CN111242892B - Method for detecting defects of solar photovoltaic cells - Google Patents
Method for detecting defects of solar photovoltaic cells Download PDFInfo
- Publication number
- CN111242892B CN111242892B CN201911379608.9A CN201911379608A CN111242892B CN 111242892 B CN111242892 B CN 111242892B CN 201911379608 A CN201911379608 A CN 201911379608A CN 111242892 B CN111242892 B CN 111242892B
- Authority
- CN
- China
- Prior art keywords
- image
- solar photovoltaic
- photovoltaic cell
- threshold
- detected
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000001514 detection method Methods 0.000 claims abstract description 43
- 230000002159 abnormal effect Effects 0.000 claims abstract description 26
- 230000011218 segmentation Effects 0.000 claims abstract description 7
- 238000012217 deletion Methods 0.000 claims description 8
- 230000037430 deletion Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000002950 deficient Effects 0.000 claims description 5
- 238000001228 spectrum Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 2
- 230000002708 enhancing effect Effects 0.000 abstract 1
- 238000002474 experimental method Methods 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000001629 suppression Effects 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
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- 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
Abstract
The invention discloses a method for detecting defects of a solar photovoltaic cell. The method mainly comprises the steps of removing horizontal buses in a produced solar photovoltaic cell image according to Fourier transform and inverse transform of the solar photovoltaic cell image, enhancing brightness and contrast of the photovoltaic cell image with the buses removed through power transform, finding out all abnormal image blocks in the image through a block data deleting model, deleting all abnormal image blocks, constructing a background of the image through a nonlinear regression model by utilizing the rest image blocks which are closer to a real background, and finally separating a defect area through threshold segmentation from a result obtained by differentiating an image to be detected and the background image. The invention has the main advantages that the background constructed by the invention is closer to the real background, the detected defect type is more comprehensive, the detection result does not contain the bus of the solar photovoltaic cell, and the detection result is more accurate.
Description
Technical Field
The invention belongs to the technical field of industrial vision detection and image processing, and relates to a method for detecting defects of a solar photovoltaic cell based on a data deletion model.
Background
The existing defect detection method of the solar photovoltaic cell mainly comprises a feature extraction method and a background suppression method. The feature extraction method is a quick and effective method for directly extracting the defect features. However, this method is less flexible because it requires the defective feature to be set in advance. The background inhibition method is a solar photovoltaic cell defect detection method based on background reconstruction and difference, can effectively divide a defect area from an original background, and is one of main directions of current researches. However, in the solar photovoltaic cell defect detection method based on background reconstruction, most of the method is used for detecting hidden crack or broken gate type defects, and the detection effect on the defects of fragments, black cores and the like is not ideal, namely the types of defects which can be detected by the existing method are not comprehensive. Meanwhile, the defect detection method of the solar photovoltaic cell has the defects of time consumption and labor consumption in the detection process, unsatisfactory busbar removal effect of the photovoltaic cell and the like.
Disclosure of Invention
The invention aims to provide a method for detecting defects of a solar photovoltaic cell and a method for removing bus bars, so that the problem that the defect type detection of the solar photovoltaic cell is incomplete is effectively solved, and meanwhile, the influence of the bus bars on a detection result is solved.
The technical scheme adopted by the invention is that the method for detecting the defects of the solar photovoltaic cell comprises the following steps:
Step 4, calculating the image I to be detected 0 With background image
Obtaining a detection result I' through threshold segmentation;
and 5, removing noise points in the detection result I' to obtain a final defect detection result.
The invention is also characterized in that:
step 1.1, the original image I x Fourier transforming to obtain spectrogram
And in the spectrogram->
The upper border of the region with constant spectral value, which is represented by the spectrogram +.>
A circular area with d as a radius is used as a circle center, and the value range of d is more than 0 and less than or equal to 8;
step 1.2, constructing a filter V, and taking the bandwidth of the filter V as D u The parameter of the control bandwidth is w, and the value range of w is
The V (u, V) value with the coordinates of any point (u, V) is calculated by a formula (1), wherein D (u, V) is the distance from the point (u, V) on the spectrogram to the center of the spectrum, D (u, V) formula is shown as a formula (2),
step 1.3, spectrogram is displayed
Convolving with the constructed filter V to obtain a spectrogram of the solar photovoltaic cell image with the horizontal bus removed>
As shown in (3),
step 1.4, pair
Performing inverse Fourier transform to obtain a solar photovoltaic cell image I with horizontal buses removed e For image I e The brightness and contrast of the image I to be detected are enhanced by performing power conversion to obtain the image I to be detected 0 Power conversionThe formula is shown as (4),
wherein the parameter C is more than 0, and r is more than 0 and less than 1.
step 2.1, the image I to be detected 0 Dividing into a plurality of non-overlapping image blocks with W multiplied by H, taking the average value of all pixel gray values of each image block as the pixel value of the whole image block to obtain an image I 1 ；
Step 2.2, record I 1 Middle position (x) i ,x j ) The pixel value at is Y i,j Establishing a nonlinear regression model between the pixel values and the pixel positions thereof as shown in formula (5), wherein epsilon-N (0, sigma) 2 )；
The coefficients can be estimated according to the least square method using (5)
Step 2.3, slave image I 1 One image block is deleted, and the average value of the rest image blocks is used for estimating the coefficient through the formula (5)
Where l=1, 2, …, k, k is the number of tiles;
step 2.4, calculating coefficients
Sum coefficient->
Cook distance Cook of (C) l ，Cook l The calculation formula of (a) is shown as formula (6),
wherein p is the number of parameters, X is the coordinate value matrix,
estimating an error variance;
step 2.5, calculate Cook l Upper quartile Q of (l=1, 2, …, k) 3 And a quartile range R 1 Taking the cut-off point as a threshold T 1 Threshold T 1 The calculation formula of (C) is T 1 ＝Q 3 +1.5R 1 ；
Step 2.6, cook calculated in step 2.3 l And threshold T of step 2.5 1 Comparison, cook was screened out l Greater than threshold T 1 Marked as abnormal block to obtain image I c As in (7),
step 3.1, from the image I to be inspected 0 Removing all the abnormal blocks marked in the step 2;
step 3.2, utilizing a nonlinear regression model type (5) to remove the to-be-detected image I after the abnormal block is removed 0 Fitting pixel values and coordinates of the non-defective region to calculate coefficients
Step 3.3, the utilization coefficient is
Performing background image reconstruction by a nonlinear regression model (5) to obtain a reconstructed background image +.>
As in (8),
wherein (x) i ,x j ) Is the same as the image I to be detected 0 The same coordinates.
Step 4 is carried out according to the following steps:
step 4.1, the image I to be inspected according to step 2 of claim 3 0 And the background image of step 3 in claim 4
Calculating a differential image Δi as in formula (9):
step 4.2, taking the threshold T 2 ＝μ+c 2 Sigma, wherein mu and sigma are the mean and standard deviation, c, respectively, of the gray values of the differential image ΔI 2 Is constant, 0 < c 2 Less than or equal to 3, utilizing a threshold T 2 Segmentation of the differential image Δi may yield a detection result I', as shown in equation (10):
where (r, c) is the position coordinates of the pixel of the differential image Δi.
step 5.1, combining the detection result I' and the area of the 8-connected region, and setting a threshold T area ，T area The value range of (2) is more than 0 and less than T area ≤600；
Step 5.2, the area of the 8-connected region is set to be a threshold T area Comparing and marking that the area of the 8-connected region is greater than or equal to a threshold T area And (3) obtaining an abnormal region from which noise is removed, namely detecting defects of the solar photovoltaic cell.
In step 1.1, d=4, in step 1.4, c=1, r=0.7.
In step 4.2, c 2 ＝2。
In step 5.1, T area ＝100。
The beneficial effects of the invention are as follows: compared with the prior art, the method has the advantages that the Fourier transform is used for removing the bus in the solar photovoltaic cell, the bus part can be well removed, the part is displayed as a non-defective area, the detection results of other areas are not influenced, and the method is more accurate in removing compared with the traditional method. When the block data deleting model in regression diagnosis is used for background reconstruction, the reconstructed background is closer to the real background, so that the defect type can be detected more comprehensively by the method, and the detection efficiency is higher.
Drawings
FIG. 1 is a flow chart of a method for detecting defects of a solar photovoltaic cell according to the present invention;
FIG. 2 is a filter V constructed by the method for detecting defects of a solar photovoltaic cell of the present invention;
FIG. 3 is a graph showing the detection results of different types of defects by the method for detecting defects of a solar photovoltaic cell according to the present invention;
FIG. 4 shows a method for detecting defects of a solar photovoltaic cell according to the present invention with respect to a threshold T 2 Taking detection results of different values;
FIG. 5 shows the detection results of different block sizes according to the method for detecting defects of a solar photovoltaic cell;
Detailed Description
The invention will be described in detail below with reference to the drawings and the detailed description.
The invention discloses a method for detecting defects of a solar photovoltaic cell, which is shown in figure 1 and is specifically implemented according to the following steps:
step 1.1, as shown in FIG. 1, the original image I is x Fourier transforming to obtain spectrogram
And in the spectrogram->
Go up to the spectrogram->
D (d is more than 0 and less than or equal to 8) is used as a circle center, a circle domain is selected by taking d (d is more than 0 and less than or equal to 8) as a radius, and the frequency spectrum value in the circle domain is kept unchanged; at this time, the horizontal buses in the airspace are shown to be positioned near a vertical line in the center of the frequency spectrum in the frequency domain, and the high-frequency components are mainly used; d=4 is selected to achieve the best detection result;
step 1.2, as shown in FIG. 2, constructing a filter V according to the Fourier transform method to eliminate the buses, taking the bandwidth of the filter V as D u The parameter controlling the bandwidth is w,
the V (u, V) value with the coordinates of any point (u, V) is calculated by a formula (1), wherein D (u, V) is the distance from any point (u, V) on the spectrogram to the center of the spectrogram, the formula of D (u, V) is shown as a formula (2),
step 1.3, spectrogram is displayed
Convolving with the constructed filter V to obtain a spectrogram of the solar photovoltaic cell image with the horizontal bus removed>
As shown in (3),
step 1.4, pair
Performing inverse Fourier transform to obtain a solar photovoltaic cell image I with horizontal buses removed e . For image I e Performing power conversion to enhance the brightness and contrast of the image to obtain an image I to be detected 0 (see fig. 1: 2), the power transform formula (4),
wherein C is more than 0, and r is more than 0 and less than 1. To obtain the best detection result, c=1 and r=0.7 are selected
step 2.1, the image I to be detected 0 Dividing into a plurality of non-overlapping image blocks with W multiplied by H, taking the average value of all pixel gray values of each image block as the pixel value of the whole image block to obtain an image I 1 (as in fig. 1: 3);
step 2.2, record I 1 At (x) i ,x j ) The pixel value at is Y i,j A nonlinear regression model between pixel values and their pixel locations can be built as in equation (5), where ε -N (0, σ) 2 )；
The coefficients can be estimated according to the least square method using (5)
Step 2.3, slave image I 1 One image block is deleted, and the average value of the rest image blocks is used for estimating the coefficient through the formula (5)
Where l=1, 2, …, k, k is the number of tiles;
step 2.4, calculating coefficients
Sum coefficient->
Cook distance Cook of (C) l ，Cook l The calculation formula of (a) is shown as formula (6),
wherein p is the number of parameters, X is the coordinate value matrix,
estimating an error variance;
step 2.5, calculate Cook l Upper quartile Q of (l=1, 2, …, k) 3 And a quartile range R 1 Taking the threshold value of the cut-off point as T 1 T, i.e 1 ＝Q 3 +1.5R 1 ；
Step 2.6, cook calculated in step 2.4 l And threshold T in step 2.5 1 Comparison, cook was screened out l Greater than threshold T 1 Marked as abnormal block to obtain image I c (see FIG. 1: 4) of formula (7),
step 3.1, from the image I to be inspected 0 Removing all the abnormal blocks marked in the step 2;
step 3.2, utilizing a nonlinear regression model type (5) to remove the to-be-detected image I after the abnormal block is removed 0 Fitting pixel values and coordinates of the non-defective region to calculate coefficients
Step 3.3, the utilization coefficient is
Performing background image reconstruction by a nonlinear regression model (5) to obtain a reconstructed background image +.>
(see FIG. 1: 5), as in equation (8),
wherein (x) i ,x j ) Is the same as the image I to be detected 0 The same coordinates.
Step 4, calculating the image I to be detected 0 With background image
The detection result I' is obtained through threshold segmentation, and the method is specifically implemented according to the following steps:
step 4.1, the image I to be detected according to the step 2 0 And step 3 background image
A differential image Δi (as in fig. 1: 6) is calculated as in equation (9):
step 4.2, taking the threshold T 2 ＝μ+c 2 Sigma, wherein mu and sigma are the mean and standard deviation, c, respectively, of the gray values of the differential image ΔI 2 Is constant, 0 < c 2 Less than or equal to 3, utilizing a threshold T 2 Segmentation of the differential image Δi may yield a detection result I', as shown in equation (10):
wherein (r, c) is the position coordinates of the pixels of the differential image ΔI, c is selected 2 The detection result is best when=2.
step 5.1, combining the detection result I' and the area of the 8-connected region, and setting a threshold T area ， 0＜T area Less than or equal to 600; to achieve the best detection result, T is selected area ＝100；
Step 5.2, the area of the 8-connected region is smaller than the threshold T area Is regarded as a noise point, and the gray value of the region is set to be 1; the area of the 8-connected region is greater than or equal to a threshold T area The partial corresponding gray value of (2) is set to 0;
step 5.3, the region with gray level value of 0 is the abnormal region (as shown in fig. 1: 8) for removing noise, namely the defect of the solar photovoltaic cell.
In the method for detecting the defects of the solar photovoltaic cell, provided by the invention, the following steps are included: the function of the step 1 is to remove the bus of the solar photovoltaic cell, and the method is realized by adopting Fourier transform and inverse transform, the principle is that an image is transformed from a space domain to a frequency domain, horizontal buses in the photovoltaic cell are intensively distributed near a vertical line passing through the center of a frequency spectrum in the frequency domain, the value of the region is set to 0 according to the principle, and then the region corresponding to the buses is removed from the space domain according to the inverse transform of Fourier transform to the space domain, so that the defect detection result of the solar photovoltaic cell for removing the buses is more accurate, and the interference of the buses is eliminated.
The step 2 is used for finding out abnormal image blocks in the segmented image to the maximum extent, the principle is that the image blocks are removed one by one according to a block data deleting model, then corresponding coefficients are obtained through a nonlinear regression model according to gray values of the rest area and the whole image and corresponding coordinates of the rest area, the Cook distance between the coefficients before and after block deletion is compared, the abnormal coefficients and the corresponding image blocks are separated by using an upper cutoff point as a threshold value, and then the found abnormal blocks are completely deleted, and the pixels which are more close to the real background are utilized to reconstruct the background.
In order to illustrate the effectiveness of the invention in detection results and accuracy, subjective evaluation and objective evaluation are adopted to carry out detection performance analysis.
Subjective evaluation
As shown in FIG. 3, the method for detecting defects of the solar photovoltaic cell according to the invention respectively carries out experiments on the solar photovoltaic cell images containing hidden cracks, broken grids and other defect types, and the experimental results show that in the method for detecting defects of the solar photovoltaic cell according to the invention, besides the hidden cracks, the broken grids and other defects, such as black cores, fragments and other defects and stains can be detected, and the final detection result can not detect bus bars carried by the solar photovoltaic cell images, so that the method has good detection effect.
Objective evaluation
In another set of experiments we also evaluated a total of 313 solar photovoltaic cell images, of which 155 were detected manually as images containing defects such as hidden cracks, broken grids, etc., and the remaining 158 were normal images without defects. In the defect images, 150 defect images are accurately detected by the method, the defect detection rate reaches 96.77%, 5 undetected images are polycrystalline images with particularly serious defects, no defects are detected for 158 samples without defects, and the defects are basically consistent with the manual detection result.
For some of the parameters involved in the present invention, examples are given below to illustrate the rationality of parameter selection.
As shown in fig. 4, four horizontal graphs are threshold T in the image segmentation process 2 Embodiments taking different values, wherein the original image, threshold T, are sequentially from left to right 2 Medium parameter c 2 ＝1、c 2 =2 and c 2 As a result of=3, when the threshold T is found by comparison with the original image 2 The middle parameter is closest to the actual result when taken to be 2.
As shown in fig. 5, four graphs in the horizontal direction are examples of the results of experiments of different block sizes in the process of deleting block data, wherein the results of the original image, the block sizes of 10×10, 20×20, and 40×40 are sequentially shown from left to right. As can be seen from fig. 5 and table 1, for the 800×800 image used in the present experiment, the defect region is most accurately removed when the block size is 10×10, but the algorithm is also the longest, and when the block size is 40×40, the algorithm is less time-consuming but the defect region coverage is not accurate enough, so we finally select the block size of 20×20 to perform the following experiment, so that the detection result is the best while ensuring the faster running time of the algorithm.
Table 1: the 4 images in fig. 5 are detected for different tile sizes.
Claims (8)
1. The method for detecting the defects of the solar photovoltaic cells is characterized by comprising the following steps of:
step 1, removing an original image I of a solar photovoltaic cell x The bus in the image is improved in brightness and contrast ratio to obtain an image I to be detected 0 The original image I x And an image I to be inspected 0 The sizes are M multiplied by N;
step 1 is carried out according to the following steps:
step 1.1, the original image I x Fourier transforming to obtain spectrogram
And in the spectrogram->
The region in which the spectral value remains unchanged is delimited by a spectrogram +.>
Is centered at the center of (2)D is a circular area with a radius, and the value range of d is more than 0 and less than or equal to 8;
step 1.2, constructing a filter V, and taking the bandwidth of the filter V as D u The parameter of the control bandwidth is w, and the value range of w is
The V (u, V) value with the coordinates of any point (u, V) is calculated by a formula (1), wherein D (u, V) is the distance from the point (u, V) on the spectrogram to the center of the spectrum, D (u, V) formula is shown as a formula (2),
step 1.3, spectrogram is displayed
Convolving with the constructed filter V to obtain a spectrogram of the solar photovoltaic cell image with the horizontal bus removed>
As shown in (3),
step 1.4, pair
Performing inverse Fourier transform to obtain a solar photovoltaic cell image I with horizontal buses removed e For image I e The brightness and contrast of the image I to be detected are enhanced by performing power conversion to obtain the image I to be detected 0 The power conversion formula is shown as (4),
wherein the parameter C is more than 0, and r is more than 0 and less than 1;
step 2, constructing a block data deletion model, and utilizing the block data deletion model to detect the image I to be detected 0 Marking an abnormal block;
step 3, eliminating abnormal blocks in the images marked in the step 2, and carrying out background reconstruction by using the images with the abnormal blocks eliminated to obtain reconstructed background images
Step 4, calculating the image I to be detected 0 With background image
Obtaining a detection result I' through threshold segmentation;
and 5, removing noise points in the detection result I' to obtain a final defect detection result.
2. The method for detecting defects of a solar photovoltaic cell according to claim 1, wherein step 2 is performed according to the following steps:
step 2.1, the image I to be detected 0 Dividing into a plurality of non-overlapping image blocks with W multiplied by H, taking the average value of all pixel gray values of each image block as the pixel value of the whole image block to obtain an image I 1 ；
Step 2.2, record I 1 Middle position (x) i ,x j ) The pixel value at is Y i,j Establishing a nonlinear regression model between the pixel values and the pixel positions thereof as shown in formula (5), wherein epsilon-N (0, sigma) 2 )；
The coefficients can be estimated according to the least square method using (5)
Step 2.3, slave image I 1 One image block is deleted, and the average value of the rest image blocks is used for estimating the coefficient through the formula (5)
Where l=1, 2, …, k, k is the number of tiles;
step 2.4, calculating coefficients
Sum coefficient->
Cook distance Cook of (C) l ，Cook l The calculation formula of (a) is shown as formula (6),
wherein p is the number of parameters, X is the coordinate value matrix,
estimating an error variance;
step 2.5, calculate Cook l Upper quartile Q of (l=1, 2, …, k) 3 And a quartile range R 1 Taking the cut-off point as a threshold T 1 The threshold T 1 The calculation formula of (C) is T 1 ＝Q 3 +1.5R 1 ；
Step 2.6, cook calculated in step 2.3 l And threshold T of step 2.5 1 Comparison, cook was screened out l Greater than threshold T 1 Marked as abnormal block to obtain image I c As in (7),
3. a method for detecting defects in a solar photovoltaic cell according to claim 2, wherein step 3 is performed according to the following steps:
step 3.1, from the image I to be inspected 0 Removing all the abnormal blocks marked in the step 2;
step 3.2, utilizing a nonlinear regression model type (5) to remove the to-be-detected image I after the abnormal block is removed 0 Fitting pixel values and coordinates of the non-defective region to calculate coefficients
Step 3.3, the utilization coefficient is
Performing background image reconstruction by a nonlinear regression model (5) to obtain a reconstructed background image +.>
As in (8),
wherein (x) i ,x j ) Is the same as the image I to be detected 0 The same coordinates.
4. A method for detecting defects in a solar photovoltaic cell according to claim 3, wherein step 4 is performed according to the following steps:
step 4.1, the image I to be inspected according to step 2 of claim 2 0 And the background image of step 3 of claim 3
Calculating a differential image Δi as in formula (9):
step 4.2, taking the threshold T 2 ＝μ+c 2 Sigma, wherein mu and sigma are the mean and standard deviation, c, respectively, of the gray values of the differential image ΔI 2 Is constant, 0 < c 2 Less than or equal to 3, utilizing a threshold T 2 Segmentation of the differential image Δi may yield a detection result I', as shown in equation (10):
where (r, c) is the position coordinates of the pixel of the differential image Δi.
5. The method for detecting defects of a solar photovoltaic cell according to claim 4, wherein step 5 is performed according to the following steps:
step 5.1, combining the detection result I' and the area of the 8-connected region, and setting a threshold T area The T is area The value range of (2) is more than 0 and less than T area ≤600；
Step 5.2, the area of the 8-connected region is set to be a threshold T area Comparing and marking the area of the 8-connected region at the position of the mark to be more than or equal to a threshold value T area And (3) obtaining an abnormal region from which noise is removed, namely detecting defects of the solar photovoltaic cell.
6. The method of claim 1, wherein d=4 in step 1.1, c=1 and r=0.7 in step 1.4.
7. The method for detecting defects of a solar photovoltaic cell according to claim 4, wherein in step 4.2, c 2 ＝2。
8. The method for detecting defects of a solar photovoltaic cell according to claim 5, wherein in step 5.1, T is area ＝100。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911379608.9A CN111242892B (en) | 2019-12-27 | 2019-12-27 | Method for detecting defects of solar photovoltaic cells |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911379608.9A CN111242892B (en) | 2019-12-27 | 2019-12-27 | Method for detecting defects of solar photovoltaic cells |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111242892A CN111242892A (en) | 2020-06-05 |
| CN111242892B true CN111242892B (en) | 2023-06-27 |
Family
ID=70866319
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911379608.9A Active CN111242892B (en) | 2019-12-27 | 2019-12-27 | Method for detecting defects of solar photovoltaic cells |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111242892B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113834816A (en) * | 2021-09-30 | 2021-12-24 | 江西省通讯终端产业技术研究院有限公司 | Machine vision-based photovoltaic cell defect online detection method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011016420A1 (en) * | 2009-08-03 | 2011-02-10 | 株式会社エヌ・ピー・シー | Solar cell defect inspection apparatus, defect inspection method and program |
| WO2013093153A1 (en) * | 2011-12-21 | 2013-06-27 | Abengoa Solar New Technologies, S.A. | Method for the automated inspection of photovoltaic solar collectors installed in plants |
| CN106650770A (en) * | 2016-09-29 | 2017-05-10 | 南京大学 | Mura defect detection method based on sample learning and human visual characteristics |
| CN108306616A (en) * | 2018-01-11 | 2018-07-20 | 厦门科华恒盛股份有限公司 | A kind of photovoltaic module method for detecting abnormality, system and photovoltaic system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8586862B2 (en) * | 2009-11-18 | 2013-11-19 | Solar Wind Technologies, Inc. | Method of manufacturing photovoltaic cells, photovoltaic cells produced thereby and uses thereof |
-
2019
- 2019-12-27 CN CN201911379608.9A patent/CN111242892B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011016420A1 (en) * | 2009-08-03 | 2011-02-10 | 株式会社エヌ・ピー・シー | Solar cell defect inspection apparatus, defect inspection method and program |
| WO2013093153A1 (en) * | 2011-12-21 | 2013-06-27 | Abengoa Solar New Technologies, S.A. | Method for the automated inspection of photovoltaic solar collectors installed in plants |
| CN106650770A (en) * | 2016-09-29 | 2017-05-10 | 南京大学 | Mura defect detection method based on sample learning and human visual characteristics |
| CN108306616A (en) * | 2018-01-11 | 2018-07-20 | 厦门科华恒盛股份有限公司 | A kind of photovoltaic module method for detecting abnormality, system and photovoltaic system |
Non-Patent Citations (1)
| Title |
|---|
| 杨元培 ; 杨奕 ; 王建山 ; 张桂红 ; .光伏发电系统电池最大功率跟踪控制仿真.计算机仿真.2018,(06),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111242892A (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107909556B (en) | Video image rain removing method based on convolutional neural network | |
| CN105004972B (en) | Porcelain insulator Condition assessment of insulation method based on day blind ultraviolet imagery characteristics of image | |
| CN106780486B (en) | Steel plate surface defect image extraction method | |
| CN107845087A (en) | The detection method and system of the uneven defect of liquid crystal panel lightness | |
| CN111833312B (en) | Ultraviolet image diagnosis method and system for detecting discharge of fault insulator | |
| JP2006162583A (en) | Crack detection method | |
| CN109472788B (en) | Method for detecting flaw on surface of airplane rivet | |
| CN113313677B (en) | Quality detection method for X-ray image of wound lithium battery | |
| CN111598897B (en) | Infrared image segmentation method based on Otsu and improved Bernsen | |
| CN111242892B (en) | Method for detecting defects of solar photovoltaic cells | |
| CN115825410B (en) | Method for estimating action noise level of tire road surface based on road surface structure | |
| CN112381759B (en) | Monocrystalline silicon solar wafer defect detection method based on optical flow method and confidence coefficient method | |
| CN114332081B (en) | Textile surface abnormity determination method based on image processing | |
| CN110363749B (en) | Evaluation method for rusting degree of vibration damper based on image processing | |
| CN115601379A (en) | Surface crack accurate detection technology based on digital image processing | |
| CN110533626B (en) | All-weather water quality identification method | |
| CN101477025A (en) | Fast evaluation method for collection exhibition materials based on image processing | |
| CN113096103A (en) | Intelligent smoke image sensing method for emptying torch | |
| CN105787955A (en) | Sparse segmentation method and device of strip steel defect | |
| Nguyen et al. | A novel automatic concrete surface crack identification using isotropic undecimated wavelet transform | |
| CN109584224B (en) | Method for analyzing and displaying X-ray image of casting | |
| CN113436216B (en) | Electrical equipment infrared image edge detection method based on Canny operator | |
| CN114166850B (en) | Light excitation infrared thermal imaging defect detection method based on differential tensor decomposition | |
| CN111091601B (en) | PM2.5 index estimation method for real-time daytime outdoor mobile phone image | |
| CN111709887A (en) | Image rain removing method based on sparse blind detection and image multiple feature restoration |
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 |