical imaging principle, adjusting the deflection angle of a camera according to the included angle, and acquiring an infrared image and a low-light image of the real-time porcelain insulator by using the adjusted camera;
step S3: converting the infrared image and the low-light image of the real-time porcelain insulator into an infrared inverse gray level image of the real-time porcelain insulator and a low-light gray level image of the real-time porcelain insulator; calculating characteristic points of a real-time porcelain insulator low-light gray level image and a real-time insulator infrared inverse gray level image by using an S-ORB operator, and solving a nonlinear scale space matrix by using a fast explicit diffusion FED algorithm;
step S4: performing image splicing on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix; performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator and double-neighborhood operator edge detection on the real-time porcelain insulator micro-light gray level map after matrix transformation;
step S5: and fusing the extracted low-light porcelain insulator edge contour binary image with an infrared image of the porcelain insulator.
The specific process for constructing the porcelain insulator infrared low-light data set comprises the following steps: the infrared image and the low-light image of the porcelain insulator are acquired through the camera carried by the unmanned aerial vehicle, and the infrared image and the porcelain insulator in the low-light image of the porcelain insulator are marked and stored by using a labelimg third-party tool to manufacture an infrared low-light data set of the porcelain insulator.
The specific process of converting the infrared image and the low-light image of the real-time porcelain insulator into the infrared inverse gray level image of the real-time porcelain insulator and the low-light gray level image of the real-time porcelain insulator is as follows: the size of the real-time porcelain insulator micro-light image is adjusted to be equal to the size of the real-time porcelain insulator infrared image;
and (3) graying the real-time porcelain insulator infrared image and the real-time porcelain insulator low-light image with the adjusted size by using a weighted average method, wherein the formula is as follows:
（1）；
in the method, in the process of the invention,、、three channels of real-time porcelain insulator micro-light images after the real-time porcelain insulator infrared images are adjusted in size;gray values at positions of i columns and j rows;the method is characterized in that the method comprises the steps of setting the i column j row positions of real-time porcelain insulator low-light images after the real-time porcelain insulator infrared images are subjected to size adjustment;
the inverse graying formula is as follows:
（2）。
as shown in fig. 2, the characteristic points of the real-time porcelain insulator low-light gray level image and the real-time insulator infrared inverse gray level image are calculated by using an S-ORB operator, each pixel of the real-time porcelain insulator low-light gray level image and the real-time insulator infrared inverse gray level image is used as a center point to select a 5×5 matrix, the center point is compared with 24 pixels around the center point, the difference is 20% less than the center point pixel value and more than 20 are one characteristic point, and a nonlinear scale space matrix is solved by using a fast explicit diffusion FED algorithm, wherein the formula is as follows:
（3）；
（4）；
（5）；
（6）；
（7）；
（8）；
in the method, in the process of the invention,representing the size space of a real-time porcelain insulator micro-light gray scale image;a gaussian function representing a varying scale;representing the gradient magnitude;representing the gradient direction;representing a convolution kernel; m, n represent the width and height of the Gaussian template convolution kernel;representing the position of a real-time porcelain insulator micro-light gray scale image element;representing a scale space factor;representing the evolution time, and making a nonlinear relation with the scale space factor;a scale-space factor square representing the current j time;representing a feature point scale space; h= { h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 And (c) represents a non-linear scale space matrix, where h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 Are all constants;representing the base of the logarithmic function ln; x is x 1 、y 1 Representing the coordinates after matrix transformation; c. c 1 Representing a constant.
The specific process for performing image splicing on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix comprises the following steps: performing matrix transformation on the real-time porcelain insulator low-light gray level image to obtain an image matrix of the real-time porcelain insulator low-light gray level image, multiplying the image matrix of the real-time porcelain insulator low-light gray level image by a nonlinear scale space matrix to obtain a real-time porcelain insulator low-light gray level transformation image, and replacing pixels of the corresponding coordinate positions of the real-time porcelain insulator low-light gray level transformation image with the real-time porcelain insulator infrared inverse gray level image.
The method comprises the specific processes of performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator edge detection on a real-time porcelain insulator low-light gray level map after matrix transformation, wherein the specific processes are as follows:
performing self-adaptive median filtering operation on the real-time porcelain insulator low-light gray level transformation diagram to remove scattered point interference, wherein the formula is as follows:
（9）；
in the method, in the process of the invention,representing a gray value;a median value representing gray values within the filter window;a minimum value representing a gray value within the filter window;representing the maximum value of the gray values within the filter window;is the position of the real-time porcelain insulator micro-light gray scale image elementPixel values of (2);
and binarizing the threshold value of the real-time porcelain insulator low-light gray level conversion chart after removing the scattered point interference to obtain a real-time porcelain insulator low-light gray level conversion binary chart, wherein the formula is as follows:
（10）；
as shown in the figure 3 of the drawings,converting the gray scale of the micro light of the real-time porcelain insulator into a binary image pixel value;is the pixel gray value;
searching a first white pixel B position (the white pixel position is an edge) of each row of pixels of the real-time porcelain insulator micro-light gray level conversion binary image, judging whether a right pixel B2 of a second white pixel B1 is consistent with a pixel value of the second white pixel B1 by using a single neighborhood operator from the second white pixel B1 position of each row of pixels, judging the edge if a gradient difference occurs in the pixel value, stopping executing, continuing executing a double neighborhood operator if the pixel value is consistent, judging that gradient differences exist in upper and lower side pixels of the white pixel B1, and stopping to obtain the micro-light porcelain insulator edge contour binary image;
by adopting the single-neighborhood operator and the double-neighborhood operator to perform eight-neighborhood search of the edge detection relative to the Canny operator, more than half of time can be saved, and the efficiency is greatly improved.
The fusion formula of the edge profile binary image of the micro-light porcelain insulator and the infrared image of the porcelain insulator is as follows:
（11）；
in the method, in the process of the invention,representing an edge enhancement diagram of the porcelain insulator after fusion;an infrared image representing the porcelain insulator;and (5) representing a two-value diagram of the edge profile of the micro-light porcelain insulator.
A non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions for executing the porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection is characterized by comprising the following steps of:
step S1: constructing an infrared shimmer data set of the porcelain insulator, and training a YOLO-V5 model by adopting the infrared shimmer data set of the porcelain insulator to obtain a trained YOLO-V5 model;
step S2: identifying the porcelain insulator in the real-time porcelain insulator image by adopting a trained YOLO-V5 model, calculating to obtain an included angle according to the pixel difference value between the identified porcelain insulator center and the real-time porcelain insulator image center and an optical imaging principle, adjusting the deflection angle of a camera according to the included angle, and acquiring an infrared image and a low-light image of the real-time porcelain insulator by using the adjusted camera;
step S3: converting the infrared image and the low-light image of the real-time porcelain insulator into an infrared inverse gray level image of the real-time porcelain insulator and a low-light gray level image of the real-time porcelain insulator; calculating characteristic points of a real-time porcelain insulator low-light gray level image and a real-time insulator infrared inverse gray level image by using an S-ORB operator, and solving a nonlinear scale space matrix by using a fast explicit diffusion FED algorithm;
step S4: performing image splicing on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix; performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator and double-neighborhood operator edge detection on the real-time porcelain insulator micro-light gray level map after matrix transformation;
the specific process of performing self-adaptive threshold binarization, filtering and selective single-neighborhood operator edge detection on the real-time porcelain insulator low-light gray level map after matrix transformation is as follows: performing self-adaptive median filtering operation on the real-time porcelain insulator low-light gray level transformation diagram to remove scattered point interference; binarizing a threshold value of the real-time porcelain insulator low-light gray level conversion chart after removing scattered point interference to obtain a real-time porcelain insulator low-light gray level conversion binary chart; searching a first white pixel position of each row of pixels of the real-time porcelain insulator low-light gray level conversion binary image, judging whether a right pixel of a second white pixel is consistent with a pixel value of the second white pixel by using a single neighborhood operator from a second white pixel position of each row of pixels, judging that the pixels are edges if gradient differences occur in the pixel values, stopping executing, continuing executing a double neighborhood operator if the pixel values are consistent, judging that the pixels on the upper side and the lower side of the white pixels are gradient differences, and stopping executing to obtain a low-light porcelain insulator edge contour binary image; wherein the white pixel locations are necessarily edges;
step S5: and fusing the extracted low-light porcelain insulator edge contour binary image with an infrared image of the porcelain insulator.
2. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 1, wherein the method is characterized by comprising the following steps: the specific process of converting the infrared image and the low-light image of the real-time porcelain insulator into the infrared inverse gray level image of the real-time porcelain insulator and the low-light gray level image of the real-time porcelain insulator is as follows: the size of the real-time porcelain insulator micro-light image is adjusted to be equal to the size of the real-time porcelain insulator infrared image;
and carrying out graying and inverse graying on the real-time porcelain insulator infrared image and the real-time porcelain insulator low-light image after the size adjustment by using a weighted average method.
3. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 2, wherein the method is characterized by comprising the following steps: calculating characteristic points of a real-time porcelain insulator micro-light gray level image and a real-time insulator infrared inverse gray level image by using an S-ORB operator, respectively selecting a 5X 5 matrix by taking each pixel of the real-time porcelain insulator micro-light gray level image and the real-time insulator infrared inverse gray level image as a central point, comparing the central point with 24 pixels around the central point, wherein the difference is 20% of pixel values lower than the central point and more than 20 are one characteristic point; the nonlinear scale space matrix is solved by using a fast explicit diffusion FED algorithm, and the formula is as follows:
；
；
；
；
；
；
in the method, in the process of the invention,representing the size space of a real-time porcelain insulator micro-light gray scale image; />A gaussian function representing a varying scale; />Representing the gradient magnitude; />Representing the gradient direction; />Representing a convolution kernel; m, n represent the width and height of the Gaussian template convolution kernel; />Representing the position of a real-time porcelain insulator micro-light gray scale image element; />Representing a scale space factor; />Representing the evolution time, and making a nonlinear relation with the scale space factor; />A scale-space factor square representing the current j time;representing a feature point scale space; h= { h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 And (c) represents a non-linear scale space matrix, where h 00 ，h 01 ，h 02 ，h 10 ，h 11 ，h 12 ，h 20 ，h 21 ，h 22 Are all constants; />Representing the base of the logarithmic function ln; x is x 1 、y 1 Representing the coordinates after matrix transformation; c. c 1 Representing a constant.
4. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 3, wherein the method comprises the following steps: the specific process for performing image splicing on the real-time porcelain insulator low-light gray level image and the real-time porcelain insulator infrared inverse gray level image according to the nonlinear scale space matrix comprises the following steps: performing matrix transformation on the real-time porcelain insulator low-light gray level image to obtain an image matrix of the real-time porcelain insulator low-light gray level image, multiplying the image matrix of the real-time porcelain insulator low-light gray level image by a nonlinear scale space matrix to obtain a real-time porcelain insulator low-light gray level transformation image, and replacing pixels of the corresponding coordinate positions of the real-time porcelain insulator low-light gray level transformation image with the real-time porcelain insulator infrared inverse gray level image.
5. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 4, wherein the method is characterized by comprising the following steps: the formula for removing scattered point interference by performing self-adaptive median filtering operation on the real-time porcelain insulator low-light gray level transformation diagram is as follows:
；
in the method, in the process of the invention,representing a gray value; />A median value representing gray values within the filter window; />A minimum value representing a gray value within the filter window; />Representing the maximum value of the gray values within the filter window; />The position of the micro-light gray level image element of the real-time porcelain insulator is +.>Is a pixel value of (a).
6. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 5, wherein the method is characterized by comprising the following steps: the formula for obtaining the real-time porcelain insulator micro-light gray level conversion binary image by binarizing the threshold value of the real-time porcelain insulator micro-light gray level conversion image after removing the scattered point interference is as follows:
；
in the method, in the process of the invention,converting the gray scale of the micro light of the real-time porcelain insulator into a binary image pixel value; />Is the pixel gray value.
7. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection of claim 6, wherein the method is characterized by comprising the following steps: the fusion formula of the edge contour binary image of the micro-light porcelain insulator and the infrared image of the porcelain insulator is as follows:
；
in the method, in the process of the invention,representing an edge enhancement diagram of the porcelain insulator after fusion; />An infrared image representing the porcelain insulator;and (5) representing a two-value diagram of the edge profile of the micro-light porcelain insulator.
8. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 1, wherein the method is characterized by comprising the following steps: the specific process for constructing the porcelain insulator infrared low-light data set is as follows: the infrared image and the low-light image of the porcelain insulator are acquired through the camera carried by the unmanned aerial vehicle, and the infrared image and the porcelain insulator in the low-light image of the porcelain insulator are marked and stored by using a labelimg third-party tool to manufacture an infrared low-light data set of the porcelain insulator.
9. The porcelain insulator micro-light infrared fusion sensing method based on single-double neighborhood edge detection according to claim 1, wherein the method is characterized by comprising the following steps: the feature extraction network of the YOLO-V5 model adopts a shuffle 2 lightweight network.
10. A non-volatile computer storage medium storing computer executable instructions for performing the porcelain insulator micro-optic infrared fusion sensing method based on single-double neighborhood edge detection according to any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410256796.0A CN117853962B (en) | 2024-03-07 | 2024-03-07 | Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410256796.0A CN117853962B (en) | 2024-03-07 | 2024-03-07 | Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117853962A true CN117853962A (en) | 2024-04-09 |
| CN117853962B CN117853962B (en) | 2024-07-05 |
Family
ID=90530531
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410256796.0A Active CN117853962B (en) | 2024-03-07 | 2024-03-07 | Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117853962B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006017233A1 (en) * | 2004-07-12 | 2006-02-16 | Lehigh University | Image fusion methods and apparatus |
| CN101620676A (en) * | 2009-07-02 | 2010-01-06 | 浙江省电力公司 | Fast image recognition method of insulator contour |
| CN101853492A (en) * | 2010-05-05 | 2010-10-06 | 浙江理工大学 | Method for fusing night-viewing twilight image and infrared image |
| CN104361571A (en) * | 2014-11-21 | 2015-02-18 | 南京理工大学 | Infrared and low-light image fusion method based on marginal information and support degree transformation |
| CN107705274A (en) * | 2017-08-21 | 2018-02-16 | 中国核电工程有限公司 | A kind of low-light and infrared image fusion method multiple dimensioned based on mathematical morphology |
| CN115588021A (en) * | 2022-11-04 | 2023-01-10 | 国网湖北省电力有限公司荆州供电公司 | Ceramic insulator and hardware unmanned aerial vehicle infrared inspection image fusion and segmentation method |
| CN115689960A (en) * | 2022-10-27 | 2023-02-03 | 长春理工大学 | Illumination self-adaptive infrared and visible light image fusion method in night scene |
-
2024
- 2024-03-07 CN CN202410256796.0A patent/CN117853962B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006017233A1 (en) * | 2004-07-12 | 2006-02-16 | Lehigh University | Image fusion methods and apparatus |
| CN101620676A (en) * | 2009-07-02 | 2010-01-06 | 浙江省电力公司 | Fast image recognition method of insulator contour |
| CN101853492A (en) * | 2010-05-05 | 2010-10-06 | 浙江理工大学 | Method for fusing night-viewing twilight image and infrared image |
| CN104361571A (en) * | 2014-11-21 | 2015-02-18 | 南京理工大学 | Infrared and low-light image fusion method based on marginal information and support degree transformation |
| CN107705274A (en) * | 2017-08-21 | 2018-02-16 | 中国核电工程有限公司 | A kind of low-light and infrared image fusion method multiple dimensioned based on mathematical morphology |
| CN115689960A (en) * | 2022-10-27 | 2023-02-03 | 长春理工大学 | Illumination self-adaptive infrared and visible light image fusion method in night scene |
| CN115588021A (en) * | 2022-11-04 | 2023-01-10 | 国网湖北省电力有限公司荆州供电公司 | Ceramic insulator and hardware unmanned aerial vehicle infrared inspection image fusion and segmentation method |
Non-Patent Citations (2)
| Title |
|---|
| ZHANG, Z ET AL: "Bionic Algorithm for Color Fusion of Infrared and Low Light Level Image Based on Rattlesnake Bimodal Cells", 《IEEE ACCESS》, 31 December 2018 (2018-12-31), pages 68981 - 68988 * |
| 康辉等: "基于融合算法的电力设备红外图像增强", 《黑龙江科技大学学报》, vol. 32, no. 4, 31 July 2022 (2022-07-31), pages 545 - 562 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117853962B (en) | 2024-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108961235B (en) | Defective insulator identification method based on YOLOv3 network and particle filter algorithm | |
| CN109785285B (en) | Insulator damage detection method based on ellipse characteristic fitting | |
| CN110084241B (en) | Automatic ammeter reading method based on image recognition | |
| CN109472776B (en) | Depth significance-based insulator detection and self-explosion identification method | |
| CN108734689B (en) | Method for detecting scattered strands of conducting wires based on region growth | |
| CN112819094A (en) | Target detection and identification method based on structural similarity measurement | |
| CN112560619B (en) | Multi-focus image fusion-based multi-distance bird accurate identification method | |
| CN116721107B (en) | Intelligent monitoring system for cable production quality | |
| CN114241364A (en) | Method for quickly calibrating foreign object target of overhead transmission line | |
| CN110427979B (en) | Road water pit identification method based on K-Means clustering algorithm | |
| CN106875430B (en) | Single moving target tracking method and device based on fixed form under dynamic background | |
| CN112561899A (en) | Electric power inspection image identification method | |
| CN115937160A (en) | Explosion fireball contour detection method based on convex hull algorithm | |
| CN116167999A (en) | Distribution line zero-value insulator infrared thermal image detection method based on image matching | |
| CN112509026A (en) | Insulator crack length identification method | |
| CN116503470A (en) | Method for segmenting, positioning and analyzing states of insulator string disc in infrared image | |
| CN114241310B (en) | Improved YOLO model-based intelligent identification method for piping dangerous case of dike | |
| CN112528994B (en) | Free angle license plate detection method, license plate recognition method and recognition system | |
| CN117409083B (en) | Cable terminal identification method and device based on infrared image and improved YOLOV5 | |
| CN117853962B (en) | Single-double neighborhood edge detection-based porcelain insulator micro-light infrared fusion sensing method | |
| CN108734054B (en) | Non-shielding citrus fruit image identification method | |
| CN116485802B (en) | Insulator flashover defect detection method, device, equipment and storage medium | |
| CN113378672A (en) | Multi-target detection method for defects of power transmission line based on improved YOLOv3 | |
| CN108830834B (en) | Automatic extraction method for video defect information of cable climbing robot | |
| CN113095309B (en) | Method for extracting road scene ground marker based on point cloud |
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 |