CN113989260A - Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image - Google Patents

Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image Download PDF

Info

Publication number
CN113989260A
CN113989260A CN202111321628.8A CN202111321628A CN113989260A CN 113989260 A CN113989260 A CN 113989260A CN 202111321628 A CN202111321628 A CN 202111321628A CN 113989260 A CN113989260 A CN 113989260A
Authority
CN
China
Prior art keywords
photovoltaic panel
image
aerial vehicle
unmanned aerial
segmentation
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
Application number
CN202111321628.8A
Other languages
Chinese (zh)
Inventor
戴敏敏
戴超超
鲍丽娟
刘刚
曹蕴瀚
郑恩辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Jiliang University
Zhejiang Zheneng Jiahua Power Generation Co Ltd
Original Assignee
China Jiliang University
Zhejiang Zheneng Jiahua Power Generation Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Jiliang University, Zhejiang Zheneng Jiahua Power Generation Co Ltd filed Critical China Jiliang University
Priority to CN202111321628.8A priority Critical patent/CN113989260A/en
Publication of CN113989260A publication Critical patent/CN113989260A/en
Pending legal-status Critical Current

Links

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
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • 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
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • 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
    • G06T2207/20076—Probabilistic image processing
    • 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
    • G06T2207/30148—Semiconductor; IC; Wafer

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images. Acquiring an infrared image of the photovoltaic panel through unmanned aerial vehicle acquisition; performing optimization iteration processing on a segmentation threshold t aiming at the infrared image to obtain a first optimal segmentation threshold; calculating probability distribution aiming at the infrared image division region to obtain a second optimal division threshold value; and performing weighted calculation on the two optimal segmentation threshold values to obtain a final segmentation threshold value, and segmenting the infrared image according to the final segmentation threshold value to obtain the edge of the photovoltaic panel. The method better considers the influence of the image gray scale, the background and the number of target pixels, thereby obtaining better segmentation effect; the improved traditional method can achieve good image segmentation effect under the conditions of high exposure and low contrast.

Description

Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image
Technical Field
The invention belongs to a photovoltaic panel image processing method in the field of computer vision and image analysis, and particularly relates to a photovoltaic panel edge segmentation method based on an unmanned aerial vehicle infrared image, which is mainly used for image segmentation processing in infrared image processing.
Background
In a power plant, solar modules are typically installed in a wide, sunny area. The shielding objects such as flying birds, dust, fallen leaves and the like are difficult to fall on for a long time, the shielding objects form shadows on the solar cell module, and the current and the voltage of certain single cells in the solar cell module are changed due to the existence of local shadows. As a result, the product of the local current and voltage of the solar cell module increases, and a local temperature rise occurs in the solar cell module. Defects of certain cell single sheets in the solar cell module can also cause the module to generate heat locally during operation, and the phenomenon is called a 'hot spot effect'. The hot spot effect of the photovoltaic panel and how to detect the existence of the hot plate are important research subjects for ensuring the normal work of the photovoltaic panel.
The detection of the boundary of the photovoltaic panel can provide the most basic position information for the thermal panel detection. In recent years, with the development of unmanned aerial vehicle technology and the development of infrared imaging devices, the price of an infrared sensor is lower and lower, the resolution is gradually improved, the image resolution of 640X512 can be achieved in M300 RTK, and the infrared image detection based on the unmanned aerial vehicle provides basic guarantee for all-weather hot spot detection.
In the process of acquiring the infrared image, due to adverse effects of factors such as uneven heating, infrared radiation of the environment and equipment, uneven surface and internal structure of the tested piece and the like, the acquired defect information in the thermal wave image is submerged by a large amount of irrelevant noise.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides an infrared image boundary segmentation method for a photovoltaic panel under the view angle of an unmanned aerial vehicle, which overcomes the defects of the traditional method, well improves the photovoltaic panel and the background and the region with larger contrast, improves the segmentation precision, realizes multi-scene segmentation and improves the flexibility of positioning the photovoltaic panel of a power plant.
As shown in fig. 1, the object of the present invention is achieved by:
(1) acquiring an infrared image of the photovoltaic panel by an unmanned aerial vehicle, preprocessing the image and converting the image into gray scale, as shown in fig. 2;
(2) optimizing iterative processing of segmentation threshold T aiming at infrared image to obtain first optimal segmentation threshold TOtsu；
(3) Calculating probability distribution aiming at infrared image divided regions to obtain second optimal segmentation threshold value TKsw；
(4) Dividing the second optimal division threshold value TKswAnd a first optimal segmentation threshold TOtsuAnd performing weighting calculation to obtain a final segmentation threshold, and segmenting the infrared image according to the final segmentation threshold to obtain the edge of the photovoltaic panel, as shown in fig. 3.
The unmanned aerial vehicle on install infrared camera, unmanned aerial vehicle flies to the sky above the solar field, infrared camera is to shooting the photovoltaic panel subassembly in the solar field downwards and obtain infrared image.
The step (2) is specifically as follows:
(2.1) initially setting a segmentation threshold t to be 1; in specific implementation, the value range of the segmentation threshold t is 0-255, 0 is pure black, and 255 is pure white.
(2.2) classifying the pixels of the infrared image smaller than the segmentation threshold t as a background part C0, otherwise classifying the pixels as a target part C1, wherein the gray level range of the background part C0 is 0-t-1, the gray level range of the target part C1 is t-L-1, and L is a preset maximum value of the gray level;
(2.3) calculating the probability, the mean gray value, the intra-class variance and the number of pixels of the background part C0 and the target part C1 respectively;
(2.4) calculating the feature parameter δ from the probabilities, mean gray-scale values, intra-class variances, and pixel counts of the background portion C0 and the target portion C12(k)；
(2.5) if t < L, making t equal to t +1, and returning to step (2.2);
(2.6) if t is more than or equal to L, correspondingly obtaining all characteristic parameters delta according to the change of t from 0 to L-12(T) finding a division threshold T in which the maximum value of the characteristic parameter corresponds to the maximum value as a first optimum division threshold TOtsu。
In the step (2.4), the characteristic parameter delta is calculated according to the following formula2(t)：
δ2(t)＝(ω0(u-u0)+ω1(u-u1))*(cnt0-cnt1)
Wherein, w0Probability of background portion C0, w1Is the probability, u, of the target portion C10Is the mean gray value of the background portion C0, u1Is the average gray value, v, of the target portion C10Within class variance, v, of background component C01The intra-class variance of the target portion C1 is denoted by cnt0 as the total number of pixels in the background portion C0 and cnt1 as the total number of pixels in the target portion C1.
According to the invention, the infrared image of the photovoltaic panel is acquired by the unmanned aerial vehicle, and then the infrared image is processed to obtain a purer image and a more accurate detection and segmentation result.
In the present invention, the infrared image I is represented by f (x, y)MXNThe gray value at the (x, y) position is a gray image, the gray level L of which is 256, f (x, y) belongs to [0, L-1 ]]。
Probability p of occurrence of a pixel with a gray level iiComprises the following steps:
Figure BDA0003345760810000021
Figure BDA0003345760810000031
wherein p isiEach represents the probability of occurrence of a pixel having a gray level i, i being 0,1iRepresenting the total number of pixels at the same gray level i; m, N respectively indicate the number of pixels in length and width of the input image, where M × N is the total number of pixels in the image.
In the steps (2.3) and (2.4), the average gray value u is calculated as:
Figure BDA0003345760810000032
wherein i represents the number of gray levels, piRespectively, the occurrence probability of a pixel having a gray level i.
The step (2.4) comprises the following steps:
(2.41) probability w of background portion C00And the probability w of the target portion C11The calculation is as follows:
Figure BDA0003345760810000033
Figure BDA0003345760810000034
w0+w1＝1
wherein, w0Probability of background portion C0, w1Probability of being the target portion C1;
(2.42) average grayscale value u of background portion C0 image0And the average gray value u of the image of the target portion C11The calculation is as follows:
Figure BDA0003345760810000035
Figure BDA0003345760810000036
the step (3) is specifically as follows:
(3) setting the gray scale range of an image as {0,1,2.,. L-1}, dividing the image into T, B two regions according to a preset segmentation threshold k, wherein the region T is a distribution of i ∈ {0, 1.,. k }, and the region B is a distribution of i ∈ { k + 1.,. L-1}, and the specific form is as follows:
Figure BDA0003345760810000037
Figure BDA0003345760810000038
Figure BDA0003345760810000041
wherein, PiRepresenting the probability of occurrence of a pixel with a gray level i;
computing total entropy of an image using entropy of probability distributions of two regions
Figure BDA0003345760810000042
Comprises the following steps:
Figure BDA0003345760810000043
Figure BDA0003345760810000044
Figure BDA0003345760810000045
solving the total entropy of an image
Figure BDA0003345760810000046
Division threshold of maximum valuek as the first optimal segmentation threshold fKswI.e. expressed as:
Figure BDA0003345760810000047
in the above (4), the second optimal division threshold value T is setKswWith a first optimal segmentation threshold TOtsuWeighting to obtain the final segmentation threshold TFinal：
TFinal＝α×TOtsu+β×TKsw
Wherein alpha and beta respectively represent a first weight coefficient and a second weight coefficient, the values of the weight coefficients alpha and beta are between (0 and 1), alpha is 0.7, and beta is 0.6.
Optimal threshold T obtained by maximum entropy methodKswAnd the optimal threshold value T obtained in the maximum characteristic parameter Otsu methodOtsuAnd (5) carrying out threshold weighting to determine the optimal threshold of the final segmentation.
The method is mainly used for image segmentation processing in infrared image processing of the photovoltaic panel.
Compared with the prior art, the invention has the beneficial effects that:
because the contrast ratio of the infrared image is low, the gray level range is narrow, and the target cannot be well segmented by using the maximum characteristic parameter method.
The invention considers the influence of the image gray level, the image background pixel number and the target pixel number on the segmentation, improves the existing image processing and method, can better make up for the defect that the infrared image segmentation algorithm cannot obtain better segmentation effect under the conditions of exposure and lower contrast between the background and the target by combining the maximum entropy method, and can better segment the region with higher contrast between the target and the background, thereby better segmenting the target.
The method better considers the influence of the image gray level, the background and the number of the target pixels, carries out weighting calculation by combining a maximum entropy method, has stronger robustness of the algorithm, and thus obtains better segmentation effect. The invention improves the traditional method, so that the image segmentation effect can be good under the conditions of high exposure and low contrast.
Drawings
FIG. 1 is a flow chart of an infrared image segmentation algorithm of the present invention.
Fig. 2 is an infrared image to be measured of the photovoltaic panel under the view angle of the unmanned aerial vehicle in the embodiment of the invention.
FIG. 3 is a diagram illustrating the results of an infrared image segmentation algorithm according to an embodiment of the present invention.
Detailed Description
The invention is described in more detail below with reference to the accompanying drawings:
as shown in fig. 1, the embodiment of the method of the present invention and its implementation are as follows:
(1) acquiring an infrared image of the photovoltaic panel by an unmanned aerial vehicle, as shown in fig. 2;
install infrared camera on the unmanned aerial vehicle, unmanned aerial vehicle flies to the sky above the solar field, and infrared camera is downwards to the photovoltaic panel subassembly in the solar field shoot and obtain infrared image.
(2) Optimizing iterative processing of segmentation threshold T aiming at infrared image to obtain first optimal segmentation threshold TOtsu；
(2.1) initially setting a segmentation threshold t to be 1; in specific implementation, the value range of the segmentation threshold t is 0-255, 0 is pure black, and 255 is pure white.
(2.2) classifying the pixels of the infrared image smaller than the segmentation threshold t as a background part C0, otherwise classifying the pixels as a target part C1, wherein the gray level range of the background part C0 is 0-t-1, the gray level range of the target part C1 is t-L-1, and L is a preset maximum value of the gray level;
(2.3) calculating the probability, the average gray value and the number of pixels of the background part C0 and the target part C1 respectively;
(2.4) calculating the characteristic parameter δ from the probabilities, mean gray values and number of pixels of the background part C0 and the target part C12(t)；
(2.41) probability w of background portion C00And the probability w of the target portion C11The calculation is as follows:
Figure BDA0003345760810000051
Figure BDA0003345760810000052
w0+w1＝1
wherein, w0Probability of background portion C0, w1Probability of being the target portion C1;
(2.42) average grayscale value u of background portion C0 image0And the average gray value u of the image of the target portion C11The calculation is as follows:
Figure BDA0003345760810000061
Figure BDA0003345760810000062
the invention improves the traditional maximum characteristic parameter (Otsu) method, considers the relation between the background gray scale and the target gray scale and the number of pixels of the background and the target part, and can ensure that the region with larger contrast ratio has better segmentation effect compared with the traditional method.
In step (2.4), the characteristic parameter δ is calculated according to the following formula2(t)：
δ2(t)＝(ω0(u-u0)+ω1(u-u1))*(cnt0-cnt1)
Wherein, w0Probability of background portion C0, w1Is the probability, u, of the target portion C10Is the mean gray value of the background portion C0, u1The cnt0 is the total number of pixels in the background portion C0, and the cnt1 is the total number of pixels in the target portion C1, which is the average gray scale value of the target portion C1.
In the present invention, the infrared image I is represented by f (x, y)MXNGray value at the (x, y) position, the image being a gray image, the gray level L of which＝256，f(x，y)∈[0，L－1]。
Probability p of occurrence of a pixel with a gray level iiComprises the following steps:
Figure BDA0003345760810000063
Figure BDA0003345760810000064
wherein p isiEach represents the probability of occurrence of a pixel having a gray level i, i being 0,1iRepresenting the total number of pixels at the same gray level i; m, N respectively indicate the number of pixels in length and width of the input image, where M × N is the total number of pixels in the image.
(2.5) if t < L, making t equal to t +1, and returning to step (2.2);
(2.6) if t is more than or equal to L, correspondingly obtaining all characteristic parameters delta according to the change of t from 0 to L-12(T) finding a division threshold T in which the maximum value of the characteristic parameter corresponds to the maximum value as an optimum division threshold TOtsu。
(3) Setting the gray scale range of an image as {0,1,2.,. L-1}, dividing the image into T, B two regions according to a preset segmentation threshold k, wherein the region T is a distribution of i ∈ {0, 1.,. k }, and the region B is a distribution of i ∈ { k + 1.,. L-1}, and the specific form is as follows:
Figure BDA0003345760810000065
Figure BDA0003345760810000066
Figure BDA0003345760810000071
wherein, PiRepresenting the probability of occurrence of a pixel with a gray level i;
computing total entropy of an image using entropy of probability distributions of two regions
Figure BDA0003345760810000072
Comprises the following steps:
Figure BDA0003345760810000073
Figure BDA0003345760810000074
Figure BDA0003345760810000075
solving the total entropy of an image
Figure BDA0003345760810000076
The division threshold k of the maximum value is used as the first optimal division threshold fKswI.e. expressed as:
Figure BDA0003345760810000077
in the step (4), the step (c),
firstly, a second optimal segmentation threshold value T is obtainedKswWith a first optimal segmentation threshold TOtsuWeighting to obtain the final segmentation threshold TFinal：
TFinal＝α×TOtsu+β×TKsw
Wherein alpha and beta respectively represent a first weight coefficient and a second weight coefficient, the values of the weight coefficients alpha and beta are between (0 and 1), alpha is 0.7, and beta is 0.6.
Then, a binary image obtained after the infrared image f (x, y) is segmented is recorded as g (x, y), and an infrared image binarization algorithm is shown as the following formula:
Figure BDA0003345760810000078
wherein g (x, y) represents when the gray level of the original image is greater than TFinalWhen the gray value is 255 and less than TFinalWhen this is the case, the gray value is 0. The image is segmented to obtain the edge of the photovoltaic panel, and the segmentation result is shown in fig. 3.
The algorithm of the embodiment completes the segmentation of the photovoltaic panel in the infrared image under the view angle of the unmanned aerial vehicle, compared with the traditional infrared image segmentation method, the method has great improvement on the precision, the edge segmentation is obviously due to the traditional image processing method, and the method has the greatest advantages of overcoming the problems that the boundary is not easy to distinguish due to continuous infrared image hot zones and the segmentation effect is poor when the contrast of the background and the target is high. The method can be applied to the fields of photovoltaic panel positioning, defect detection and the like under the infrared scene under the visual angle of the unmanned aerial vehicle.

Claims (8)

1. A photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images is characterized by comprising the following steps: the method comprises the following steps:
(1) acquiring an infrared image of the photovoltaic panel;
(2) optimizing iterative processing of segmentation threshold T aiming at infrared image to obtain first optimal segmentation threshold TOtsu；
(3) Calculating probability distribution aiming at infrared image divided regions to obtain second optimal segmentation threshold value TKsw；
(4) Dividing the second optimal division threshold value TKswAnd a first optimal segmentation threshold TOtsuAnd performing weighting calculation to obtain a final segmentation threshold, and segmenting the infrared image according to the final segmentation threshold to obtain the edge of the photovoltaic panel.
2. The photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 1, characterized in that: the unmanned aerial vehicle on install infrared camera, unmanned aerial vehicle flies to the sky above the solar field, infrared camera is to shooting the photovoltaic panel subassembly in the solar field downwards and obtain infrared image.
3. The photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 1, characterized in that: the step (2) is specifically as follows:
(2.1) initially setting a segmentation threshold t to be 1;
(2.2) classifying the pixels of the infrared image smaller than the segmentation threshold t as a background part C0, otherwise classifying the pixels as a target part C1, wherein the gray scale range of the background part C0 is 0-t-1, the gray scale range of the target part C1 is t-L-1, and L is the maximum value of the gray scale;
(2.3) calculating the probability, the mean gray value, the intra-class variance and the number of pixels of the background part C0 and the target part C1 respectively;
(2.4) calculating the feature parameter δ from the probabilities, mean gray-scale values, intra-class variances, and pixel counts of the background portion C0 and the target portion C12(k)；
(2.5) if t < L, making t equal to t +1, and returning to the step (2.2);
(2.6) if t is more than or equal to L, correspondingly obtaining all characteristic parameters delta according to the change of t from 0 to L-12(T), finding a division threshold T in which the maximum value corresponds to the maximum value as a first optimal division threshold TOtsu。
4. The photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 3, characterized in that: in the step (2.4), the characteristic parameter delta is calculated according to the following formula2(t)：
δ2(t)＝(ω0(u-u0)+ω1(u-u1))*(cnt0-cnt1)
Wherein, w0Probability of background portion C0, w1Is the probability, u, of the target portion C10Is the mean gray value of the background portion C0, u1Is the average gray value, v, of the target portion C10Within class variance, v, of background component C01The intra-class variance of the target portion C1 is denoted by cnt0 as the total number of pixels in the background portion C0 and cnt1 as the total number of pixels in the target portion C1.
5. The photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 3, characterized in that: in the steps (2.3) and (2.4), the average gray value u is calculated as:
Figure FDA0003345760800000021
wherein i represents the number of gray levels, piRespectively, the occurrence probability of a pixel having a gray level i.
6. The photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 1, characterized in that: the step (2.4) comprises the following steps:
(2.41) probability w of background portion C00And the probability w of the target portion C11The calculation is as follows:
Figure FDA0003345760800000022
Figure FDA0003345760800000023
w0+w1＝1
wherein, w0Probability of background portion C0, w1Probability of being the target portion C1;
(2.42) average grayscale value u of background portion C0 image0And the average gray value u of the image of the target portion C11The calculation is as follows:
Figure FDA0003345760800000024
Figure FDA0003345760800000025
7. the photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 1, characterized in that: the step (3) is specifically as follows:
(3) setting the gray scale range of the image as {0,1,2.., L-1}, and dividing the image into T, B two regions according to a preset segmentation threshold k, wherein the specific form is as follows:
T：
Figure FDA0003345760800000026
B：
Figure FDA0003345760800000027
Figure FDA0003345760800000031
wherein, PiRepresenting the probability of occurrence of a pixel with a gray level i;
computing total entropy of an image using entropy of probability distributions of two regions
Figure FDA0003345760800000032
Comprises the following steps:
Figure FDA0003345760800000033
Figure FDA0003345760800000034
Figure FDA0003345760800000035
solving the total entropy of an image
Figure FDA0003345760800000036
The maximum value of the division threshold k is used as the first optimal division threshold TKswI.e. expressed as:
Figure FDA0003345760800000037
8. the photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared images according to claim 1, characterized in that: in the above (4), the second optimal division threshold value T is setKswWith a first optimal segmentation threshold TOtsuWeighting to obtain the final segmentation threshold TFinal：
TFinal＝α×TOtsu+β×TKsw
Where α and β represent a first weight coefficient and a second weight coefficient, respectively.
CN202111321628.8A 2021-11-09 2021-11-09 Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image Pending CN113989260A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111321628.8A CN113989260A (en) 2021-11-09 2021-11-09 Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111321628.8A CN113989260A (en) 2021-11-09 2021-11-09 Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image

Publications (1)

Publication Number Publication Date
CN113989260A true CN113989260A (en) 2022-01-28

Family

ID=79747453

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111321628.8A Pending CN113989260A (en) 2021-11-09 2021-11-09 Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image

Country Status (1)

Country Link
CN (1) CN113989260A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
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

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
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

Similar Documents

Publication Publication Date Title
CN111179334B (en) Sea surface small-area oil spill area detection system and detection method based on multi-sensor fusion
CN107680054B (en) Multi-source image fusion method in haze environment
CN108230264B (en) Single image defogging method based on ResNet neural network
CN103996209B (en) Infrared vessel object segmentation method based on salient region detection
Stromer et al. Enhanced crack segmentation (eCS): a reference algorithm for segmenting cracks in multicrystalline silicon solar cells
Di Tommaso et al. A multi-stage model based on YOLOv3 for defect detection in PV panels based on IR and visible imaging by unmanned aerial vehicle
CN111428748A (en) Infrared image insulator recognition and detection method based on HOG characteristics and SVM
CN107507215B (en) Power equipment infrared heat map segmentation method based on adaptive quantization enhancement
CN116977329B (en) Photovoltaic grid line detection method based on machine vision
CN107944354B (en) Vehicle detection method based on deep learning
CN112258490A (en) Low-emissivity coating intelligent damage detection method based on optical and infrared image fusion
CN112446863A (en) Photovoltaic array hot spot detection method and detection system based on image processing
CN113989260A (en) Photovoltaic panel edge segmentation method based on unmanned aerial vehicle infrared image
CN114677336A (en) Curtain wall panel damage identification method based on infrared image
CN115601379A (en) Surface crack accurate detection technology based on digital image processing
Wang et al. Effective segmentation approach for solar photovoltaic panels in uneven illuminated color infrared images
CN112541478A (en) Insulator string stain detection method and system based on binocular camera
CN117058182A (en) Target detection frame optimization method for visual ranging
CN113807238B (en) Visual measurement method for river channel water surface floater area
CN114581407B (en) Self-adaptive defect detection method for photovoltaic module
CN111709887B (en) Image rain removing method based on sparse blind detection and image multiple feature restoration
CN111983710B (en) Non-uniformity correction method for scanning type infrared search system
CN114565564A (en) Fitting centroid sub-pixel positioning method based on correlation method threshold iteration
CN112651936A (en) Steel plate surface defect image segmentation method and system based on image local entropy
Yang et al. Estimating PV Soiling Loss Using Panel Images and a Feature-Based Regression Model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination