CN116862918B - Real-time detection method for condensation inside ring main unit based on artificial intelligence - Google Patents
Real-time detection method for condensation inside ring main unit based on artificial intelligence Download PDFInfo
- Publication number
- CN116862918B CN116862918B CN202311133033.9A CN202311133033A CN116862918B CN 116862918 B CN116862918 B CN 116862918B CN 202311133033 A CN202311133033 A CN 202311133033A CN 116862918 B CN116862918 B CN 116862918B
- Authority
- CN
- China
- Prior art keywords
- window
- pixel point
- heat source
- determining
- temperature
- 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
- 238000009833 condensation Methods 0.000 title claims abstract description 27
- 230000005494 condensation Effects 0.000 title claims abstract description 27
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 12
- 238000011897 real-time detection Methods 0.000 title claims abstract description 10
- 238000001931 thermography Methods 0.000 claims abstract description 16
- 230000005855 radiation Effects 0.000 claims description 53
- 238000004364 calculation method Methods 0.000 claims description 17
- 101150017422 HTR1 gene Proteins 0.000 claims description 7
- 101100024116 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) MPT5 gene Proteins 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 3
- 230000007306 turnover Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 4
- 101150017459 fer1 gene Proteins 0.000 description 4
- UJHBVMHOBZBWMX-UHFFFAOYSA-N ferrostatin-1 Chemical compound NC1=CC(C(=O)OCC)=CC=C1NC1CCCCC1 UJHBVMHOBZBWMX-UHFFFAOYSA-N 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000010438 heat treatment Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 239000000428 dust Substances 0.000 description 2
- 230000017525 heat dissipation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009529 body temperature measurement Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual 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
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Radiation Pyrometers (AREA)
Abstract
The application relates to the field of image processing, and provides an artificial intelligence-based ring main unit interior condensation real-time detection method, which comprises the following steps: determining a first confidence coefficient that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence coefficient that a similar window in the search window comprises the heat source pixel point; determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window; determining the gray weight of the central pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, so as to determine the gray weight of each pixel point in the image to be detected; denoising the image to be detected based on the gray weight of each pixel point in the image to be detected. According to the method, the thermal imaging image is improved by denoising the thermal imaging image, and the accurate condition inside the ring main unit is detected in real time through the image.
Description
Technical Field
The application relates to the field of image processing, in particular to an artificial intelligence-based real-time detection method for condensation inside a ring main unit.
Background
The ring main unit is power voltage converting equipment, is widely applied to various power systems such as urban power grids and industrial power grids at present, and gradually becomes an important component of the urban power systems. Because the power equipment in the looped netowrk cabinet continuously generates electricity to generate heat, the temperature in the looped netowrk cabinet is higher than the ambient temperature, when the external ambient temperature reduces, when the inside and outside temperature difference of producing to a certain extent of looped netowrk cabinet, the vapor in the looped netowrk cabinet can form the condensation on the inner wall, and the condensation can drop when certain dust is accumulated in the condensation, and the power equipment operation is problematic.
At present, condensation detection in the ring main unit mainly depends on monitoring humidity and temperature in the ring main unit, and when the humidity and the temperature are at a certain threshold value, monitoring personnel judge that condensation possibly occurs in the ring main unit through calculation or table lookup and go to the ring main unit for verification. Because the environment in the looped netowrk cabinet is complicated, the running power of power equipment is the condition such as not equity, just through temperature and humidity to in the looped netowrk cabinet monitor the accurate condition in the accuse looped netowrk cabinet of being difficult to.
Disclosure of Invention
The application provides a real-time detection method of condensation inside a ring main unit based on artificial intelligence.
In a first aspect, the application provides an artificial intelligence-based real-time detection method for condensation inside a ring main unit, which comprises the following steps:
determining a first confidence coefficient that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence coefficient that a similar window in the search window comprises the heat source pixel point; the window to be denoised is the same as the central pixel point of the search window, and the central pixel point of the similar window is the pixel point in the search window;
determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window;
determining the gray weight of the central pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, so as to determine the gray weight of each pixel point in the image to be detected;
denoising the image to be detected based on the gray weight of each pixel point in the image to be detected.
In an alternative embodiment, determining a first confidence that a window to be denoised in each search window in the image to be detected includes a heat source pixel point and a second confidence that a similar window in the search window includes a heat source pixel point includes:
determining a hot zone position influence factor based on a first Euclidean distance between the window to be denoised and a reference heat source pixel point and a second Euclidean distance between the similar window and the reference heat source pixel point; the reference heat source pixel point is a central pixel point of a reference window with the maximum average gray value;
determining a first heat zone body influence factor corresponding to a window to be denoised, and determining a second heat zone body influence factor corresponding to a similar window, wherein the first heat zone body influence factor represents whether the window to be denoised is a real heat source area, and the second heat zone body influence factor represents whether the similar window is a real heat source area;
determining a first confidence that the window to be denoised includes a heat source pixel point based on the hot zone position impact factor and the first hot zone body impact factor, and determining a second confidence that the similar window includes a heat source pixel point based on the hot zone position impact factor and the second hot zone body impact factor.
In an alternative embodiment, determining a first hot zone body influence factor corresponding to the window to be denoised includes:
determining a first heat radiation range and a first heat source fluctuation index corresponding to the window to be denoised;
the first hot zone body impact factor is determined based on the first heat radiation range and the first heat source fluctuation index.
In an optional embodiment, determining the first heat radiation range corresponding to the window to be denoised includes:
determining a high Wen Xiangsu point and a low-temperature pixel point in the image to be detected, and determining a heat source pixel point based on the Gao Wenxiang pixel point and the low-temperature pixel point;
determining a first heat radiation range of the Gao Wenxiang pixel in the window to be denoised based on the distance between the Gao Wenxiang pixel in the window to be denoised and the low-temperature pixel corresponding to the Gao Wenxiang pixel and the distance between the Gao Wenxiang pixel and the heat source pixel corresponding to the Gao Wenxiang pixel; the low-temperature pixel point corresponding to the Gao Wenxiang pixel point is the low-temperature pixel point closest to the high-temperature pixel point in each direction of the high-temperature pixel point;
determining a first heat source fluctuation index corresponding to the window to be denoised, including:
determining a temperature gradient value and a temperature gradient mean value of the heat source pixel points in all directions in the window to be denoised, and determining a first heat source fluctuation index corresponding to the window to be denoised based on the temperature gradient value and the temperature gradient mean value; wherein, when calculating the temperature gradient value, if a low-temperature pixel point is encountered, the temperature gradient calculation is stopped.
In an alternative embodiment, determining the high Wen Xiangsu point and the low temperature pixel point in the image to be detected includes:
determining a gray threshold value based on the maximum gray value and the minimum gray value in the image to be detected, taking a pixel point with the gray value larger than the gray threshold value as a high Wen Xiangsu point, and taking a pixel point with the gray value smaller than the gray threshold value as a low-temperature pixel point;
determining a heat source pixel based on the Gao Wenxiang pixel and the low temperature pixel includes:
determining a low-temperature pixel point corresponding to the Gao Wenxiang pixel point, and taking the opposite direction from the Gao Wenxiang pixel point to the direction of the corresponding low-temperature pixel point as an inverse radiation direction; and calculating the heat gradient value in the reverse radiation direction, wherein if the heat gradient value is the smallest pixel point, the pixel point is a heat source pixel point, and if the heat gradient value is not the smallest point, the Gao Wenxiang pixel point is a heat source pixel point.
In an alternative embodiment, determining the second hot zone body influence factor corresponding to the similar window includes:
determining a second heat radiation range and a second heat source fluctuation index corresponding to the similar window;
the second hot zone body impact factor is determined based on the second heat radiation range and the second heat source fluctuation index.
In an alternative embodiment, determining the second heat radiation range corresponding to the similar window includes:
and determining a second heat radiation range of the Gao Wenxiang pixel point in the similar window based on the distance between the Gao Wenxiang pixel point and the Gao Wenxiang pixel point corresponding low-temperature pixel point and the distance between the Gao Wenxiang pixel point and the Gao Wenxiang pixel point corresponding heat source pixel point in the similar window.
In an alternative embodiment, determining a second heat source fluctuation index corresponding to the similar window includes:
determining a temperature gradient value and a temperature gradient mean value of the heat source pixel points in all directions in the similar window, and determining a second heat source fluctuation index corresponding to the similar window based on the temperature gradient value and the temperature gradient mean value; wherein, when calculating the temperature gradient value, if a low-temperature pixel point is encountered, the temperature gradient calculation is stopped.
In an alternative embodiment, determining the heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window includes:
calculating a first average gray value of a window to be denoised and a second average gray value of a similar window;
and calculating to obtain a heat source temperature difference coefficient based on the first average gray value, the second average gray value and the average gray value of the reference window corresponding to the reference heat source pixel point.
In an alternative embodiment, determining the gray weight of the center pixel of the window to be denoised based on the first confidence, the second confidence and the heat source temperature difference coefficient includes:
and determining the gray weight of the central pixel point of the window to be denoised by using the following formula:
wherein,is the central pixel point of the window to be denoised in the NLM algorithmIs the normalized parameter in NLM algorithm, Z (x)>For the first confidence level, ++>For the second confidence level, HER is the heat source temperature difference coefficient, h is the smoothing parameter, ++>Is an exponential function of natural constant.
The application has the beneficial effects that the method is different from the prior art, and the method for detecting the condensation in the ring main unit based on artificial intelligence comprises the following steps: determining a first confidence coefficient that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence coefficient that a similar window in the search window comprises the heat source pixel point; the window to be denoised is the same as the central pixel point of the search window, and the central pixel point of the similar window is the pixel point in the search window; determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window; determining the gray weight of the central pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, so as to determine the gray weight of each pixel point in the image to be detected; denoising the image to be detected based on the gray weight of each pixel point in the image to be detected. According to the method, the thermal imaging image is improved by denoising the thermal imaging image, and the accurate condition inside the ring main unit is detected in real time through the image.
Drawings
FIG. 1 is a flow chart of an embodiment of a real-time detection method of condensation inside a ring main unit based on artificial intelligence;
FIG. 2 is a flowchart illustrating an embodiment of the step S11 in FIG. 1;
fig. 3 is a schematic diagram of searching a heat source pixel point in the real-time detection method of the condensation inside the ring main unit based on artificial intelligence.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The present application will be described in detail with reference to the accompanying drawings and examples.
Referring to fig. 1, a flow chart of an embodiment of a real-time detection method for condensation inside a ring main unit based on artificial intelligence according to the present application specifically includes:
step S11: and determining a first confidence that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence that a similar window in the search window comprises the heat source pixel point.
And arranging a small temperature measurement thermal imager on the upper part of the inner wall, close to the cabinet door, in the ring main unit, performing thermal imaging shooting on the condition inside the ring main unit once every preset time, for example, 15 minutes, and only acquiring original data shot by thermal imaging, namely, a thermal imaging gray image inside the ring main unit, wherein a region with a larger gray value is a region with a higher temperature, and a region with a smaller gray value is a region with a lower temperature. And the image is sharpened by using the Laplacian operator, so that the region with larger temperature difference in the image is more obvious.
The main source of heat in the looped netowrk cabinet is the heat that produces when the power equipment in the cabinet operates, mainly distributes in the inner wall around the power equipment and the position that power equipment is located, and partial heat comes from the ambient temperature that the looped netowrk cabinet is located like sunlight directly, and it mainly distributes in cabinet door and inner wall department. The power equipment in the ring main unit usually works with a relatively balanced power, the generated heat is mainly concentrated around the equipment, and the heat of the external environment can cause certain change of the temperature of the inner wall of the ring main unit due to external time, weather and other factors. When the temperature of the inner wall of the ring main unit is reduced due to the reduction of the external environment temperature, condensation can be formed on the inner wall of the ring main unit when the vapor with higher internal temperature contacts with the inner wall with lower temperature. When the accumulated dust in the condensation becomes more, the condensation can drop on the power equipment, so that the power equipment is damaged. Most of the current detection methods for the condensation in the ring main unit are to monitor the temperature and the humidity in the ring main unit, calculate and look up the obtained data to roughly judge the generation condition of the condensation, and finally determine whether the condensation is generated in the ring main unit or not, and the condensation needs to be surveyed by workers in the field, so that the labor time and the cost are wasted and the detection method is not visual.
According to the application, the interior of the ring main unit is intuitively monitored in real time by adopting the small thermal imager, so that whether condensation is formed in the ring main unit can be rapidly judged. However, the operation of the power equipment not only generates heat, but also generates thermal drift noise on a thermal imaging image shot by the thermal imaging instrument, so that the final imaging quality is poor, and a worker is difficult to better judge the situation in the cabinet. So that in order to improve the quality of the image, it is necessary to perform denoising processing on the image. In the application, an NLM non-local mean filtering method is mainly adopted to carry out denoising treatment on the image. The main reason for the thermal drift is that the external environment temperature of the power equipment changes during operation, so that the heat generated by the power equipment deviates in the image, and a high Wen Bankuai appears in a place which is originally in a low-temperature area, and the plaque is also called a low-temperature contrast plaque, thereby affecting the judgment of the internal condition of the ring main unit.
Aiming at the characteristics, the acquired thermal imaging image is recorded as an image to be detected, a 55 x 55 search window is constructed by taking each pixel point in the image to be detected as a central pixel point, a window to be denoised and a similar window are constructed in the search window, specifically, the window to be denoised is the same as the central pixel point of the search window, and the central pixel point of the similar window is the pixel point in the search window. In a specific embodiment, a window to be denoised of 5*5 is built in the search window by taking the central pixel point of the search window as the centerFurther constructing 5*5 similar window by taking each pixel point in the search window as a central pixel point. And determining a first confidence that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point, and determining a second confidence that a similar window in the search window comprises the heat source pixel point.
In one embodiment, referring to fig. 2, step S11 includes:
step S21: and determining a hot zone position influence factor based on a first Euclidean distance between the window to be denoised and a reference heat source pixel point and a second Euclidean distance between the similar window and the reference heat source pixel point.
It should be noted that, the reference heat source pixel point is the position of the heat source in the image to be detected, the heat generated by the power equipment is generally uniform under the condition of normal operation, but because a plurality of power equipment is generally arranged in the ring main unit, the internal space is limited, the heat dissipation space is smaller, the heat is generally accumulated in a certain area to form a local heat dissipation point, so that a 5*5 reference window is constructed by taking each pixel point of the image to be detected as the center, the average gray value T of each reference window is calculated, the average gray value calculated by counting all the pixel points to form the reference window is calculated, and the maximum value in the average gray values is found outCorresponding reference window, the central pixel point of the reference window, record the coordinate of the central pixel point +.>The pixel point is the reference heat source pixel point in the current ring main unit.
The phenomenon of thermal drift is usually manifested as a high Wen Bankuai in the original low temperature region, i.e. the low temperature contrast plaque is in an abnormal position. If a certain height Wen Bankuai occurs in a region farther from the heat source, it is possible that the plaque is a low temperature contrast plaque, and the closer the distance is, it is possible that the plaque is a normal high heat region. For the window to be denoised in the search windowAnd similar window->Firstly, recording the coordinates of the central pixel points of the two pixels as +.>And->Respectively calculating Euclidean distance between the pixel points and the reference heat source pixel point>And->Wherein->For window to be denoised->Center pixel of +.>Pixel point of reference heat source>First Euclidean distance between +_>Is a similar window->Center pixel of +.>Pixel point of reference heat source>A second euclidean distance therebetween. Based on the first Euclidean distance->And a second Euclidean distance->A hot zone position influence factor TER is determined. Specifically, the calculation mode of the hot zone position influence factor TER is as follows:
when the hot zone position influence factor TER is smaller, the window to be denoised is describedAnd similar Window->The more similar the window to be denoised +>And similar Window->The areas belong to the same temperature area, thermal drift noise possibly exists, and the larger the thermal area position influence factor TER is, the window to be denoised is indicated>And similar Window->The larger the difference in the regions, the less likely thermal drift noise is present.
In order to be able to more accurately confirm whether thermal drift noise is present in this region, further analysis of the hot zone body impact characteristics is performed below.
Step S22: and determining a first heat zone body influence factor corresponding to the window to be denoised, and determining a second heat zone body influence factor corresponding to the similar window.
Firstly, a calculation mode of a first hot zone body influence factor corresponding to a window to be denoised is explained. Specifically, a first heat radiation range and a first heat source fluctuation index corresponding to the window to be denoised are determined.
In an embodiment, high Wen Xiangsu and low temperature pixel points in the image to be detected are determined. Specifically, a gray threshold is determined based on a maximum gray value and a minimum gray value in the image to be detected, a pixel point with a gray value larger than the gray threshold is taken as a high Wen Xiangsu point, and a pixel point with a gray value smaller than the gray threshold is taken as a low-temperature pixel point.
In one embodiment of the application, the thermal region typically has a range of thermal emissions that falls within that region, whereas for low temperature contrast regions, the range of thermal emissions is smaller because it is not a real heat source. Aiming at the characteristics of the radiation range of the heat source, counting the pixel gray values of the image to be detected, searching the minimum gray value and recording the minimum gray value asSearching the maximum gray value to record asThe gray threshold P is:
according to the gray threshold P, pixels with gray values greater than P are classified as high Wen Xiangsu points, and pixels with gray values less than P are classified as low temperature pixels.
Further, a heat source pixel is determined based on the high temperature pixel and the low temperature pixel. Specifically, determining a low-temperature pixel point corresponding to the Gao Wenxiang pixel point, and taking the direction opposite to the direction from the Gao Wenxiang pixel point to the corresponding low-temperature pixel point as an inverse radiation direction; and calculating the heat gradient value in the reverse radiation direction, wherein if the heat gradient value is the smallest pixel point, the pixel point is a heat source pixel point, and if the heat gradient value is not the smallest point, the Gao Wenxiang pixel point is a heat source pixel point.
Specifically, please refer to fig. 3, for a certain high Wen Xiangsu point, the high Wen Xiangsu point (referred to as high Wen Xiangsu point in fig. 3) is taken as the center, and the search is performed in eight directionsA nearest low-temperature pixel point (called a low-temperature point in fig. 3) and obtaining the distance between the high Wen Xiangsu point and the corresponding second low-temperature pixel pointAnd determining the direction of the high Wen Xiangsu point to the low temperature point as a forward radiation direction, and taking the 180-degree turnover direction of the direction as a reverse radiation direction. If the high Wen Xiangsu point is not the heat source pixel point of the high temperature region, the high Wen Xiangsu point has the heat gradient change minimum pixel point in the reverse radiation direction, namely the heat source pixel point, and if the high Wen Xiangsu point is the heat source pixel point of the region, the heat gradient change minimum pixel point in the reverse radiation direction does not exist. Specifically, the thermal gradient value +.>And finding out the point with the smallest heat gradient, wherein the point is the heat source pixel point in the heat region, and if the point cannot be found, the point with the height Wen Xiangsu is the heat source pixel point. Let A1 be the gray value of a pixel point in the reverse radiation direction, and the gray values of the front and rear pixels in the reverse radiation direction be A2 and A0, then the heat gradient value +.>Because the heat of the heat source pixel point is diffused from the heat source pixel point to the surrounding, the heat of the points in the neighborhood of the heat source pixel point is similar, and according to the calculation method, the heat gradient value of the heat source pixel point is minimum, so that only one heat source pixel point meeting the condition can be found. If the point with the smallest heat gradient is found, the point is the heat source pixel point in the heat region, and if the point cannot be found, the point with the height Wen Xiangsu is the heat source pixel point.
And determining a first heat radiation range FER1 of the Gao Wenxiang pixel in the window to be denoised based on the distance between the Gao Wenxiang pixel and the low-temperature pixel corresponding to the Gao Wenxiang pixel in the window to be denoised and the distance between the Gao Wenxiang pixel and the heat source pixel corresponding to the Gao Wenxiang pixel. Specifically, the first heat radiation range FER1 is calculated by:
wherein FER1 is the first heat radiation range, pi is the circumference ratio,distance ∈ Wen Xiangsu for high and low-temperature pixel closest to the high Wen Xiangsu (i.e., low-temperature pixel corresponding to the high Wen Xiangsu) by eight directions>For the distance between a certain high Wen Xiangsu point and the heat source pixel point with the smallest heat gradient found by the reverse radiation direction with the high Wen Xiangsu point as the center, if the heat source pixel point with the smallest heat gradient is not found, the high Wen Xiangsu point is the heat source pixel point, at this time ∈ ->. The larger the first heat radiation range, the more likely the point is a real heat source, and the smaller the first heat radiation range, the more likely is a false heat source due to thermal drift.
In order to further confirm the authenticity of the heat source, judging heat fluctuation in a high-temperature area, and calculating a first heat source fluctuation index corresponding to the window to be denoised. For a certain high-temperature region, the smaller the temperature gradient is, the more gradual the temperature change is, the lower the heat source diffusivity is, and the high-temperature region may not be a real heat source high-temperature region; when the temperature gradient is larger, the temperature change is more severe, the heat diffusivity is larger, and the heat source is probably a real heat source high-temperature region. Based on the characteristics, the heat source pixel points determined in the above are utilized to determine the temperature gradient value and the temperature gradient mean value of the heat source pixel points in all directions, such as the eight directions of the heat source pixel points, in the window to be denoised, and the temperature gradient value is recorded asThe average value of the temperature gradient is recorded asThe calculation method of the temperature gradient value is the same as the calculation method of the heat gradient value, and is not described herein. In calculating the temperature gradient value, if a low-temperature pixel point is encountered, the temperature gradient calculation is stopped.
And determining a first heat source fluctuation index HTR1 corresponding to the window to be denoised based on the temperature gradient value and the temperature gradient mean value. Specifically, the first heat source fluctuation index HTR1 is calculated by:
wherein HTR1 is a first heat source fluctuation index,for the ith temperature gradient value, +.>And n is the number of total temperature gradient values. When the first heat source fluctuation index is larger, the diffusivity of the high-temperature region represented by the window to be denoised is larger, the temperature change is obvious, the high-temperature region is more likely to be a real heat source region, and when the first heat source fluctuation index is smaller, the diffusivity of the high-temperature region represented by the window to be denoised is smaller, the temperature change is gentle, and the high-temperature region is more likely to be a false heat source region.
The first heat zone body influence factor HB1 is determined based on the first heat radiation range FER1 and the first heat source fluctuation index HTR1. Specifically, the first hot zone body influence factor HB1 is calculated by:
the first heat zone body influence factor represents whether the window to be denoised is a real heat source area or not, specifically, when the first heat zone body influence factor is larger, the more likely that the area represented by the window to be denoised is a real heat source area, and when the first heat zone body influence factor is smaller, the more likely that the area represented by the window to be denoised is a false heat source area, and the more likely that the area represented by the window to be denoised is a thermal drift noise area.
Further, a second heat zone body influence factor corresponding to the similar window is calculated, and the second heat zone body influence factor characterizes whether the similar window is a real heat source area or not. The second hot zone body influence factor is calculated in the same manner as the first hot zone body influence factor. It will be appreciated that the larger the second hot zone body impact factor, the more likely the region characterized by the similar window is a real heat source region, and that the smaller the second hot zone body impact factor, the more likely the region characterized by the similar window is a false heat source region, which is a thermal drift noise region.
When a second heat zone body influence factor corresponding to a similar window is calculated, determining a second heat radiation range and a second heat source fluctuation index corresponding to the similar window; the second hot zone body impact factor is determined based on the second heat radiation range and the second heat source fluctuation index.
Wherein determining a second heat radiation range corresponding to the similar window includes: and determining a second heat radiation range of the Gao Wenxiang pixel point in the similar window based on the distance between the Gao Wenxiang pixel point and the Gao Wenxiang pixel point corresponding low-temperature pixel point and the distance between the Gao Wenxiang pixel point and the Gao Wenxiang pixel point corresponding heat source pixel point in the similar window. The calculation method of the second heat radiation range is the same as that of the first heat radiation range, and specific reference is made to the calculation process of the first heat radiation range, which is not described herein.
Further, determining a second heat source fluctuation index corresponding to the similar window includes: and determining the temperature gradient value and the temperature gradient mean value of the heat source pixel points in all directions in the similar window, and determining a second heat source fluctuation index corresponding to the similar window based on the temperature gradient value and the temperature gradient mean value. The calculation method of the second heat source fluctuation index is the same as the calculation method of the first heat source fluctuation index, and the detailed description thereof is omitted herein.
Step S23: determining a first confidence that the window to be denoised includes a heat source pixel point based on the hot zone position impact factor and the first hot zone body impact factor, and determining a second confidence that the similar window includes a heat source pixel point based on the hot zone position impact factor and the second hot zone body impact factor.
Specifically, the window to be denoised includes a first confidence level of the heat source pixel pointThe calculation mode of (a) is as follows:
wherein the method comprises the steps ofFor the first confidence that the window to be denoised includes the heat source pixel point, HB1 is the first hot zone body influence factor, and TER is the hot zone position influence factor. When the first confidence coefficient of the heat source pixel point is larger, the probability that the region corresponding to the window to be denoised is a real heat source is larger, and the probability that thermal drift noise exists is smaller; when the first confidence coefficient of the heat source pixel point included in the window to be denoised is smaller, the probability that the region corresponding to the window to be denoised is a real heat source is smaller, and the probability that thermal drift noise exists is larger.
Further, the similar window includes a second confidence level for the heat source pixel pointThe calculation mode of (a) is as follows:
wherein the method comprises the steps ofFor a second confidence that the similar window includes heat source pixels, HB2 is a second hot zone ontology influencing factor, and TER is a hot zone location influencing factor. When the second confidence coefficient of the heat source pixel points included in the similar window is larger, the possibility that the corresponding region of the similar window is a real heat source is higher, and the possibility that thermal drift noise exists is lower; when the second confidence coefficient of the pixel point of the heat source included in the similar window is smaller, the probability that the region corresponding to the similar window is a real heat source is smaller, and the probability that thermal drift noise exists is larger.
Step S12: and determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window.
Since there are many monitoring indicator lamps in the power equipment, and part of the indicator lamps are located far away from the main heating power equipment, and the indicator lamps are also power equipment, there are also heating situations, but the heating amount of the indicator lamps is far less than that of the power equipment relative to the main ring main unit power equipment, and the radiation range of the heat source is also small, and when the indicator lamps exist in the low-temperature area, the indicator lamps in the thermal imaging image appear as high Wen Fancha patches in the low-temperature area, but the indicator lamp area does not belong to noise generated by thermal drift, so that less weight is required to prevent the indicator lamps from being treated as noise erroneously.
For the above case, for the window to be denoisedAnd similar window->The pixel gray values in the window to be denoised are averaged to calculate a first average gray value +.>And a second average gray value of the similar window +.>For Wen Fancha plaque, the average gray value and heat within its windowThe difference in gray value of the source is small, while the difference in average gray value of the indicator light region and the heat source is large, and the difference in heat of the indicator light region to the low-temperature region is also relatively large. And calculating to obtain a heat source temperature difference coefficient based on the first average gray value, the second average gray value and the average gray value of the reference window corresponding to the reference heat source pixel point. Specifically, the heat source temperature difference coefficient HER is calculated by:
wherein HER is the heat source temperature difference coefficient,for the average gray value of the reference window corresponding to the reference heat source pixel point,is the second average gray value of the similar window, < >>Is the first average gray value of the window to be denoised. When the heat source temperature difference coefficient is larger, the region corresponding to the similar window is more likely to be an indicator light region, a smaller weight should be given to avoid erroneous denoising, and when the heat source temperature difference coefficient is smaller, the region is more likely to be a common inner wall region, and denoising is normally performed.
Step S13: and determining the gray weight of the central pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, so as to determine the gray weight of each pixel point in the image to be detected.
In a specific embodiment, the gray weight of the center pixel of the window to be denoised is determined by using the following formula:
wherein,for the gray weight of the central pixel point of the window to be denoised in the NLM algorithm, Z (x) is the normalized parameter in the NLM algorithm, < ->For the first confidence level, ++>For the second confidence level, HER is the heat source temperature difference coefficient, h is the smoothing parameter, ++>Is an exponential function of natural constant.
When a high-temperature block appears in the low-temperature region, if the confidence coefficient of the heat source of the high-temperature block is larger and the temperature difference between the high-temperature block and the heat source is smaller, the search window is possibly positioned at the edges of the high-temperature region and the low-temperature region, and the denoising treatment is normally carried out; if the confidence coefficient of the heat source of the high-temperature block is smaller and the temperature difference between the heat source and the high-temperature block is smaller, thermal drift noise possibly occurs, and the weight should be increased so as to achieve the aim of better noise control; if the confidence coefficient of the heat source of the high-temperature block is larger and the temperature difference between the high-temperature block and the heat source is larger, the high-temperature block possibly belongs to the indicator light area, and is not thermal drift noise, and denoising treatment is normally carried out.
Step S14: denoising the image to be detected based on the gray weight of each pixel point in the image to be detected.
Through the method, the gray weight of each pixel point in the image to be detected in the NLM algorithm is determined, and denoising is carried out on the image to be detected based on the gray weight of each pixel point in the image to be detected. The method finally improves the quality of the shot thermal imaging image, utilizes the morphological characteristics of artificial intelligence based on condensation as characteristic values, detects and positions in the image, and finally divides the position where the condensation exists and gives early warning in real time. The processing efficiency of condensation phenomenon generated on the ring main unit is improved.
In the existing NLM algorithm, the gray weight is determined by the similarity difference characterized by the distance between the window to be denoised and the similar window. According to the application, the thermal imaging camera is arranged in the ring main unit, the real-time condition inside the ring main unit is shot, the gray weight in the NLM algorithm is determined by utilizing the noise characteristic in the ring main unit, the similarity criterion in the NLM algorithm is improved, the thermal imaging image is efficiently denoised, the thermal imaging image quality is improved, and the staff can more intuitively and clearly detect the accurate condition inside the ring main unit in real time.
The foregoing is only the embodiments of the present application, and therefore, the scope of the present application is not limited by the above embodiments, and all equivalent structures or equivalent processes using the descriptions and the drawings of the present application or direct or indirect application in other related technical fields are included in the scope of the present application.
Claims (1)
1. The real-time detection method for the condensation inside the ring main unit based on artificial intelligence is characterized by comprising the following steps of:
determining a first confidence coefficient that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence coefficient that a similar window in the search window comprises the heat source pixel point;
determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window;
determining the gray weight of the central pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, so as to determine the gray weight of each pixel point in the image to be detected;
denoising the image to be detected based on the gray weight of each pixel point in the image to be detected;
the acquired thermal imaging image is recorded as an image to be detected, a 55 x 55 search window is constructed by taking each pixel point in the image to be detected as a central pixel point, a window to be denoised and a similar window are constructed in the search window, the window to be denoised is the same as the central pixel point of the search window, and the central pixel point of the similar window is the pixel point in the search window;
determining a first confidence that a window to be denoised in each search window in the image to be detected comprises a heat source pixel point and a second confidence that a similar window in the search window comprises a heat source pixel point comprises the following steps:
determining a hot zone position influence factor based on a first Euclidean distance between the window to be denoised and a reference heat source pixel point and a second Euclidean distance between the similar window and the reference heat source pixel point; the reference heat source pixel point is a central pixel point of a reference window with the maximum average gray value;
determining a first heat zone body influence factor corresponding to a window to be denoised, and determining a second heat zone body influence factor corresponding to a similar window, wherein the first heat zone body influence factor represents whether the window to be denoised is a real heat source area, and the second heat zone body influence factor represents whether the similar window is a real heat source area;
determining a first confidence that the window to be denoised comprises a heat source pixel point based on the hot zone position impact factor and the first hot zone body impact factor, and determining a second confidence that the similar window comprises a heat source pixel point based on the hot zone position impact factor and the second hot zone body impact factor;
constructing 5*5 reference windows with each pixel point of the image to be detected as the center, calculating the average gray value T of each reference window, counting the average gray values calculated by all the pixel points forming the reference window, and finding the maximum value in the average gray valuesCorresponding reference window, the central pixel point of the reference window records coordinates +.>The pixel point is a reference heat source pixel point;
for the window to be denoised in the search windowAnd similar window->Recording coordinates of central pixel points of the twoIs thatAnd->Respectively calculating Euclidean distance between the pixel points and the reference heat source pixel point>And->Wherein->For window to be denoised->Center pixel of +.>Pixel point of reference heat source>First Euclidean distance between +_>Is a similar window->Center pixel of +.>Pixel point of reference heat source>A second Euclidean distance therebetween;
determining a first hot zone body influence factor corresponding to a window to be denoised, including:
determining a first heat radiation range and a first heat source fluctuation index corresponding to the window to be denoised;
determining the first hot zone body impact factor based on the first heat radiation range and the first heat source fluctuation index;
determining a first heat radiation range corresponding to the window to be denoised, including:
determining a high Wen Xiangsu point and a low-temperature pixel point in the image to be detected, and determining a heat source pixel point based on the Gao Wenxiang pixel point and the low-temperature pixel point;
determining a first heat radiation range of the Gao Wenxiang pixel in the window to be denoised based on the distance between the Gao Wenxiang pixel in the window to be denoised and the low-temperature pixel corresponding to the Gao Wenxiang pixel and the distance between the Gao Wenxiang pixel and the heat source pixel corresponding to the Gao Wenxiang pixel; the low-temperature pixel point corresponding to the Gao Wenxiang pixel point is the low-temperature pixel point closest to the high-temperature pixel point in each direction of the high-temperature pixel point;
determining a first heat source fluctuation index corresponding to the window to be denoised, including:
determining a temperature gradient value and a temperature gradient mean value of the heat source pixel points in all directions in the window to be denoised, and determining a first heat source fluctuation index corresponding to the window to be denoised based on the temperature gradient value and the temperature gradient mean value; when the temperature gradient value is calculated, if a low-temperature pixel point is encountered, the temperature gradient calculation is stopped;
for any high-temperature pixel point, taking any high Wen Xiangsu point as the center, finding the nearest low-temperature pixel point in eight directions, and obtaining the distance between any high Wen Xiangsu point and the corresponding second low-temperature pixel pointDetermining the direction of any high-temperature pixel point to the low-temperature point as a forward radiation direction, and taking the 180-degree turnover direction along the radiation direction as a reverse radiation direction; if any high-temperature pixel point is not the heat source pixel point of the high-temperature region, the pixel point with the minimum heat gradient change exists at any high Wen Xiangsu point in the reverse radiation directionAs a heat source pixel; if any high-temperature pixel point is a heat source pixel point in a high-temperature area, a minimum pixel point with heat gradient change does not exist in the reverse radiation direction; calculating the heat gradient value in the reverse radiation directionFinding out the pixel point with the minimum heat gradient as a heat source pixel point in the heat area, and if the pixel point cannot be found out, taking any point Wen Xiangsu as the heat source pixel point; let a pixel gray value in the reverse radiation direction be A1, and the gray values of the front and rear pixels in the reverse radiation direction be A2 and A0, then the heat gradient value of the pixel with gray value of A1；
Determining a first heat source fluctuation index HTR1 corresponding to the window to be denoised based on the temperature gradient value and the temperature gradient mean value; the first heat source fluctuation index HTR1 is calculated by:
wherein HTR1 is a first heat source fluctuation index,for the ith temperature gradient value, +.>N is the number of total temperature gradient values;
determining a high Wen Xiangsu point and a low-temperature pixel point in the image to be detected comprises:
determining a gray threshold value based on the maximum gray value and the minimum gray value in the image to be detected, taking a pixel point with the gray value larger than the gray threshold value as a high Wen Xiangsu point, and taking a pixel point with the gray value smaller than the gray threshold value as a low-temperature pixel point;
determining a heat source pixel based on the Gao Wenxiang pixel and the low temperature pixel includes:
determining a low-temperature pixel point corresponding to the Gao Wenxiang pixel point, and taking the opposite direction from the Gao Wenxiang pixel point to the direction of the corresponding low-temperature pixel point as an inverse radiation direction; calculating the heat gradient value in the reverse radiation direction, wherein if the pixel point with the minimum heat gradient is provided, the pixel point is a heat source pixel point, and if the pixel point with the minimum heat gradient is not provided, the Gao Wenxiang pixel point is a heat source pixel point;
determining a second hot zone body influence factor corresponding to the similar window, including:
determining a second heat radiation range and a second heat source fluctuation index corresponding to the similar window;
determining the second hot zone body impact factor based on the second heat radiation range and the second heat source fluctuation index;
determining a second heat radiation range corresponding to the similar window comprises:
determining a second heat radiation range of the Gao Wenxiang pixel in the similar window based on the distance between the Gao Wenxiang pixel and the Gao Wenxiang pixel and the distance between the Gao Wenxiang pixel and the Gao Wenxiang pixel;
determining a second heat source fluctuation index corresponding to the similar window, including:
determining a temperature gradient value and a temperature gradient mean value of the heat source pixel points in all directions in the similar window, and determining a second heat source fluctuation index corresponding to the similar window based on the temperature gradient value and the temperature gradient mean value; when the temperature gradient value is calculated, if a low-temperature pixel point is encountered, the temperature gradient calculation is stopped;
determining a heat source temperature difference coefficient based on the pixel gray scale characteristics in the window to be denoised and the similar window comprises the following steps:
calculating a first average gray value of a window to be denoised and a second average gray value of a similar window;
calculating to obtain a heat source temperature difference coefficient based on the first average gray value, the second average gray value and the average gray value of a reference window corresponding to the reference heat source pixel point;
for window to be denoisedAnd similar window->The pixel gray values in the window to be denoised are averaged to calculate a first average gray value +.>And a second average gray value of the similar window +.>；
The heat source temperature difference coefficient HER is calculated by the following steps:
wherein HER is the heat source temperature difference coefficient,for the average gray value of the reference window corresponding to the reference heat source pixel point,/for>Is the second average gray value of the similar window, < >>The first average gray value of the window to be denoised;
determining the gray scale weight of the center pixel point of the window to be denoised based on the first confidence coefficient, the second confidence coefficient and the heat source temperature difference coefficient, wherein the gray scale weight comprises the following steps:
and determining the gray weight of the central pixel point of the window to be denoised by using the following formula:
wherein,is the gray weight of the central pixel point of the window to be denoised in NLM algorithm, wherein +.>Representing window to be denoised->Is>Representing a similar Window->Is defined by a center pixel point of the display panel; z (x) represents the central pixel point of the window to be denoised in NLM algorithm +.>Is>For the first confidence level, ++>For the second confidence level, HER is the heat source temperature difference coefficient, h is the smoothing parameter, ++>Is an exponential function of natural constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311133033.9A CN116862918B (en) | 2023-09-05 | 2023-09-05 | Real-time detection method for condensation inside ring main unit based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311133033.9A CN116862918B (en) | 2023-09-05 | 2023-09-05 | Real-time detection method for condensation inside ring main unit based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116862918A CN116862918A (en) | 2023-10-10 |
| CN116862918B true CN116862918B (en) | 2023-11-24 |
Family
ID=88225360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311133033.9A Active CN116862918B (en) | 2023-09-05 | 2023-09-05 | Real-time detection method for condensation inside ring main unit based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116862918B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111578999A (en) * | 2020-05-25 | 2020-08-25 | 东北电力大学 | Integrated information monitoring system of ring main unit |
| CN113643201A (en) * | 2021-07-27 | 2021-11-12 | 西安理工大学 | Image denoising method of self-adaptive non-local mean value |
| CN115358951A (en) * | 2022-10-19 | 2022-11-18 | 广东电网有限责任公司佛山供电局 | Intelligent ring main unit monitoring system based on image recognition |
| CN115830106A (en) * | 2023-02-16 | 2023-03-21 | 智联信通科技股份有限公司 | Auxiliary positioning method for electrified cleaning of machine room equipment |
| CN115841624A (en) * | 2023-02-23 | 2023-03-24 | 山东洲蓝环保科技有限公司 | Blast furnace gas flow distribution identification method based on infrared image |
| CN116012265A (en) * | 2023-03-28 | 2023-04-25 | 国科天成科技股份有限公司 | Infrared video denoising method and device based on time-space domain adaptive filtering |
| CN116363133A (en) * | 2023-06-01 | 2023-06-30 | 无锡斯达新能源科技股份有限公司 | Illuminator accessory defect detection method based on machine vision |
| CN116433035A (en) * | 2023-06-13 | 2023-07-14 | 中科数创(临沂)数字科技有限公司 | Building electrical fire risk assessment prediction method based on artificial intelligence |
| CN116433533A (en) * | 2023-06-09 | 2023-07-14 | 青岛宇通管业有限公司 | Water heating pipeline maintenance and detection method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013145799A1 (en) * | 2012-03-30 | 2013-10-03 | 富士フイルム株式会社 | Image processing device, photographing device, program, and image processing method |
| CN109804119B (en) * | 2016-12-30 | 2021-03-19 | 同济大学 | Asphalt pavement crack development degree detection method based on infrared thermography analysis |
-
2023
- 2023-09-05 CN CN202311133033.9A patent/CN116862918B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111578999A (en) * | 2020-05-25 | 2020-08-25 | 东北电力大学 | Integrated information monitoring system of ring main unit |
| CN113643201A (en) * | 2021-07-27 | 2021-11-12 | 西安理工大学 | Image denoising method of self-adaptive non-local mean value |
| CN115358951A (en) * | 2022-10-19 | 2022-11-18 | 广东电网有限责任公司佛山供电局 | Intelligent ring main unit monitoring system based on image recognition |
| CN115830106A (en) * | 2023-02-16 | 2023-03-21 | 智联信通科技股份有限公司 | Auxiliary positioning method for electrified cleaning of machine room equipment |
| CN115841624A (en) * | 2023-02-23 | 2023-03-24 | 山东洲蓝环保科技有限公司 | Blast furnace gas flow distribution identification method based on infrared image |
| CN116012265A (en) * | 2023-03-28 | 2023-04-25 | 国科天成科技股份有限公司 | Infrared video denoising method and device based on time-space domain adaptive filtering |
| CN116363133A (en) * | 2023-06-01 | 2023-06-30 | 无锡斯达新能源科技股份有限公司 | Illuminator accessory defect detection method based on machine vision |
| CN116433533A (en) * | 2023-06-09 | 2023-07-14 | 青岛宇通管业有限公司 | Water heating pipeline maintenance and detection method |
| CN116433035A (en) * | 2023-06-13 | 2023-07-14 | 中科数创(临沂)数字科技有限公司 | Building electrical fire risk assessment prediction method based on artificial intelligence |
Non-Patent Citations (4)
| Title |
|---|
| Research on a Bilateral Filtering Based Image Denoising Algorithm for Nightly Infrared Monitoring Images;Wang Zhong-ce等;《 PROCEEDINGS OF THE 3RD INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE AND SERVICE SYSTEM (CSSS)》;第109卷;全文 * |
| 基于FPGA的柜内电力设备广角红外图像处理与通信电路设计;许崇亮;《长江信息通信》(第2期);全文 * |
| 基于图像识别的高压电力电缆接头缺陷识别研究;牟久红;《中国优秀硕士学位论文全文数据库（工程科技Ⅱ辑）》(第10期);全文 * |
| 基于梯度特征的双核非局部均值去噪算法;张玉征;杨词慧;林泉;;计算机应用研究(第05期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116862918A (en) | 2023-10-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10914653B2 (en) | Infrared imaging systems and methods for oil leak detection | |
| CN109743879B (en) | Underground pipe gallery leakage detection method based on dynamic infrared thermography processing | |
| US20190340914A1 (en) | Infrared imaging systems and methods for gas leak detection | |
| US8175336B2 (en) | Target tracker | |
| CN104330164A (en) | Infrared focal plane array blind pixel detection method and device | |
| CN116721107B (en) | Intelligent monitoring system for cable production quality | |
| CN111667470B (en) | Industrial pipeline flaw detection inner wall detection method based on digital image | |
| CN117252868B (en) | Direct current screen defect detection method based on machine vision | |
| CN116843682B (en) | Boiler thermal efficiency on-line detection and analysis system using thermal infrared imager | |
| CN117274722A (en) | Intelligent detection method for distribution box based on infrared image | |
| CN113095114A (en) | Method, device and equipment for detecting oil leakage of hydraulic system | |
| CN116862918B (en) | Real-time detection method for condensation inside ring main unit based on artificial intelligence | |
| CN114463389A (en) | Moving target detection method and detection system | |
| CN115830106B (en) | Auxiliary positioning method for electrified cleaning of equipment in machine room | |
| CN112733646B (en) | Liquid medium leakage automatic detection method and system based on thermal imaging | |
| CN116883446B (en) | Real-time monitoring system for grinding degree of vehicle-mounted camera lens | |
| CN115619838B (en) | Missing risk monitoring system for medical waste refrigeration house | |
| CN113963013B (en) | Markless power transmission tower displacement vibration identification method based on computer vision | |
| CN112541403B (en) | Indoor personnel falling detection method by utilizing infrared camera | |
| CN111723808A (en) | Infrared weak target extraction method based on maximum value background prediction | |
| CN110309838B (en) | Object contour detection preprocessing method for video detection area based on exponential transformation | |
| CN112465850A (en) | Peripheral boundary modeling method, intelligent monitoring method and device | |
| JP6457727B2 (en) | Laminar smoke detection device and laminar smoke detection method | |
| CN116246222A (en) | Forest fire detection method based on extremum detection and self-adaptive segmentation algorithm | |
| CN117830032B (en) | Method and system for monitoring snapshot and risk assessment of power transmission line network |
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 |