Method and apparatus for detecting apparent defects of solar cells

Abstract

The invention discloses a method for detecting appearance defects of a solar cell. The method comprises the following steps: receiving a first image of the solar cell shot by a first camera under a first light source; receiving a second image of the solar cell shot by the first camera under a second light source, wherein a first angle between the first light source and the solar cell is smaller than a second angle between the second light source and the solar cell; identifying a first defect of the solar cell in the first image and a second defect of the solar cell in the second image; and determining the appearance grade of the solar cell based on the first defect and the second defect.

Description

Technical Field

[0001] This application relates to the field of solar cells, and more particularly, to a method and device for detecting appearance defects of solar cells. Background Art

[0002] A solar cell is a new energy battery that can convert light energy into electrical energy through the photovoltaic effect. It has developed from the first-generation solar cells of single-crystalline silicon or polycrystalline silicon to the current third-generation solar cells, with a significant increase in energy conversion efficiency. A typical third-generation solar cell includes a perovskite solar cell, which uses an organometallic halide semiconductor of the perovskite type as a light-absorbing material. When receiving sunlight irradiation, it can convert the received photon energy into photocurrent.

[0003] Perovskite solar cells have a very wide range of application fields, including photovoltaics, new energy vehicles, wearable devices, indoor electronic devices, the Internet of Things, remote control devices, etc. Whether there are defects in the appearance of perovskite solar cells will affect the appearance grade of the final product and the photoelectric conversion performance of the product. In addition, since ordinary consumers can also come into close contact with perovskite solar cells, the requirements for the appearance of perovskite solar cells are also increasing day by day. The traditional method mainly judges the appearance of perovskite solar cell products through manual visual inspection. This method has problems such as low efficiency, inconsistent standards, easy misjudgment and missed detection, and secondary pollution.

[0004] Therefore, it is necessary to provide a method and device for detecting appearance defects of solar cells to overcome the defects of the above-mentioned manual visual inspection. Summary of the Invention

[0005] An object of this application is to provide a method and device for detecting appearance defects of solar cells, which can quickly sort the appearance grades of solar cells, reduce misjudgment and missed detection caused by manual visual inspection, ensure the consistency of detection standards, and avoid secondary pollution that may be caused to the surface of solar cells by manual inspection.

[0006] In one aspect of this application, a method for detecting appearance defects of solar cells is provided. The method includes: receiving a first image of the solar cell captured by a first camera under a first light source; receiving a second image of the solar cell captured by the first camera under a second light source, where a first angle between the first light source and the solar cell is less than a second angle between the second light source and the solar cell; identifying a first defect of the solar cell in the first image and a second defect of the solar cell in the second image; and determining the appearance grade of the solar cell based on the first defect and the second defect.

[0007] In another aspect of the present application, there is also provided a device for detecting appearance defects of a solar cell. The device includes a non-transitory computer storage medium on which one or more executable instructions are stored. After being executed by a processor, the one or more executable instructions perform the following steps: receiving a first image of the solar cell taken by a first camera under a first light source; receiving a second image of the solar cell taken by the first camera under a second light source, wherein a first angle between the first light source and the solar cell is less than a second angle between the second light source and the solar cell; identifying a first defect of the solar cell in the first image and a second defect of the solar cell in the second image; and determining an appearance grade of the solar cell based on the first defect and the second defect.

[0008] The above is an overview of the present application. There may be simplifications, generalizations, and omissions of details. Therefore, those skilled in the art should recognize that this part is only illustrative and is not intended to limit the scope of the present application in any way. This overview section is neither intended to identify the key features or essential features of the claimed subject matter nor intended to be used as an aid in determining the scope of the claimed subject matter. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and other features of the present application will be more fully and clearly understood by referring to the following description, the appended claims, and the accompanying drawings. It can be understood that these drawings only depict several embodiments of the present application and should not be considered as limiting the scope of the present application. By using the drawings, the present application will be described more clearly and in detail.

[0010] Figure 1 and Figure 2 respectively show a cross-sectional view and an exploded view of a solar cell 10 according to an embodiment of the present application;

[0011] Figure 3 shows Figure 1 and Figure 2 a top view of the solar cell 10 shown in

[0012] Figure 4 shows a device 100 for detecting appearance defects of a solar cell 10 according to an embodiment of the present application;

[0013] Figure 5 shows Figure 4 a loading module 1000 of the device 100 shown in

[0014] Figure 6 shows Figure 5 a tray 200 of the loading module shown in

[0015] Figure 7 shows the first detection platform 2004 on the first detection module 2000 of the device 100 shown; Figure 4 shown;

[0016] Figure 8 shows the first camera module 2010 on the first detection module 2000 of the device 100 shown; Figure 4 shown;

[0017] Figure 9 shows Figure 8 a side view of the first camera module 2010 shown;

[0018] Figure 10 shows Figure 8 and Figure 9 a schematic diagram of the detection optical path and light source of the first camera module 2010 shown;

[0019] Figure 11a and Figure 11b respectively show a photo 300 of the solar cell 10 obtained by the first line scan and a photo 400 of the solar cell 10 obtained by the second line scan according to an embodiment of the present application;

[0020] Figure 12 shows Figure 4 the second detection module 3000 of the device 100 shown;

[0021] Figure 13 shows an exemplary photo 500 of the solar cell 10 obtained by area scan according to an embodiment of the present application;

[0022] Figure 14 Exemplarily shows the threshold setting and corresponding appearance grades of dot impurities, bubbles, linear scratches, chipping, corner breakage, and edge burr defects;

[0023] Figure 15 shows Figure 4 the inkjet component 4010 in the inkjet module 4000 of the device 100 shown; and

[0024] Figure 16 shows Figure 4 the material receiving module 5000 of the device 100 shown.

[0025] Before explaining in detail any embodiment of the present invention, it should be understood that the application of the present invention is not limited to the details of the construction and the arrangement of components set forth in the following description or shown in the following drawings. The present invention is capable of other embodiments and of being practiced or carried out in various ways. Moreover, it should be understood that the language and terminology used herein are for the purpose of description and should not be regarded as limiting. Detailed implementation manners

[0026] In the following detailed description, reference is made to the accompanying drawings which form a part hereof. In the drawings, like reference numerals generally represent like components unless the context dictates otherwise. The illustrative embodiments described in the detailed description, the drawings, and the claims are not intended to be limiting. Other embodiments may be utilized and other changes may be made without departing from the spirit or scope of the subject matter of this application. It will be understood that various configurations, substitutions, combinations, and designs of the aspects of the subject matter of this application generally described herein and illustrated in the drawings can be made, all of which are expressly part of the subject matter of this application.

[0027] Figure 1 and Figure 2 respectively show a cross-sectional view and an exploded view of a solar cell 10 according to an embodiment of the present application. As Figure 1 and Figure 2 shown, the solar cell 10 includes: an absorption layer 12, which may be, for example, a perovskite-type light-absorbing material layer for absorbing light and generating carrier pairs; a first conductive layer 14a and a second conductive layer 14b located on both sides of the absorption layer 12. The first conductive layer 14a extracts electrons from the absorption layer 12 via an electron transport layer (not shown) located between the first conductive layer 14a and the absorption layer 12, while the second conductive layer 14b extracts holes from the absorption layer 12 via a hole transport layer (not shown) located between the second conductive layer 14b and the absorption layer 12. The first conductive layer 14b and the second conductive layer 14b are respectively connected to an external circuit to transfer the current formed by photoelectric conversion to the outside. In one embodiment, the solar cell 10 further includes a transparent glass substrate 18a, and the foregoing first conductive layer 14a, electron transport layer, absorption layer 12, hole transport layer, and second conductive layer 14b are all formed on the glass substrate 18a by, for example, coating, deposition, or other suitable methods.

[0028] Further referring to Figure 1 and Figure 2 , the solar cell 10 further includes a transparent glass cover plate 18b, on which an encapsulation glue 16 may be coated to allow the glass cover plate 18b to be bonded to the first conductive layer 14a. The glass substrate 18a and the glass cover plate 18b together encapsulate the absorption layer 12 and the like therein.

[0029] Figure 3 shows Figure 1 and Figure 2 the top view of the solar cell 10 shown in Figure 3 shown, the solar cell 10 has a generally rectangular shape. Although Figure 3Exemplarily, a generally rectangular shape is shown, but those skilled in the art can understand that the solar cell 10 can be selected in other shapes as needed, such as square, oval, racetrack type, etc.

[0030] Continuing to refer to Figure 3 , the glass substrate (not shown in the figure, the substrate is on the side closest to the observer in Figure 3 , and its specific structure can be referred to Figure 1 and Figure 2 ) forms the overall outer contour of the solar cell 10, which has a length A5 and a width B3. Preferably, the absorption layer 12 is located in the exact middle of the glass substrate and is preferably arranged substantially symmetrically. There are distances A2 and (A5 - A3) between each long side of the absorption layer 12 and the corresponding long side of the glass substrate, while there are distances B1 and (B3 - B2) between each short side of the absorption layer 12 and the short side of the substrate.

[0031] The electrode region 14 exposed between the substrate and the cover plate (indicated by diagonal lines in Figure 3 ) is used for connection to an external circuit. Figure 3 Exemplarily shown in

[0032] both sides of the solar cell 10 have electrode regions 14. The length of the electrode region 14 is the same as the width B3 of the substrate, and its width is A1 and (A5 - A4). It can be understood that in the solar cell 10, there are distances (A2 - A1) and A6 (i.e., A4 - A3) between the absorption layer 12 and the electrode region 14 in the length direction. It can be understood that for a solar cell 10 with excellent appearance grade, the above-mentioned dimensions A1 - A6 and B1 - B3 should all be consistent with the corresponding dimensions of the standard sample to ensure the consistency of the product appearance dimensions. If the dimensions are inconsistent, it may affect the battery output characteristics of the solar cell and may also affect its connection to the external circuit, etc.

[0033] Generally speaking, the above-mentioned size of the solar cell does not meet the requirements (appearance defects related to size) and other process defects can be identified by inspecting the appearance of the solar cell without complex circuit detection. Therefore, the inventors of the present application conceived an automated detection device for detecting the appearance defects of solar cells to implement the inspection of solar cells.

[0034] Figure 4 FIG. 4 shows a device 100 for detecting appearance defects of a solar cell 10 according to an embodiment of the present application. The detection device 100 can detect, for example, Figures 1 to 3 the solar cell shown, and preferably, can detect a plurality of solar cells in batches.

[0035] As Figure 4 shown, the device 100 includes five modules from upstream to downstream, namely a loading module 1000, a first detection module 2000, a second detection module 3000, a coding module 4000, and a receiving module 5000. Although Figure 4 the above five modules are exemplarily shown as being divided into five modules, those skilled in the art can understand that two or more of the above five modules can be combined into one module, and each module may be divided into multiple sub-modules; and, according to different embodiments, the order of the first detection module 2000 and the second detection module 3000 can also be replaced.

[0036] Figure 5 FIG. 18 shows Figure 4 the loading module 1000 of the device 100 shown in FIG. 17. The loading module 1000 includes a loading bin 1002 for storing one or more pallets 200. Figure 6 FIG. 20 shows Figure 4 the pallet 200 of the device 100 shown in FIG. 21, which has one or more grooves 202 for carrying one or more solar cells 10 to be detected. Continuing to refer to Figure 5, the loading bin 1002 further includes a frame 1004 and a driving mechanism 1006 installed on the upper surface of the frame 1004. A tray carriage 1008 is installed on the driving mechanism 1006; driven by the driving mechanism 1006, the tray carriage 1008 can move the standard tray 200 from a certain position in the loading bin 1002 to the position to be transferred. The loading module 1000 further includes a first handling mechanism 1010, which can handle the solar cells 10 on the tray 200 from the loading module 1000 to the first detection module 2000. The first handling mechanism 1010 includes a first clamp 1012, and the first clamp 1012 can be driven to move in the horizontal direction to align with the solar cells 10 to be clamped in the tray 200. When aligned, the first clamp 1012 is driven to move in the vertical direction to clamp the corresponding solar cells 10 in the tray 200.

[0037] Figure 7 Shows the location at Figure 4 The first detection platform 2004 on the first detection module 2000 of the device 100 shown. As described above, after the first clamp 1012 clamps the solar cells 10, it transfers the solar cells 10 to the first detection platform 2004 of the first detection module 2000 for detection on the first detection module 2000.

[0038] Figure 8 Shows the location at Figure 4 The first camera module 2010 on the first detection module 2000 of the device 100 shown. As Figure 8 shown, the first camera module 2010 is located directly above the first detection platform 2004 and is configured to be able to move back and forth through the camera carriage 2012. The first camera module 2010 includes a first camera 2020, and when the first camera module 2010 moves past the first detection platform 2004, the first camera 2020 can take pictures of the solar cells 10 located on the first detection platform 2004. In one embodiment, the first camera 2020 can be a line scan camera with a pixel accuracy of 0.02mm / pixel and 8K color. Those skilled in the art can understand that the first camera 2020 can use other specifications of line scan cameras.

[0039] Figure 9 Shows Figure 8 The side view of the first camera module 2010 shown. As Figure 8 and 9As shown, the first camera module 2010 further includes a pair of first light sources 2022 arranged on both sides of the vertical detection optical path 2021 of the first camera 2020, and a second light source 2024 located on one side of the vertical detection optical path 2021. During the forward and backward movement of the first camera module 2010 through the camera carriage 2012, the first light sources 2022 and the second light source 2024 move together with the first camera 2020, thereby providing additional light for the shooting of the camera 2020 and improving the imaging effect. In one embodiment, the first light sources 2022 and the second light source 2024 are strip-shaped LED light sources.

[0040] Figure 10 As shown Figure 8 and Figure 9 a schematic diagram of the detection optical path and light sources of the first camera module 2010 shown in the figure. As Figure 10 shown in the figure, when observed in a direction perpendicular to the movement direction of the first camera module 2010, a pair of first light sources 2022 form a first angle α with the plane of the solar cell 10, and the second light source 2024 forms a second angle β with the plane of the solar cell 10. In one embodiment, the first angle α can be 40° to 50°, which is a relatively small reflection angle, and the second angle β can be 75° to 85°, which is a relatively large reflection angle compared to the first angle.

[0041] In one embodiment, the first camera module 2010 performs two line scans on the solar cell 10 on the first detection platform 2004 to obtain two photos of the solar cell 10. During the first line scan, the first light sources 2022 are turned on and the second light source 2024 is turned off; and during the second line scan, the second light source 2024 is turned on and the first light sources 2022 are turned off. It can be understood that when the solar cell 10 is irradiated with different light sources, the presentation of the appearance details (including defects) of the solar cell 10 is also different.

[0042] Figure 11a and Figure 11b respectively show an exemplary photo 300 of the solar cell 10 obtained through the first line scan and an exemplary photo 400 of the solar cell 10 obtained through the second line scan according to an embodiment of the present application. As Figure 11aAs shown, during the first line scan, a pair of first light sources 2022 irradiate the surface of the solar cell 10 at opposite angles. Therefore, the photo 300 taken by the camera 10 has less shadow and higher brightness, and can present the details of the solar cell 10 more clearly. Those skilled in the art can understand that when light shines on a shiny object, the reflected light will appear in two forms: specular reflection and diffuse reflection. By avoiding the specular reflection light and detecting the diffuse reflection light, the camera can capture a more realistic color. Designing the first light source 2022 to have an angle of 40° to 50° can improve the color accuracy. Correspondingly, the photo 300 can be used for detecting the color uniformity and color difference of the absorption layer 12 of the solar cell 10, detecting bubbles or foreign objects in the solar cell 10, etc.

[0043] As Figure 11b shown, during the second line scan, since the second light source 2024 irradiates the surface of the solar cell 10 at a relatively high angle, the photo 400 obtained by the camera 2020 can better show the surface defects of the solar cell 10. Those skilled in the art can understand that when a high-angle light source irradiates the surface of the object to be measured, the flat area can smoothly reflect light, making the image look brighter and clearer; while the uneven part of the surface (such as scratches, etc.) reflects light poorly, so it is easier to be detected compared to the flat part. Correspondingly, the photo 400 can be better used to detect scratches and foreign objects (such as dirt, fingerprints, and oil stains) on the surface of the solar cell 10. Therefore, based on the irradiation of the light source at two different angles, the device 100 can achieve multi-level appearance defect detection of the solar cell, thereby improving the detection accuracy. It should be noted that in this embodiment, the first light source and the second light source are two separate light sources, but in some other embodiments, these two light sources can also be integrated into one light source; multi-angle irradiation and detection can be achieved by rotating the integrated light source or rotating the solar cell 10 (such as through its tray).

[0044] Figure 12 shows Figure 4 the second detection module 3000 of the device 100 shown. As Figure 12 shown, the second detection module 3000 includes a second camera 3002 and a second detection platform 3004 located below the second camera 3002. After the solar cell 10 completes two line scans in the first detection module 2000, the second handling mechanism 3050 (see Figure 4)Transfer the solar cell 10 located on the first detection platform 2004 to the second detection platform 3002. In one embodiment, the second camera 3002 is an area array camera, such as a black and white area array camera with a pixel accuracy of 0.017 mm / pixel and 150 million pixels. Those skilled in the art can understand that the black and white area array camera has the advantages of high speed and high resolution, and has great advantages in high-precision edge recognition. However, an area array camera with a suitable resolution can also be used as the second camera 3002.

[0045] Continue to refer to Figure 12 , the second detection module 3000 further includes a third light source 3006 located on the side of the second detection platform 3002 close to the second camera 3002 and a fourth light source 3008 located on the side of the second detection platform 3002 far from the second camera 3002. In one embodiment, the third light source 3006 can be composed of four strip light sources, and the angle of each strip light source is adjustable. Similar to the aforementioned first light source, the angle of the third light source 3006 relative to the second detection platform 3002 can be 40° to 50°. In one embodiment, the fourth light source 3008 is a white backlight source and the second detection platform 3004 is transparent. The fourth light source 3008 irradiates the back surface of the solar cell 10 located on the second detection platform 3002. Therefore, the second camera 3002 can clearly capture the edges of the main structural regions of the solar cell 10 under the cooperation of the third light source 3006 and the fourth light source 3008.

[0046] Figure 13 Shows an exemplary photograph 500 of the solar cell 10 obtained by area scanning according to an embodiment of the present application. Due to the transparency of the substrate and the semi-transparency of the encapsulation glue, the boundary of the substrate of the solar cell 10, the boundary between the absorption layer and the encapsulation glue, the boundary between the absorption layers, and the boundary between the electrode region and the encapsulation glue can be clearly identified from Figure 13 the photograph 500. Referring to Figure 2 , the above-mentioned boundaries identified from Figure 13 the photograph 500 basically correspond to Figure 2The boundaries depicted in each part. Therefore, the captured photo 500 can be used to detect whether the corresponding values of A1 - A6 and B1 - B3 of the solar cell 10 are the same as those of the standard sample, thereby determining the accuracy of the size of the solar cell 10. The accuracy of the size referred to in this application includes the accuracy of the sizes of different parts of the solar cell and the accuracy of the relative positions with respect to other parts. It can be understood that in some embodiments, alignment marks or standard sizes can also be set on the tray for placing the solar cell, which can be located in the peripheral area of the solar cell, so that the alignment marks and standard sizes can be imaged synchronously when imaging the solar cell, facilitating the measurement of the values of A1 - A6 and B1 - B3.

[0047] In one embodiment, since the photo 500 clearly captures the edge of the solar cell 10, by comparing with the standard sample, it can be determined whether there are any missing corners, chipped edges, or edge burrs on the solar cell 10. In one embodiment, the aforementioned photos 300 and 400 captured by the first camera can also be used to determine whether there are any missing corners, chipped edges, or edge burrs on the solar cell 10.

[0048] After the first detection module 2000 and the second detection module 3000 have captured the three photos 300, 400, and 500 of the solar cell 10, the vision detection processing software running on the device 100 or the vision detection processing software (not shown) running on a computer device coupled to the device 100 can identify various defects of the solar 10 based on the photos 300, 400, and 500, and further determine what different appearance grades the solar cell 10 belongs to.

[0049] In one embodiment, the user can first identify and classify various defects of the solar cell 10 and construct a defect classification data set. Further, the user can use deep learning tools to train a neural network model based on the constructed data set to obtain a trained vision detection processing software. In one embodiment, for example, 100, 1000, or 5000 samples can be provided for each type of defect. Those skilled in the art understand that the more the number of samples, the more accurate the detection results of the trained model. Finally, the user can use the trained vision detection processing software to identify various defects of the solar cell 10 in the captured photo and determine the appearance grade of the solar cell 10.

[0050] In one embodiment, for the detection of the color difference of the absorption layer 12 of the solar cell 10, the visual inspection processing software first converts the photo 300 from the RGB color channel to the Lab color channel. Wherein, L represents the lightness of the color; a is a positive value indicating a reddish color, and a negative value indicating a greenish color; b is a positive value indicating a yellowish color, and a negative value indicating a bluish color. The visual inspection processing software is able to identify the effective area (e.g., the absorption layer 12) and calculate the corresponding Lab parameter value of the effective area. In one embodiment, the user can provide a standard sample, and the visual inspection processing software is able to detect the color difference ΔE between the current solar cell 10 and the standard sample (refer to the following equation 1).

[0051] △E=sqrt((LL 标 )2+(aa 标 )2+(bb 标 )2)(Equation 1)

[0052] In one embodiment, for the detection of color uniformity of the absorber layer 12 of the solar cell 10, the visual inspection processing software first converts the photo 300 from RGB channels to HSV channels. Wherein, H represents chromaticity; S represents saturation; and V represents brightness. The visual inspection processing software can identify the active area (e.g., the absorber layer 12) and calculate the overall average variance ΔH of the H channel of the active area (see equation 2 below).

[0053]

[0054] Among them, n represents the number of valid pixels, xi represents the gray value of each pixel in the H channel, Represents the average grayscale of the H channel. The lower the △H, the more uniform the color, and the higher the △H, the more uneven the color.

[0055] Therefore, through the thresholds of ΔE and ΔH set in advance by the user, the visual inspection processing software can determine whether the color of the absorption layer 12 of the solar cell 10 has color difference (for example, when the calculated color difference ΔE is greater than the threshold, it is determined that there is color difference), and whether the color is uniform (for example, when the calculated overall average variance ΔH of the color is greater than the threshold, it is determined that there is color difference), and then determine the appearance grade of the solar cell 10. In one embodiment, the visual inspection processing software can identify the color difference between different sheets in the absorption layer 12 (for example, exemplarily composed of 5 sheets) or the color difference of the area within the sheet, and determine whether the color difference between multiple sheets is too large, or determine whether the number of sheets with excessive color difference is too large. In one embodiment, the visual inspection processing software can detect the area or size of color difference and color unevenness, and determine the appearance grade of the solar cell based on the specifications of the specific area or size set by the user.

[0056] In one embodiment, the visual detection processing software converts photos 300, 400, and 500 from the RGB channel to the HSV channel, and segments the images based on the gray values of each pixel point in the H channel and extracts suspected defect regions. In one embodiment, the trained neural network model in the visual detection processing software can screen the extracted suspected defect regions, extract valid defects, and eliminate invalid defects. For example, the gray value of the suspected defect region can be compared with the gray value of the background, and when the difference between the two gray values is greater than 10, it is determined as a defect region. In one embodiment, the visual detection processing software can determine the size (e.g., length, width, height) and area of the defect, as well as the shape of the defect (e.g., roundness, straightness, etc.). For example, scratches are generally slender strips, bubbles are generally circular, and missing corners are generally irregular in shape. Therefore, through the neural network model trained with the dataset, the visual detection processing software can classify the identified defects into appropriate types.

[0057] In one embodiment, the visual detection processing software can locate the positions of the solar cells 10 in photos 300, 400, and 500, and construct the contour coordinates of the solar cells 10. In one embodiment, the user can provide a standard sample, and the visual detection processing software can compare the contour of the solar cell 10 with the contour of the standard sample to identify missing corners, chipped edges, or edge burrs of the solar cell 10.

[0058] In one embodiment, the user can also set in advance the judgment criteria for the appearance grade of the solar cell, and the judgment criteria can be integrated in the device 100. Figure 14 Exemplarily shown are the threshold settings and corresponding appearance grades for point impurities, bubbles, linear scratches, chipped edges, missing corners, and edge burr defects. As Figure 14 shown, for point impurities, bubbles, and linear scratches, D represents diameter, W represents width, L represents length, DS represents spacing, and N represents quantity; for chipped edges, missing corners, and edge burrs, X and Y respectively represent the dimensions in the figure. When the corresponding values detected by the visual detection processing software fall within the corresponding threshold ranges, the visual detection processing software can determine the corresponding appearance grade of the solar cell 10. For example, the highest appearance grade can be grade S, a better appearance grade can be grade A, and an unqualified appearance grade can be grade NG. Although Figure 14 exemplarily shown are the judgment criteria for the appearance grades of different defect types, those skilled in the art understand that the user can set other defect types and other threshold criteria according to requirements.

[0059] As described above, photo 300 is more likely to extract defects inside the substrate such as air bubbles, photo 400 is more likely to extract defects on the surface of the substrate such as scratches and dirt, and photo 500 is more likely to extract defects on the contour edge and can assist in extracting defects inside the substrate such as air bubbles. Therefore, through the information provided by the three photos, the visual inspection processing software can comprehensively judge the appearance defects of the solar cell 10.

[0060] In one embodiment, the user can use the second detection module 3000 to photograph the standard sample in advance to obtain an image of the standard sample, and calibrate the dimensions A1 - A6 and B1 - B3 of the standard sample as described above. In one embodiment, the user can set the parameters of the measurement line (e.g., edge detection intensity, width of the measurement line), and draw the pre - set measurement line. The visual recognition software can select the measurement points in photo 500 based on the measurement line pre - set by the user, and perform a two - point measurement on the corresponding measurement points of the solar cell 10 in photo 500. Finally, the visual recognition software can compare the measurement line recorded in the solar cell 10 in photo 500 with the measurement line of the standard sample to obtain the relative position and relative direction between the two, and compare the relative position and relative direction with the threshold value to judge whether the size of the solar cell 10 meets the requirements.

[0061] In one embodiment, the visual inspection processing software can record defect photos of the detected solar cells 10, and record the defect types, defect locations, defect sizes, etc. of the defect photos to construct a historical data set. On the one hand, the historical data set can be viewed by users, and on the other hand, it can be processed by analysis software to obtain a statistical analysis of the appearance of the batch of solar cells. The analysis software can be embedded in the visual inspection processing software or can be another independent software. The analysis software can count the yield rate of the solar cells, the discovery ratio of different defect types, the discovery location, the deviation of the size, etc. The data statistics provided by the analysis software can help the manufacturers of solar cells optimize the production process and quality control, etc. In particular, since the detection device of the present application can automatically detect solar cells, it can be integrated into the production line of solar cells to be used for real-time online monitoring of the solar cells produced in the production line, and can feedback the statistical data of the defects of the monitored solar cells to the production line, especially different types of defects. The production line can be provided with a main control device, which can judge the processes or steps that may cause defects in the production line based on the received defect feedback information, and accordingly adjust or replace these processes or steps. For example, some processes or steps that are prone to defects can be equipped with a replacement module for replacement when the currently running module needs to be repaired, so as to keep the entire production line running continuously. It can be understood that the main control device can be integrated with a neural network model or a similar artificial intelligence engine, which can be trained to automatically judge the processes or steps that cause a certain or certain types of defects.

[0062] Back to Figure 4 , after the solar cell 10 finishes shooting in the second detection module 3000, the third handling mechanism 4050 transports the detected solar cell 10 to the coding table 4002 of the coding module 4000. It can be understood that when the solar cell 10 is transported to the coding table 4002, the aforementioned visual recognition software has completed the recognition and judgment of the color, defects and size of the solar cell 10, and provided the appearance grade of the solar cell 10 according to the user's predetermined standard.

[0063] Figure 15 Shows Figure 4The inkjet component 4010 in the inkjet module 4000 of the device 100 shown. The inkjet component 4010 is configured to perform inkjet on the solar cells located on the inkjet table 4002. The inkjet component 4010 includes a support plate 4012, a sliding lead screw 4014 disposed on the support plate 4012, and an inkjet carriage 4016 that can slide on the sliding lead screw 4014. An inkjet head 4018 and a code reader 4020 are provided on the inkjet carriage 4016. As described above, after the trained module of the device 100 determines the appearance grade of the solar cell 10, it instructs the inkjet component 4100 to spray information including the grade and code of the solar cell 10 on the solar cell 2010. In one embodiment, the sprayed content can be a two-dimensional code, a bar code, or other identifiers capable of containing information. After the solar cell 10 is placed on the inkjet table 4002, the inkjet head 4018 is driven to the working position to perform inkjet on the solar cell 10. After the inkjet is completed, the code reader 4020 is driven to this working position to read the code of the solar cell 10, so as to verify whether the inkjet of the solar cell 10 is completed and whether the inkjet information is accurate.

[0064] Figure 16 is shown Figure 4 The receiving module 5000 of the device 100 shown is shown. After the code reader 4020 completes the code reading, the fourth handling mechanism 5002 of the receiving module 5000 transports the solar cell 10 to the corresponding tray 200 located on the receiving module 5000 according to the appearance grade of the solar cell 10. After the tray 200 is filled with solar cells 10, the tray 200 is moved to the receiving bin corresponding to the corresponding appearance grade. Figure 16 Exemplarily, receiving bins 5010, 5020, and 5030 for three appearance grades are shown to accommodate solar cells 10 corresponding to S grade, A grade, and NG grade. In one embodiment, when the inkjet printer 4020 fails to complete the code reading, the fourth handling mechanism 5002 transports the solar cell to a temporary storage box for detecting abnormal products (not shown in the figure) for subsequent manual inspection.

[0065] Therefore, the detection device disclosed in the present application can complete the rapid sorting of the appearance grades of solar cells, reduce misjudgment and missed detection caused by manual visual inspection, ensure the consistency of detection standards, and avoid secondary pollution to the glass sheets that may be caused by manual inspection. In addition, the detection method disclosed in the present application can not only detect defects such as bubbles, impurities, scratches, foreign objects, and missing corners of solar cells, but also detect color difference and color uniformity, as well as the accuracy of dimensions. Therefore, it can more comprehensively judge the appearance grade of solar cells.

[0066] Embodiments of the present invention can be implemented through hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated designed hardware. Those of ordinary skill in the art can understand that the above devices and methods can be implemented using computer-executable instructions and / or included in processor control code, such as provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above hardware circuits and software such as firmware.

[0067] It should be noted that although several steps or modules of the methods and devices for detecting appearance defects of solar cells are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more of the above-mentioned modules can be embodied in one module. Conversely, the features and functions of one module described above can be further divided and embodied by multiple modules.

[0068] Those of ordinary skill in the art can understand and implement other changes to the disclosed embodiments by studying the specification, the disclosed content, the drawings, and the appended claims. In the claims, the term "comprising" does not exclude other elements and steps, and the terms "a" and "an" do not exclude a plurality. In the actual application of the present application, a part may perform the functions of multiple technical features recited in the claims. Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A method for detecting appearance defects of a solar cell, characterized in that: The method comprises: Receiving a first image of a solar cell captured by a first camera under a first light source; receiving a second image of the solar cell captured by the first camera under a second light source, wherein a first angle between the first light source and the solar cell is smaller than a second angle between the second light source and the solar cell; identifying a first defect of the solar cell in the first image and a second defect of the solar cell in the second image; and An appearance grade of the solar cell is determined based on the first defect and the second defect.

2. The method according to claim 1, characterized in that The identification step comprises: Determine the grayscale value of each pixel of the first image and the second image; The first defect and the second defect are determined based on the grayscale values ​​of the respective pixels.

3. The method according to claim 1 or 2, characterized in that: The determining step comprises: determining the size of the first defect and the second defect; receiving a threshold value for the first defect and the second defect; and Based on the comparison of the sizes of the first defect and the second defect with the threshold, an appearance grade of the solar cell is determined.

4. The method according to claim 3, characterized in that The first defect is selected from one or more of the group consisting of: color non-uniformity of an absorber layer of the solar cell, bubbles in the solar cell, and impurities in the solar cell.

5. The method according to claim 3, characterized in that: The second defect is selected from one or more of the group consisting of scratches, foreign matter and dirt on the surface of the solar cell.

6. The method according to claim 1, characterized in that The first angle is 40° to 50°, and the second angle is 75° to 85°.

7. The method according to claim 1, characterized in that The method further comprises: receiving a standard image of a standard sample associated with the solar cell; Determining a standard color of an absorber layer of the solar cell in the standard image; A color difference between a color of an absorber layer of the solar cell in the first image and the standard color is determined, wherein the first defect is the color difference.

8. The method according to claim 1, characterized in that The method further comprises: receiving a third image of the solar cell captured by a second camera under a third light source and a fourth light source, wherein the fourth light source is a backlight source; receiving a standard image of a standard sample associated with the solar cell; determining a third defect of the solar cell based on the third image and the standard image; and Based on the third defect, an appearance grade of the solar cell is determined.

9. The method according to claim 8, characterized in that The third defect is one or more selected from the group consisting of: a chipped corner, a chipped edge, and an edge burr of the solar cell.

10. The method according to claim 8, characterized in that The third defect is a difference in size between the solar cell and the standard sample.

11. The method according to claim 10, characterized in that The size difference includes a difference in size and position of an absorber layer of the solar cell and a difference in size and position of an absorber layer of the standard sample.

12. The method according to claim 8, characterized in that The method further comprises: identifying the first defect of the solar cell in the third image; and An appearance grade of the solar cell is determined based on the first defect in the first image and the third image.

13. The method according to claim 8, characterized in that The method further comprises: identifying the third defect of the solar cell in the first image and the second image; and Based on the third defect in the first image, the second image, and the third image, an appearance grade of the solar cell is determined.

14. The method according to claim 1, characterized in that The first camera is a line scan camera.

15. The method according to claim 8, characterized in that The second camera is an area array camera.

16. A device for detecting appearance defects of solar cells, characterized in that: The device includes a non-transitory computer storage medium having one or more executable instructions stored thereon, and the one or more executable instructions are executed by a processor to perform the following steps: Receiving a first image of a solar cell captured by a first camera under a first light source; receiving a second image of the solar cell captured by the first camera under a second light source, wherein a first angle between the first light source and the solar cell is smaller than a second angle between the second light source and the solar cell; identifying a first defect of the solar cell in the first image and a second defect of the solar cell in the second image; as well as An appearance grade of the solar cell is determined based on the first defect and the second defect.

Images