CN111080691A - Infrared hot spot detection method and device for photovoltaic module - Google Patents
Infrared hot spot detection method and device for photovoltaic module Download PDFInfo
- Publication number
- CN111080691A CN111080691A CN201911300787.2A CN201911300787A CN111080691A CN 111080691 A CN111080691 A CN 111080691A CN 201911300787 A CN201911300787 A CN 201911300787A CN 111080691 A CN111080691 A CN 111080691A
- Authority
- CN
- China
- Prior art keywords
- hot spot
- area
- infrared
- photovoltaic
- image
- 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.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 50
- 238000004364 calculation method Methods 0.000 claims abstract description 48
- 238000005286 illumination Methods 0.000 claims abstract description 37
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 238000009432 framing Methods 0.000 claims abstract description 10
- 230000002159 abnormal effect Effects 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 32
- 230000011218 segmentation Effects 0.000 claims description 24
- 230000008569 process Effects 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 4
- 239000013598 vector Substances 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000000873 masking effect Effects 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 238000002372 labelling Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 17
- 238000012706 support-vector machine Methods 0.000 description 13
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010248 power generation Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010224 classification analysis Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 210000003608 fece Anatomy 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/64—Analysis of geometric attributes of convexity or concavity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02S—GENERATION OF ELECTRIC POWER BY CONVERSION OF INFRARED RADIATION, VISIBLE LIGHT OR ULTRAVIOLET LIGHT, e.g. USING PHOTOVOLTAIC [PV] MODULES
- H02S50/00—Monitoring or testing of PV systems, e.g. load balancing or fault identification
- H02S50/10—Testing of PV devices, e.g. of PV modules or single PV cells
- H02S50/15—Testing of PV devices, e.g. of PV modules or single PV cells using optical means, e.g. using electroluminescence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a photovoltaic module infrared hot spot detection method, which comprises the following steps: performing video framing processing on an infrared video corresponding to the photovoltaic module, and taking a frame image as an infrared image to be detected; segmenting the infrared image to be detected to obtain a photovoltaic region infrared image; converting the infrared image of the photovoltaic area into a gray image; carrying out binarization processing on the gray level images aiming at each gray level threshold value to obtain a first number of binarization images; determining abnormal areas of the photovoltaic module according to the first number of binary images, and extracting the outline by using an outline extraction function; and performing feature calculation on the extracted contour, judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, removing the hot spot caused by the junction box in the circular large hot spot, and removing the hot spot caused by illumination in the rectangular large hot spot. By applying the embodiment of the invention, the accuracy of hot spot analysis is improved and the analysis efficiency is improved.
Description
Technical Field
The invention relates to the technical field of solar power generation, in particular to a method and a device for detecting infrared hot spots of a photovoltaic module.
Background
The hot spot effect is that in the actual power generation process, the photovoltaic module is shielded by foreign matters such as bird droppings, fallen leaves and dust on the surface of the module for a long time to cause the photovoltaic module to generate a local shadow phenomenon, or the defects of the photovoltaic module cause a certain battery or a certain group of batteries in the photovoltaic module to be used as a load, the electric energy generated by other battery modules with illumination is consumed, and the output characteristic of the photovoltaic module is directly influenced. The hot spot effect can cause serious damage to the photovoltaic module battery plate, the service life of the photovoltaic module battery plate is shortened, and the power generation cost of a large photovoltaic power station is increased.
If the hot spots cannot be timely checked and effectively eliminated, the generated energy of a large photovoltaic power station is greatly reduced, economic loss is brought to the power station, and even the problems of battery plate burning, welding spot melting, glass cover plate cracking and the like are caused. Therefore, the detection of the hot spots is important for the operation and maintenance of the photovoltaic power station.
In recent years, with the increase of the number and scale of photovoltaic power stations, more and more photovoltaic enterprises gradually get rid of the traditional manual operation and maintenance method, and carry on the infrared camera with the help of the unmanned aerial vehicle to carry on the operation and maintenance detection to the photovoltaic power stations. But at present, the infrared video shot by the unmanned aerial vehicle can only be manually identified and checked, the automation degree is lower, and the analysis accuracy is low.
Disclosure of Invention
The embodiment of the invention aims to provide a photovoltaic module infrared hot spot detection method and device so as to improve the accuracy of hot spot analysis and improve the analysis efficiency. The specific technical scheme is as follows:
in order to achieve the above object, an embodiment of the present invention provides a method for detecting an infrared hot spot of a photovoltaic module, where the method includes:
performing video framing processing on an infrared video corresponding to the photovoltaic module, and taking a frame image as an infrared image to be detected;
dividing the upper boundary and the lower boundary of the photovoltaic array by using horizontal projection segmentation, dividing the left boundary and the right boundary by using vertical projection segmentation, combining the horizontal projection segmentation and the vertical projection segmentation, and segmenting the infrared image to be detected to segment the infrared image of the photovoltaic region;
converting the photovoltaic area infrared image into a gray level image;
setting a first number of gray level threshold values, and carrying out binarization processing on the gray level images aiming at each gray level threshold value to obtain a first number of binarization images;
determining an abnormal area of the photovoltaic module according to the first number of binary images, and extracting the outline by using an outline extraction function;
and performing feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness and main shaft direction;
and removing hot spots caused by the junction box in the round large hot spots and removing hot spots caused by illumination in the rectangular large hot spots.
In one implementation, the step of removing the hot spot caused by illumination in the large rectangular hot spot includes:
carrying out normalization processing on the image corresponding to the rectangular large hot spot;
calculating the horizontal gradient and the vertical gradient of the image in a differential mode;
dividing the window into cell units, calculating the gradient and direction of the cell units, and grouping and weighting the cell units;
combining adjacent cell units into pixel blocks, and counting pixel block histograms;
normalizing the gradient histogram and collecting HOG characteristics;
inputting the HOG characteristics into an SVM model to obtain an output result of the SVM model so as to determine whether the hot spot is caused by illumination;
the SVM model is characterized in that all samples are converted into sample pictures with the width and the height of 32 x 64 according to pre-collected solar illumination false detection and rectangular large hot spot samples in a large number of different types of photovoltaic power stations, the illumination and rectangular large hot spot sample pictures are respectively marked and provided with labels, HOG characteristic vectors of each sample are extracted, a calculation function is called to calculate a result array of each sample, and a sample matrix and a type mark of the sample matrix are stored; and meanwhile, SVM training parameters are set for SVM classification training to generate an xml file, the xml file is called in the hot spot detection process, HOG characteristic calculation is carried out on the detection component divided from the video frame to obtain a result array, and the result array is compared with an SVM training result to carry out recognition classification, so that rectangular large hot spots and illumination false detection can be accurately distinguished, and illumination interference similar to the rectangular large hot spots is removed.
In one implementation, the step of removing the hot spot caused by the junction box in the large circular hot spot includes:
according to the layout of the photovoltaic assembly junction box, a junction box detection function is designed by adopting an image masking method, and circular hot spots caused by the junction box are removed.
In one implementation, the circularity calculation is expressed by the following formula:
wherein E iscsIs the degree of circularity of the steel,area is the Area of the contour, perimeter is the perimeter of the contour; .
The formula used for the area calculation is expressed as:
wherein S is the area, D represents the plane area enclosed by the outline, L represents the outline, and P, Q represents the function on the area;
the formula used for the convexity calculation is expressed as:
wherein, TuIs convexity, SAreaIndicating the size of the area of the contour, SConvexHullRepresenting the area of the outline convex hull;
the formula used for the inertia ratio calculation is expressed as:
wherein m isuIs the ratio of inertia, mu20Is the x-direction 2-step invariant moment of the profile, mu02Is a constant moment of 2 steps in the y direction, mu112-step central invariant moment;
the formula adopted by the squareness calculation is expressed as:
wherein, JuIs a rectangular degree, SAreaIs the area of the outline of the spot, SMinRectAreaThe minimum circumscribed rectangle area of the spot outline;
the formula adopted by the calculation of the main shaft direction is expressed as
WhereinAnd theta denotes a principal axis direction angle, M11Representing the second central moment, M20The expression represents the second central moment in the x-direction, M02Representing the second central moment in the y-direction.
In addition, the invention also discloses a photovoltaic module infrared hot spot detection device, which comprises:
the framing module is used for carrying out video framing processing on the infrared video corresponding to the photovoltaic module and taking the frame image as an infrared image to be detected;
the segmentation module is used for segmenting the upper boundary and the lower boundary of the photovoltaic array by using horizontal projection segmentation, segmenting the left boundary and the right boundary by using vertical projection segmentation, combining the horizontal projection segmentation with the vertical projection segmentation, and segmenting the infrared image to be detected so as to segment the infrared image of the photovoltaic region;
the conversion module is used for converting the infrared image of the photovoltaic area into a gray image;
the acquiring module is used for setting a first number of gray level threshold values and carrying out binarization processing on the gray level image aiming at each gray level threshold value to acquire a first number of binarization images;
the determining module is used for determining the abnormal area of the photovoltaic module according to the first number of the binary images and extracting the outline by using an outline extracting function;
the extraction module is used for carrying out feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness and main shaft direction;
and the removing module is used for removing the hot spots caused by the junction boxes in the round large hot spots and removing the hot spots caused by illumination in the rectangular large hot spots.
The embodiment of the invention provides a method and a device for detecting infrared hot spots of a photovoltaic module, which comprises the steps of firstly obtaining an infrared image to be detected corresponding to the photovoltaic module; then, dividing the infrared image to be detected to divide the infrared image of the photovoltaic region and converting the infrared image into a gray image; then, carrying out binarization processing on the gray level images by using a first number of gray level thresholds to obtain a first number of binarized images; determining abnormal areas of the photovoltaic module according to the first number of binary images, and extracting the outline by using an outline extraction function; and performing feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness and main shaft direction; and then removing the hot spots caused by the junction boxes in the circular large hot spots and removing the hot spots caused by illumination in the rectangular large hot spots. The method and the device solve the problem that in the prior art, hot spot analysis is inaccurate due to interference caused by limit sum and illumination, and therefore the accuracy of hot spot analysis is improved by the embodiment of the invention.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a first method for detecting an infrared hot spot of a photovoltaic module according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a gray scale of an infrared imaging apparatus according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of the location of a hot spot for a preliminary analysis provided by an embodiment of the present invention;
fig. 4 is a schematic diagram of a finally determined hot spot provided by an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to solve the problems in the prior art, the embodiment of the invention provides a photovoltaic module infrared hot spot detection method.
Fig. 1 is a first method for detecting an infrared hot spot of a photovoltaic module, provided by an embodiment of the present invention, where the method includes:
s101, performing video framing processing on the infrared video corresponding to the photovoltaic module, and taking the frame image as an infrared image to be detected.
Carry the unmanned aerial vehicle of infrared camera device from beginning flight to other check points, at the flight in-process, unmanned aerial vehicle can be the speed operation of invariant, the infrared video of photovoltaic module in its infrared camera device that carries whole shooting photovoltaic power plant.
It can be appreciated that the drone may also operate at variable speeds. When the unmanned aerial vehicle operates at a variable speed, the change curve of the speed of the unmanned aerial vehicle along with time needs to be recorded, and the operating distance of the unmanned aerial vehicle is calculated according to the curve. In practical application, can adjust unmanned aerial vehicle's height or infrared camera device's shooting scope to make infrared camera device shoot the photovoltaic module of predetermineeing the quantity row, for example, can make the infrared camera device that unmanned aerial vehicle carried only shoot two rows of photovoltaic module with unmanned aerial vehicle's high reduction. That is to say, in the video frame of the infrared video shot by the infrared camera device, only two rows of photovoltaic modules are contained in the longitudinal direction of each video frame. In addition, the infrared camera device can be a device with the function of shooting infrared video, such as an infrared camera, a thermal infrared imager and the like.
The obtained infrared video is subjected to framing processing, so that a plurality of video frame images can be obtained, and specifically, how to perform framing processing of the infrared video images is the prior art, which is not described herein again in the embodiments of the present invention.
And S102, dividing the upper and lower boundaries of the photovoltaic array by horizontal projection segmentation, dividing the left and right boundaries by vertical projection segmentation, combining the horizontal projection segmentation and the vertical projection segmentation, and segmenting the infrared image to be detected to segment the infrared image of the photovoltaic region.
The infrared image contains a large amount of redundant information, in order to eliminate non-photovoltaic region interference and reduce the calculation data amount, the upper and lower boundaries of the photovoltaic array are divided by horizontal projection segmentation, the left and right boundaries are divided by vertical projection segmentation, the horizontal projection segmentation and the vertical projection segmentation are combined, and the intersection of the photovoltaic array in the horizontal projection region and the vertical projection region is taken to segment a detected photovoltaic region, namely a detected region of interest (ROI).
And S103, converting the infrared image of the photovoltaic region into a gray image.
To further reduce the amount of calculation data, the divided infrared image of the photovoltaic region is converted into a grayscale image by using a color space conversion function, as shown in fig. 2 for example. Specifically, the gray image is converted into the prior art, and the embodiment of the present invention is not described herein again.
S104, setting a first number of gray level threshold values, and carrying out binarization processing on the gray level image aiming at each gray level threshold value to obtain a first number of binarization images.
In order to accurately detect hot spots of the assembly and prevent the missing detection of darker hot spots or detect brighter hot spots far larger than the contour of the detected brighter hot spots, the gray level image is subjected to binary segmentation of multiple gray level threshold levels by using a threshold function. Illustratively, decomposing a picture at [50,200] with gray value 20 as an interval can obtain 8 images with segmentation thresholds [50,70,90,110,130,150,170,190], and the 8 binary images are processed respectively.
Taking the threshold value of 50 as an example, when the pixel value of the pixel point is greater than 50, the pixel value is represented as 50, otherwise, the pixel value is 0; then, the gray value 20 is added on the basis of the threshold value 50, that is, the gray threshold value is 70, for example, when the pixel value of the pixel point is greater than 70, the pixel value is 70, otherwise, the pixel value is 0, and so on, the corresponding binary image can be obtained after each gray threshold value is binarized, and therefore, how many binary images can be obtained by how many gray threshold values.
And S105, determining the abnormal area of the photovoltaic module according to the first number of binary images, and extracting the outline by using an outline extraction function.
And performing contour extraction on each gray image in the set by using a findContours () contour extraction function, setting the function mode as CV _ RETR _ LIST to detect all contours, and setting the method as CV _ CHAIN _ APPROX _ NONE to store all continuous contour points on the boundary of the object into a contours vector, namely performing contour extraction on the abnormal region of the photovoltaic component in each binary image by using the contour extraction function.
S106, carrying out feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness, and main axis direction.
In the embodiment of the invention, the same video frame infrared image is divided into two binary images with different gray levels and different thresholds, the number of times of the same hot spot appears is calculated, the two binary images are judged to be the hot spot when the number of times is more than a preset number, for example, the preset number is three.
And performing feature calculation on the contour extracted from each binary image, wherein the features at least comprise circularity, area, convexity, inertia ratio, rectangularity and main axis direction, and judging whether the contour is a circular large hot spot or a rectangular large hot spot according to the features.
The extracted contour is analyzed from the following aspects:
(1) circularity calculation
The circularity is used to describe the degree of approximation between the extracted contour and a circle, and the maximum value of the calculation formula is 1, that is, when the extracted contour is exactly a circle, the calculation formula is expressed as:
wherein, Area is the Area of the outline, perimeter is the perimeter of the outline, and the suspected hot spot can be determined when the range of the set circularity is 0.2-1 due to the existence of the oval and circular outlines.
(2) Calculation of area
There are many kinds of modes in the calculation of area, through what of statistics target region pixel number comes as the area of outline, utilize the relation of the curve integral of closed region and area to calculate the size of area, calculate the number of pixel in the image outline promptly, this kind of calculating speed is generally faster, and the area of calculating is more accurate, expresses as follows:
wherein: d denotes the area of the plane enclosed by the outline, L denotes the outline, and P, Q denotes the function over the area.
(3) Calculation of convexity
The convexity is mainly used for representing the size of the radian of one vertex, the radian is smaller and closer to a straight line, when the radian is 1, the radian is represented as a circle, the convexity is frequently used in calculation as an important parameter for screening graphs with radians, and the convexity is calculated as follows:
wherein: sAreaIndicating the size of the area of the contour, SConvexHullThe area of the contour convex hull is indicated. The value of the formula is usually 0.3 or more.
(4) Inertia ratio calculation
The inertia ratio is mainly used for looking at the ellipse of the contour region and is also an important parameter in the identification of the hot spot, and the calculation formula is as follows:
wherein: mu (u)20Is the x-direction 2-step invariant moment of the profile, mu02Is a constant moment of 2 steps in the y direction, mu11Is 2-step central invariant moment.
(5) Squareness calculation
The calculation of the squareness is directed at the identification of a large rectangular hot spot, the formation of which is generally related to the electrical connection failure of the battery of the module, the image is displayed as a rectangular highlight region extending along the long side of the module, and the calculation formula of the squareness is as follows:
wherein S isAreaIs the area of the outline of the spot, SMinRectAreaThe smallest circumscribed rectangular area of the spot profile.
(6) Spindle direction calculation
The principal axis represents the direction of the longest axis in an area, and is mainly used for identifying the finger directions of objects with different shapes, and the calculation formula 6 is as follows:
wherein: theta denotes the principal axis direction angle, M11Representing the second central moment, M20The representation represents the second-order central moment in the x-direction, and M02 represents the second-order central moment in the y-direction.
The detection result of the circular hot spot comprises hot spot false detection caused by junction box interference. The detection result of the rectangular large hot spot comprises false detection of the rectangular large hot spot caused by illumination interference, as shown in fig. 3.
The hot spot detection result in fig. 3 may include a circular hot spot, a large rectangular hot spot, illumination interference, and junction box interference, where the range where the rectangular frame in fig. 3 ends is that the spots in the middle of the string are regularly arranged as photovoltaic module junction boxes, and the junction boxes may emit more heat and exhibit the characteristic of a hot spot when the module normally works due to the presence of a large number of resistive devices, which may cause serious interference to the detection result. For the interference of the junction boxes, the arrangement of the junction boxes on the assembly is regular and mainly focuses on the upper and lower parts of the middle line of the cluster.
S107, removing the hot spots caused by the junction boxes in the circular large hot spots and removing the hot spots caused by light irradiation in the rectangular large hot spots.
Due to the similarity between the large rectangular hot spot and the illumination area, the misjudgment influence caused by the illumination area in the large rectangular hot spot needs to be removed, so that the real rectangular hot spot is obtained.
And each photovoltaic module is provided with a junction box which contains a large number of resistive components, so that a large amount of heat can be generated in the power generation process, and hot spot characteristics can be presented in an infrared image, thereby causing hot spot detection interference. The arrangement of the junction box on the assembly is regular, so that a junction box detection function is designed according to the arrangement when hot spot detection is carried out, and the junction box detection function is eliminated from hot spot detection results.
For the large rectangular hot spot, the interference of illumination and the large rectangular hot spot are mainly different in the change of the edge, the difference is not obvious in the outline, and only the difference of the texture change in the large rectangular hot spot can be analyzed. HOG characteristics are adopted to analyze the internal characteristics of the hot spot area and the illumination area, then the SVM is trained by using the obtained characteristic data, and the trained model is used for distinguishing the rectangular large hot spot from the illumination interference, so that the rectangular large hot spot is determined.
The histogram of oriented gradients feature is a feature descriptor used for object detection in computer vision and image processing. The HOG features are constructed by calculating and counting the histogram of gradient direction of local area of image. The gradients of the image in the horizontal and vertical directions are calculated by difference:
fx(x，y)＝f(x，y)-f(x+1，y)
fy(x，y)＝f(x，y)-f(x，y+1)
wherein f (x, y) is an image function, fx(x, y) is the gradient in the x-direction, fy(x, y) is the y-direction gradient, and (x, y) is the coordinates of the pixel.
Then obtaining the amplitude and direction of the gradient of each pixel point, and expressing as:
wherein,
is the magnitude of the gradient and is,
is the direction angle.
The SVM model is a support vector machine and is a common discrimination method. In the field of machine learning, the method is a supervised learning model, is generally used for pattern recognition, classification and regression analysis, and is a relatively mature classifier. The method is used for classifying the hot spots and the illumination in the hot spot detection.
The method disclosed by the invention is used for identifying the rectangular large hot spots by using a method of combining the HOG and the SVM, and firstly, a plurality of rectangular large hot spot images (positive samples) and illumination interference images (negative samples) with the same scale are prepared, and HOG characteristics of the rectangular large hot spot images and the illumination interference images are respectively extracted. And (4) providing the collected HOG feature vectors for SVM classification training to generate an xml file. The xml file is called in the hot spot detection process, feature calculation and recognition classification are carried out on the detection assembly, rectangular large hot spots and illumination false detection can be accurately distinguished, and illumination interference similar to the rectangular large hot spots is removed.
And (4) sorting hot spot detection results, and determining that the hot spot false detection caused by illumination is the result if traversing the hot spot coordinate for only 1 time in the detection results according to the characteristic that the illumination moves along with the video. Aiming at the rectangular large hot spot false detection caused by illumination, the rectangular large hot spot false detection caused by illumination is removed by adopting a method of extracting and analyzing HOG characteristics of a photovoltaic module and classifying and identifying the rectangular large hot spot and illumination interference by using an SVM (support vector machine), and the hot spot detection result after the interference is removed is shown in figure 4.
Illustratively, a hot spot algorithm is used for detecting a certain fishing complementary power station in Jiangsu, 45 video materials are selected for testing, and the performance of the detection algorithm can be obtained by comparing results obtained by manual detection and algorithm detection. The hot spot detection algorithm test results are shown in table 1.
TABLE 1
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the apparatus, the electronic device, the storage medium, and the system embodiment, since they are substantially similar to the method embodiment, the description is relatively simple, and in relation to the description, reference may be made to part of the description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (5)
1. A photovoltaic module infrared hot spot detection method is characterized by comprising the following steps:
performing video framing processing on an infrared video corresponding to the photovoltaic module, and taking a frame image as an infrared image to be detected;
dividing the upper boundary and the lower boundary of the photovoltaic array by using horizontal projection division, dividing the left boundary and the right boundary by using vertical projection division, combining the horizontal projection division and the vertical projection division, and dividing the infrared image to be detected to divide the infrared image of the photovoltaic region;
converting the photovoltaic area infrared image into a gray level image;
setting a first number of gray level threshold values, and carrying out binarization processing on the gray level images aiming at each gray level threshold value to obtain a first number of binarization images;
determining an abnormal area of the photovoltaic module according to the first number of binary images, and extracting the outline by using an outline extraction function;
and performing feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness and main shaft direction;
and removing hot spots caused by the junction box in the round large hot spots and removing hot spots caused by illumination in the rectangular large hot spots.
2. The infrared hot spot detection method for the photovoltaic module as claimed in claim 1, wherein the step of removing the hot spots caused by the illumination in the large rectangular hot spots comprises:
carrying out normalization processing on the image corresponding to the rectangular large hot spot;
calculating the horizontal gradient and the vertical gradient of the image in a differential mode;
dividing the window into cell units, calculating the gradient and direction of the cell units, and grouping and weighting the cell units;
combining adjacent cell units into pixel blocks, and counting pixel block histograms;
normalizing the gradient histogram and collecting HOG characteristics;
inputting the HOG characteristics into an SVM model to obtain an output result of the SVM model so as to determine whether the hot spot is caused by illumination;
the method comprises the steps that an SVM model is used for converting all samples into sample pictures with the width of 32 x 64 according to pre-collected solar illumination false detection and rectangular large hot spot samples in a large number of different types of photovoltaic power stations, labeling the illumination and rectangular large hot spot sample pictures respectively and setting labels, extracting HOG feature vectors of each sample, calling a calculation function to calculate a result array of each sample, storing a sample matrix and type labels of the sample matrix, setting SVM training parameters for SVM classification training to generate an xml file, calling the xml file in a hot spot detection process, carrying out HOG feature calculation on detection components segmented from a video frame to obtain the result array, comparing the result array with SVM training results to carry out identification and classification, and being capable of accurately distinguishing rectangular large hot spots from illumination false detection and removing illumination interference similar to rectangular large hot spots.
3. The infrared hot spot detection method for the photovoltaic module as claimed in claim 1 or 2, wherein the step of removing the hot spots caused by the junction boxes in the large circular hot spots comprises:
according to the layout of the photovoltaic assembly junction box, a junction box detection function is designed by adopting an image masking method, and circular hot spots caused by the junction box are removed.
4. The infrared hot spot detection method of the photovoltaic module as claimed in claim 3, characterized in that the circularity calculation uses a formula expressed as:
wherein E iscsIs the circularity, Area is the Area of the profile, perimeter is the perimeter of the profile;
the formula used for the area calculation is expressed as:
wherein S is the area, D represents the plane area enclosed by the outline, L represents the outline, and P, Q represents the function on the area;
the formula used for the convexity calculation is expressed as:
wherein, TuIs convexity, SAreaIndicating the size of the area of the contour, SConvexHullRepresenting the area of the outline convex hull;
the formula used for the inertia ratio calculation is expressed as:
wherein m isuIs the ratio of inertia, mu20Is the x-direction 2-step invariant moment of the profile, mu02Is a constant moment of 2 steps in the y direction, mu112-step central invariant moment;
the formula adopted by the squareness calculation is expressed as:
wherein, JuIs a rectangular degree, SAreaIs the area of the outline of the spot, SMinRectAreaThe minimum circumscribed rectangle area of the spot outline;
the formula adopted by the calculation of the main shaft direction is expressed as
Where θ represents the principal axis direction angle, M11Representing the second central moment, M20The expression represents the second central moment in the x-direction, M02Representing the second central moment in the y-direction.
5. The utility model provides an infrared hot spot detection device of photovoltaic module which characterized in that includes:
the framing module is used for carrying out video framing processing on the infrared video corresponding to the photovoltaic module and taking the frame image as an infrared image to be detected;
the segmentation module is used for segmenting the upper boundary and the lower boundary of the photovoltaic array by using horizontal projection segmentation, segmenting the left boundary and the right boundary by using vertical projection segmentation, combining the horizontal projection segmentation with the vertical projection segmentation, and segmenting the infrared image to be detected so as to segment the infrared image of the photovoltaic region;
the conversion module is used for converting the infrared image of the photovoltaic area into a gray image;
the acquiring module is used for setting a first number of gray level threshold values and carrying out binarization processing on the gray level image aiming at each gray level threshold value to acquire a first number of binarization images;
the determining module is used for determining the abnormal area of the photovoltaic module according to the first number of the binary images and extracting the outline by using an outline extracting function;
the extraction module is used for carrying out feature calculation on the extracted contour, and judging a circular large hot spot and a rectangular large hot spot according to a calculation result and a preset value range, wherein the features at least comprise: circularity, area, convexity, inertia ratio, squareness and main shaft direction;
and the removing module is used for removing the hot spots caused by the junction boxes in the round large hot spots and removing the hot spots caused by illumination in the rectangular large hot spots.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911300787.2A CN111080691A (en) | 2019-12-17 | 2019-12-17 | Infrared hot spot detection method and device for photovoltaic module |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911300787.2A CN111080691A (en) | 2019-12-17 | 2019-12-17 | Infrared hot spot detection method and device for photovoltaic module |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111080691A true CN111080691A (en) | 2020-04-28 |
Family
ID=70315166
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911300787.2A Pending CN111080691A (en) | 2019-12-17 | 2019-12-17 | Infrared hot spot detection method and device for photovoltaic module |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111080691A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112037191A (en) * | 2020-08-28 | 2020-12-04 | 晶科能源有限公司 | Method and device for determining local leakage current density threshold and computer equipment |
| CN112101365A (en) * | 2020-09-10 | 2020-12-18 | 国网辽宁省电力有限公司电力科学研究院 | Power equipment key feature extraction method and system based on infrared thermal image processing |
| CN112116600A (en) * | 2020-08-17 | 2020-12-22 | 内蒙古能建数字信息科技有限公司 | Photovoltaic panel counting method based on image processing |
| CN112182967A (en) * | 2020-09-28 | 2021-01-05 | 四川渔光物联技术有限公司 | Automatic photovoltaic module modeling and hot spot positioning method based on thermal imaging instrument |
| CN113155288A (en) * | 2020-11-30 | 2021-07-23 | 齐鲁工业大学 | Image identification method for hot spots of photovoltaic cell |
| CN114565517A (en) * | 2021-12-29 | 2022-05-31 | 骨圣元化机器人(深圳)有限公司 | Image denoising method and device for infrared camera and computer equipment |
| CN114663389A (en) * | 2022-03-21 | 2022-06-24 | 上海电气集团股份有限公司 | Photovoltaic module hot spot detection method and device and storage medium |
| CN115049648A (en) * | 2022-08-12 | 2022-09-13 | 北京中拓新源科技有限公司 | Hot spot defect detection method and device for photovoltaic module |
| CN116660317A (en) * | 2023-07-25 | 2023-08-29 | 北京智芯微电子科技有限公司 | Hot spot detection method, system, processor and storage medium of photovoltaic array |
| EP4290454A1 (en) * | 2022-06-10 | 2023-12-13 | Commissariat à l'énergie atomique et aux énergies alternatives | Method and device for segmenting at least one color source image representative of a photovoltaic plant |
| CN118096748A (en) * | 2024-04-25 | 2024-05-28 | 国网山东省电力公司鱼台县供电公司 | Fault detection method and system for photovoltaic array |
| US12044576B2 (en) | 2021-04-20 | 2024-07-23 | Abb Schweiz Ag | System for monitoring a device |
-
2019
- 2019-12-17 CN CN201911300787.2A patent/CN111080691A/en active Pending
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112116600A (en) * | 2020-08-17 | 2020-12-22 | 内蒙古能建数字信息科技有限公司 | Photovoltaic panel counting method based on image processing |
| CN112037191A (en) * | 2020-08-28 | 2020-12-04 | 晶科能源有限公司 | Method and device for determining local leakage current density threshold and computer equipment |
| CN112037191B (en) * | 2020-08-28 | 2023-08-25 | 晶科能源股份有限公司 | Method and device for determining local leakage current density threshold value and computer equipment |
| CN112101365A (en) * | 2020-09-10 | 2020-12-18 | 国网辽宁省电力有限公司电力科学研究院 | Power equipment key feature extraction method and system based on infrared thermal image processing |
| CN112182967A (en) * | 2020-09-28 | 2021-01-05 | 四川渔光物联技术有限公司 | Automatic photovoltaic module modeling and hot spot positioning method based on thermal imaging instrument |
| CN113155288A (en) * | 2020-11-30 | 2021-07-23 | 齐鲁工业大学 | Image identification method for hot spots of photovoltaic cell |
| US12044576B2 (en) | 2021-04-20 | 2024-07-23 | Abb Schweiz Ag | System for monitoring a device |
| CN114565517B (en) * | 2021-12-29 | 2023-09-29 | 骨圣元化机器人(深圳)有限公司 | Image denoising method and device of infrared camera and computer equipment |
| CN114565517A (en) * | 2021-12-29 | 2022-05-31 | 骨圣元化机器人(深圳)有限公司 | Image denoising method and device for infrared camera and computer equipment |
| CN114663389A (en) * | 2022-03-21 | 2022-06-24 | 上海电气集团股份有限公司 | Photovoltaic module hot spot detection method and device and storage medium |
| EP4290454A1 (en) * | 2022-06-10 | 2023-12-13 | Commissariat à l'énergie atomique et aux énergies alternatives | Method and device for segmenting at least one color source image representative of a photovoltaic plant |
| FR3136576A1 (en) * | 2022-06-10 | 2023-12-15 | Commissariat à l'énergie atomique et aux énergies alternatives | Method and device for segmenting at least one source color image representative of a photovoltaic power plant |
| CN115049648B (en) * | 2022-08-12 | 2022-11-01 | 北京中拓新源科技有限公司 | Hot spot defect detection method and device for photovoltaic module |
| CN115049648A (en) * | 2022-08-12 | 2022-09-13 | 北京中拓新源科技有限公司 | Hot spot defect detection method and device for photovoltaic module |
| CN116660317A (en) * | 2023-07-25 | 2023-08-29 | 北京智芯微电子科技有限公司 | Hot spot detection method, system, processor and storage medium of photovoltaic array |
| CN116660317B (en) * | 2023-07-25 | 2023-12-22 | 北京智芯微电子科技有限公司 | Hot spot detection method, system, processor and storage medium of photovoltaic array |
| CN118096748A (en) * | 2024-04-25 | 2024-05-28 | 国网山东省电力公司鱼台县供电公司 | Fault detection method and system for photovoltaic array |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111080691A (en) | Infrared hot spot detection method and device for photovoltaic module | |
| Zhao et al. | Cloud shape classification system based on multi-channel cnn and improved fdm | |
| CN108537154B (en) | Power transmission line bird nest identification method based on HOG characteristics and machine learning | |
| CN109344753A (en) | A kind of tiny fitting recognition methods of Aerial Images transmission line of electricity based on deep learning | |
| CN111260629A (en) | Pantograph structure abnormity detection algorithm based on image processing | |
| CN108711148B (en) | Tire defect intelligent detection method based on deep learning | |
| CN113324864B (en) | Pantograph carbon slide plate abrasion detection method based on deep learning target detection | |
| CN108680833B (en) | Composite insulator defect detection system based on unmanned aerial vehicle | |
| CN111784633A (en) | Insulator defect automatic detection algorithm for power inspection video | |
| CN111852792B (en) | Fan blade defect self-diagnosis positioning method based on machine vision | |
| CN111242899B (en) | Image-based flaw detection method and computer-readable storage medium | |
| CN110751619A (en) | Insulator defect detection method | |
| CN110533654A (en) | The method for detecting abnormality and device of components | |
| CN117677969A (en) | Defect detection method and device | |
| CN109120230B (en) | A kind of solar battery sheet EL image detection and defect identification method | |
| CN105303162A (en) | Target proposed algorithm-based insulator recognition algorithm for aerial images | |
| CN116309407A (en) | Method for detecting abnormal state of railway contact net bolt | |
| CN115841633A (en) | Power tower and power line associated correction power tower and power line detection method | |
| Haoran et al. | Egg crack detection based on support vector machine | |
| CN117252840B (en) | Photovoltaic array defect elimination evaluation method and device and computer equipment | |
| CN111597939B (en) | High-speed rail line nest defect detection method based on deep learning | |
| CN110070520B (en) | Pavement crack detection model construction and detection method based on deep neural network | |
| CN117036314A (en) | Method for detecting oxidation area of high-density flexible IC substrate | |
| CN115187878A (en) | Unmanned aerial vehicle image analysis-based blade defect detection method for wind power generation device | |
| CN112991252A (en) | Method for detecting loss of vibration damper on power transmission line |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200428 |
|
| WD01 | Invention patent application deemed withdrawn after publication |