A method, device, electronic device, storage medium and program product for photovoltaic module image stitching

Abstract

The present application provides a method, apparatus, electronic device, storage medium, and program product for photovoltaic module image stitching. The method includes: determining an initial stitching line from the images to be stitched, where the images to be stitched are partial images of a photovoltaic module, and the initial stitching line represents the boundary of the solar cell in the images to be stitched close to the stitching position; adjusting the position of the initial stitching line according to a target object to obtain an adjusted stitching line; and stitching the images to be stitched according to the adjusted stitching line to obtain a stitched image. In the implementation process of the above solution, the initial stitching line determined from the images to be stitched provides the ability to perform position adjustment, and the images to be stitched are stitched according to the adjusted stitching line, which improves the problem of distortion of the photovoltaic module images due to reasons such as image quality. This ability to adjust the position of the stitching line makes the stitched image closer to the real photovoltaic module, thereby improving the authenticity of the stitched image.

Description

Technical Field

[0001] This application relates to the technical field of computer vision and image processing. Specifically, it relates to a method, device, electronic device, storage medium, and program product for splicing photovoltaic module images. Background Art

[0002] Currently, in the production process of photovoltaic modules, image acquisition and processing are key links in the quality inspection of photovoltaic modules. Due to the large size of photovoltaic modules, multiple industrial cameras (such as line scan cameras) are usually used to segmentally acquire local images of photovoltaic modules, and then image stitching technology is used to generate a complete module image. In related image stitching technologies, the stitching line determined based on global image features is usually relied on for stitching. However, in the process of practice, it is found that due to reasons such as image quality, the stitching line determined based on global image features is not accurate, so the stitched photovoltaic module image is distorted. Summary of the Invention

[0003] The purpose of the embodiments of this application is to provide a method, device, electronic device, storage medium, and program product for splicing photovoltaic module images, which is used to improve the problem of distortion of the stitched photovoltaic module image.

[0004] The embodiments of this application provide a method for splicing photovoltaic module images, including: determining an initial stitching line from the images to be stitched, where the images to be stitched are local images of photovoltaic modules, and the initial stitching line represents the boundary of the solar cells in the images to be stitched close to the stitching position; adjusting the position of the initial stitching line according to target objects, where the target objects include: solar cells, gap regions, bus bar regions, and / or segmentation information, and the segmentation information represents the relative position relationship of the images to be stitched; stitching the images to be stitched according to the adjusted stitching line to obtain the stitched image. In the implementation process of the above solution, the initial stitching line determined from the images to be stitched provides the ability to perform position adjustment, and the images to be stitched are stitched according to the adjusted stitching line, which improves the problem of distortion of the photovoltaic module image due to reasons such as image quality. This ability to adjust the position of the stitching line can obtain better image stitching processing results, making the stitched image closer to the real photovoltaic module, thereby improving the authenticity of the stitched image.

[0005] Optionally, in the embodiments of the present application, the target object includes segmentation information; determining an initial stitching line from the to-be-stitched images includes: determining the initial stitching line from the to-be-stitched images according to the segmentation information of the to-be-stitched images; adjusting the position of the initial stitching line according to the target object includes: determining the moving direction and the number of pixels of the initial stitching line according to the segmentation information; moving the position of the initial stitching line according to the moving direction and the number of pixels of the initial stitching line. In the implementation process of the above solution, by combining the image segmentation information to determine and adjust the initial stitching line, the adjusted stitching line can be made to be compatible with the stitching requirements of the images collected by the line-scan camera and the area array camera, effectively improving the applicable scenarios and flexibility of the solution.

[0006] Optionally, in the embodiments of the present application, the target object includes: unpowered battery cells in the to-be-stitched images; adjusting the position of the initial stitching line according to the target object includes: moving the position of the stitching line on the side close to the unpowered battery cell in the initial stitching line by a preset number of spacing pixels towards the side of the unpowered battery cell to obtain the moved stitching line, and the preset number of spacing pixels is an empirical value set according to the gap width between the battery cells. In the implementation process of the above solution, by moving the position of the stitching line on the side close to the unpowered battery cell in the initial stitching line towards the side of the unpowered battery cell, the function of dynamically adjusting the height of the gaps between the battery cells in the to-be-stitched images is realized, the actual gap width between the battery cells is restored more accurately, and the gap distortion at the stitching seam caused by the lack of information in the unpowered area is avoided, making the image after stitching processing closer to the real photovoltaic module, thereby improving the authenticity of the image after stitching processing.

[0007] Optionally, in the embodiments of the present application, the target object includes a gap area; adjusting the position of the initial stitching line according to the target object includes: moving the initial stitching line inward towards the battery cell by a first preset distance to retain the gap area between adjacent battery cells, where the first preset distance is less than the size of the battery cell in the moving direction; or moving the initial stitching line outward towards the outside of the battery cell by a second preset distance to exceed the gap area between adjacent battery cells, where the second preset distance is greater than the gap width between the battery cells and less than the sum of the size of the adjacent battery cell and the gap width in the moving direction. In the implementation process of the above solution, by moving the initial stitching line to retain the gap area between adjacent battery cells, the gap area between adjacent battery cells in the image after stitching with the moved stitching line is made more clearly visible, so that the re-inspection personnel or quality inspection personnel can accurately detect the gap defects in the stitched image. The gap defects in the photovoltaic module include, for example, foreign objects, string spacing, cell spacing, etc.

[0008] Optionally, in the embodiments of the present application, the target object includes: the bus bar region in the image to be stitched; adjusting the position of the initial stitching line includes: moving the stitching line position on the side of the initial stitching line close to the bus bar region towards the side of the bus bar region to obtain the moved stitching line. In the implementation process of the above solution, by moving the stitching line position on the side of the initial stitching line close to the bus bar region towards the side of the bus bar region, the bus bar region is presented in the finally stitched image, making the whole image more comprehensive and capable of presenting the complete structural information of the component.

[0009] Optionally, in the embodiments of the present application, moving the stitching line position on the side of the initial stitching line close to the bus bar region towards the side of the bus bar region includes: moving the stitching line position on the side of the initial stitching line close to the bus bar region towards the side of the bus bar region by the number of pixels of the region width; wherein, the number of pixels of the region width is an empirical value set according to the width of the bus bar region in the image, or the number of pixels of the region width is determined by the bus bar region detected from the image to be stitched by the target detection neural network model, or the number of pixels of the region width is determined by the bus bar region segmented from the image to be stitched by the image segmentation model. In the traditional method, the stitching line is usually fixed at the boundary of the battery cell, which may cut or ignore the bus bar region, resulting in the missing of bus bar information in the stitched image. Such an image missing the bus bar region cannot comprehensively reflect the structural information of the component and affects the accuracy of quality detection. In the implementation process of the above solution, by moving the initial stitching line towards the side of the bus bar region, it is ensured that the bus bar region is completely included in the stitched image, making the whole image more comprehensive and completely reflecting the structural information of the component, improving the problem of information loss caused by the fixed stitching line, and enhancing the authenticity of the image and the accuracy of quality detection. Further, according to the actual width of the bus bar region (determined by, for example, an empirical value, a target detection neural network model, or an image segmentation model), the position of the stitching line is accurately adjusted to adapt to the bus bar layouts of different photovoltaic modules (such as single-sided bus bars and double-sided bus bars), avoiding cutting the bus bar region due to different bus bar layouts of photovoltaic modules, and effectively increasing the flexibility and applicability of the solution.

[0010] Optionally, in the embodiments of the present application, stitching the image to be stitched according to the adjusted stitching line includes: aligning the image to be stitched according to the adjusted stitching line to obtain the aligned image; performing weighted fusion on the aligned image to obtain the stitched image. In the implementation process of the above solution, the aligned image is weighted and fused by the adjusted stitching line. This smooth processing method of weighted fusion effectively eliminates phenomena such as edge traces and inconsistent illumination in the stitching region, thereby achieving the effect of smooth image transition.

[0011] Optionally, in the embodiments of the present application, weighted fusion of the aligned images includes: obtaining a weight matrix, which is generated according to a Gaussian weight template or a cosine weight template with a preset fusion width; using the weight matrix to perform weighted fusion on the aligned images. In the implementation process of the above solution, the aligned images are weighted and fused by the weight matrix generated according to the weight template with the preset fusion width. Since the values of the weight matrix gradually transition from one side of the image to the other side, this smoothing processing method using the weight matrix effectively eliminates phenomena such as edge traces and inconsistent illumination in the splicing area. The design of this weight template ensures that the pixel values in the seam area gradually change, quickly achieving the effect of smooth image transition and improving the efficiency of weighted fusion. In addition, by generating the weight matrix according to the Gaussian weight template or the cosine weight template with the preset fusion width, the weight template can be selected according to the fusion requirements of different photovoltaic module images, effectively improving the applicable scenarios and flexibility of this solution.

[0012] The embodiments of the present application further provide a photovoltaic module image splicing device, including: a splicing line determination module, configured to determine an initial splicing line from the images to be spliced, where the images to be spliced are partial images of photovoltaic modules, and the initial splicing line represents the boundary of the solar cell near the splicing position in the images to be spliced; a splicing line adjustment module, configured to adjust the position of the initial splicing line according to the target object to obtain an adjusted splicing line, where the target object includes: solar cells, gap regions, bus bar regions, and / or segmentation information, and the segmentation information represents the relative positional relationship of the images to be spliced; an image splicing module, configured to splice the images to be spliced according to the adjusted splicing line to obtain a spliced image.

[0013] Optionally, in the embodiments of the present application, the target object includes segmentation information; the splicing line determination module includes: an initial splicing line determination sub-module, configured to determine an initial splicing line from the images to be spliced according to the segmentation information of the images to be spliced; the splicing line adjustment module includes: a direction pixel determination sub-module, configured to determine the moving direction and the number of pixels of the initial splicing line according to the segmentation information; an initial splicing line moving sub-module, configured to move the position of the initial splicing line according to the moving direction and the number of pixels of the initial splicing line.

[0014] Optionally, in the embodiments of the present application, the target object includes the unpowered solar cells in the images to be spliced; the splicing line adjustment module includes: a first splicing line moving sub-module, configured to move the position of the splicing line on the side of the initial splicing line close to the unpowered solar cells to the side of the unpowered solar cells by a preset number of spaced pixels, where the preset number of spaced pixels is an empirical value set according to the gap width between the solar cells.

[0015] Optionally, in the embodiments of the present application, the target object includes a gap region; the splicing line adjustment module includes: a first splicing line moving sub-module, configured to move the initial splicing line inward of the battery cell by a first preset distance to reserve the gap region between adjacent battery cells, where the first preset distance is less than the size of the battery cell in the moving direction; or, a second splicing line moving sub-module, configured to move the initial splicing line outward of the battery cell by a second preset distance to reserve the gap region between adjacent battery cells, where the second preset distance is greater than the gap width between the battery cells and less than the sum of the size of the adjacent battery cells and the gap width in the moving direction.

[0016] Optionally, in the embodiments of the present application, the target object includes: the bus bar region in the image to be spliced; the splicing line adjustment module includes: a third splicing line moving sub-module, configured to move the position of the splicing line on the side close to the bus bar region in the initial splicing line to one side of the bus bar region to obtain the moved splicing line.

[0017] Optionally, in the embodiments of the present application, the third splicing line moving sub-module includes: a third splicing line moving unit, configured to move the position of the splicing line on the side close to the bus bar region in the initial splicing line to one side of the bus bar region by the number of pixels of the region width; where the number of pixels of the region width is an empirical value set according to the width of the bus bar region in the image, or the number of pixels of the region width is determined by the bus bar region detected from the image to be spliced by a target detection neural network model, or the number of pixels of the region width is determined by the bus bar region segmented from the image to be spliced by an image segmentation model.

[0018] Optionally, in the embodiments of the present application, the image splicing module includes: an image alignment sub-module, configured to align the image to be spliced according to the adjusted splicing line to obtain the aligned image; a weighted fusion sub-module, configured to perform weighted fusion on the aligned image to obtain the spliced image.

[0019] Optionally, in the embodiments of the present application, the weighted fusion sub-module includes: a weight matrix obtaining unit, configured to obtain a weight matrix, where the weight matrix is generated according to a Gaussian weight template or a cosine weight template of a preset fusion width; an image weighted fusion unit, configured to perform weighted fusion on the aligned image using the weight matrix.

[0020] The embodiments of the present application further provide an electronic device, including: a processor and a memory, where the memory stores machine-readable instructions executable by the processor, and when the machine-readable instructions are run by the processor, the above-described method is executed.

[0021] The embodiments of the present application further provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the above-described method is executed.

[0022] The embodiments of the present application also provide a computer program product, including: a computer program or computer instructions, and when the computer program or computer instructions are run by a processor, they execute the method described above. Description of the Drawings

[0023] To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the embodiments of the present application, and thus should not be regarded as a limitation of the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative efforts.

[0024] Figure 1 Showing a schematic partial image of the photovoltaic module provided by the comparative example;

[0025] Figure 2 Showing a schematic diagram of the comparative example provided that cannot reflect the actual gap of the photovoltaic module;

[0026] Figure 3 Showing a schematic flowchart of the photovoltaic module image stitching method provided by the embodiments of the present application;

[0027] Figure 4 Showing a schematic diagram of the corner positions of all the solar cells powered on provided by the embodiments of the present application;

[0028] Figure 5 Showing a schematic diagram of the initial stitching line of all the solar cells powered on provided by the embodiments of the present application;

[0029] Figure 6 Showing a schematic diagram of adjusting the initial stitching line when all the solar cells are powered on provided by the embodiments of the present application; <>

[0030] Figure 7 Showing a schematic diagram of the image to be stitched of some of the solar cells powered on provided by the embodiments of the present application; <>

[0031] <><> Figure 8 Showing a schematic diagram of the initial stitching line of some of the solar cells powered on provided by the embodiments of the present application; <>

[0032] <><> Figure 9 Showing a schematic diagram of adjusting the initial stitching line when some of the solar cells are powered on provided by the embodiments of the present application; <>

[0033] <><> Figure 10 Showing a schematic diagram of the stitching line offsetting towards the inner side of the solar cell provided by the embodiments of the present application; <>

[0034] <><> Figure 11Schematic diagram showing that the splicing line of the embodiment of the present application deviates outward from the cell;

[0035] Figure 12 Schematic diagram of the appearance image of the bus bar area in the image to be spliced provided by the embodiment of the present application;

[0036] Figure 13 Schematic diagram of the appearance image after splicing provided by the embodiment of the present application;

[0037] Figure 14 Schematic diagram of the structure of the photovoltaic module image splicing device provided by the embodiment of the present application.

[0038] Icons: 100 - photovoltaic module; 101 - cell array; 102 - main grid line; 103 - cell; 104 - cell gap; 105 - frame; 106 - bus bar area; 200 - photovoltaic module image splicing device; 210 - splicing line determination module; 220 - splicing line adjustment module; 230 - image splicing module. Detailed implementation manners

[0039] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application usually described and shown in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application provided in the accompanying drawings below is not intended to limit the scope of the present application claimed, but only represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application. It should be noted that similar reference numerals and letters indicate similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, the terms "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance. It should be noted that, without conflict, the embodiments or technical features in the embodiments of the present application can be combined.

[0040] Please refer to Figure 1Schematic diagram of a partial image of a photovoltaic module provided by the shown comparative example; the photovoltaic module 100 is a device that converts solar energy into electrical energy. The photovoltaic module 100 includes: a frame 105 and a cell array 101 embedded in the frame 105. The cell array 101 can be formed by splicing a plurality of cells 103. There are a plurality of main grid lines 102 extending in the horizontal direction in each cell 103. It can be understood that there is a cell gap 104 (also simply referred to as a gap) between two adjacent cells 103 in the cell array 101. In addition, there is also a cell gap 104 between the cell 103 and the frame 105. In addition, the photovoltaic module may also include components such as an EVA film, a backsheet, a glass cover plate, and / or a junction box.

[0041] To ensure the image quality requirements, an industrial camera (such as a line scan camera) is usually used to take a picture of a cell array 101 in the photovoltaic module 100 to obtain an image to be stitched. The number of cells 103 included in the cell array 101 can be determined according to specific scenario requirements such as camera performance parameters and image quality, for example, 6, 10, 12, 16, or 18 cells 103, etc. During the image acquisition process, the industrial camera for acquiring the image to be stitched can use multiple area array cameras or multiple line scan cameras. Among them, the area array camera can specifically use an industrial camera with a near-infrared monochrome charge-coupled device (CCD) to acquire an electroluminescence (EL) image, or a color industrial camera to acquire the appearance image of the photovoltaic module. Among them, the number of industrial cameras can be 2, 4, or 6, and the specific number of cameras can be determined according to specific scenario requirements such as camera performance parameters and image quality.

[0042] For ease of understanding, the following takes the example of using four area array cameras to photograph a photovoltaic module. Assume that it takes three steps to photograph a complete photovoltaic module using four area array cameras. The image to be stitched in the figure is the image taken by the first area array camera when using four area array cameras and three steps of photographing. That is to say, for each step, these four area array cameras need to respectively photograph and image the area within the field of view of the camera lens. Then each area array camera collects approximately 1 / 12 of the image of the photovoltaic module each time. To obtain a complete image of the entire photovoltaic module 100, image stitching technology is usually required to stitch the multiple images to be stitched taken.

[0043] Please refer to Figure 2The schematic diagram of the actual gap of the photovoltaic module provided by the shown comparative example cannot be reflected; in some related technologies of image stitching, it is usually stitched without moving the stitching line, and finally the stitched photovoltaic module image is distorted. There are many reasons for image distortion. For example, when using multiple cameras for segmented shooting, the photovoltaic module is usually driven by a stepping motor to move, and there is a positioning error (such as a micron-level error) in the mechanical system of the stepping motor. In addition to the positioning error, other error sources can also be camera lens distortion, uneven illumination, image noise, etc. These may all cause the physical positions at the cell arrays of adjacent segmented information to be not completely aligned. If the stitching line is fixed, this positioning error will form misalignment or overlap at the gap, resulting in image distortion. For example, in the gap area of the stitched photovoltaic module image in the row direction or column direction, the actual gap of the photovoltaic module cannot be reflected.

[0044] The defects existing in the above solutions in the related technologies are all the results obtained by the inventor through practice and careful research. Therefore, the process of discovering the above problems and the solutions proposed by the embodiments of the present invention below for the above problems should both be the contributions made by the inventor to the present invention in the process of the present invention.

[0045] To improve the above problems, the following is a photovoltaic module image stitching method, device, electronic device, storage medium and program product provided by an embodiment of the present application. By determining an initial stitching line from the image to be stitched, the ability to perform position adjustment is provided, and the image to be stitched is stitched according to the adjusted stitching line, improving the problem of image distortion of the photovoltaic module image due to reasons such as image quality. This position adjustment ability of the stitching line can obtain a better image stitching processing result, making the stitched image closer to the real photovoltaic module, so that the stitched photovoltaic module image can truly reflect the actual gap of the photovoltaic module in the row direction or column direction.

[0046] In some application scenarios, the above photovoltaic module image stitching method can be applied to a terminal device. The above terminal device can, for example, include a PC, a tablet computer, a mobile phone, etc.; in some other application scenarios, the above photovoltaic module image stitching method can also be applied to a server device. The above server device can, for example, include a server, a server cluster or a cloud platform, etc. In addition, the above terminal device can be a physical device or a virtual device, such as a virtual machine. In the following description of the embodiments of the present application, an example of being applied to a server is used for elaboration.

[0047] Please refer to Figure 3 The flow schematic diagram of the photovoltaic module image stitching method provided by the embodiment of the present application shown; the implementation manner of the photovoltaic module image stitching method can include:

[0048] Step S110: Determine an initial splicing line from the images to be spliced. The images to be spliced are partial images of a photovoltaic module, and the initial splicing line represents the boundary of the solar cell in the images to be spliced near the splicing position.

[0049] It can be understood that the above images to be spliced can be partial images currently captured by an industrial camera for a photovoltaic module, or historical images captured by the industrial camera, which is not limited here. The above server can obtain at least two images to be spliced from the industrial camera, or obtain at least two images to be spliced from a preset storage location for historical images. The above images to be spliced can be, for example, an image of a solar cell array composed of a preset number of solar cells. In this case, the object to be spliced in the images to be spliced is the solar cell array. Of course, the above images to be spliced can also be an image of a single solar cell. In this case, the object to be spliced in the images to be spliced is the solar cell itself.

[0050] Please refer to Figure 4 the schematic diagram showing the corner positions of all the powered-on solar cells provided by the embodiment of the present application; all the corner positions in the partial image of the photovoltaic module are shown in the figure. However, in fact, during the fitting process of the corner positions, only some of the corner positions can be fitted because the initial splicing line is a straight line, and theoretically, only two or more corner positions are required to fit this straight line. For example, only any two of the points EFHG in the figure can be used to fit the EF splicing line, the FH splicing line, the HG splicing line, and / or the GE splicing line as the initial splicing line. Of course, in some implementation scenarios, all the corner positions can also be fitted according to needs. For example, when fitting the EF splicing line, not only the E corner position and the F corner position are used, but also the other five corner positions between E and F are required to fit an EF splicing line as the initial splicing line together. The process of determining the initial splicing line from the images to be spliced can include: first, determining the corner positions according to the images to be spliced, and then using the least squares algorithm to fit the corner positions to obtain the initial splicing line. Among them, the corner positions can be directly marked in the images to be spliced, or detected from the images to be spliced through a corner detection algorithm. The corner detection algorithms that can be used include, for example, the Harris corner detection, SIFT, SURF, etc. algorithms. Optionally, before fitting the corner positions, the corner positions can be screened first. For example, the corner positions of the points EFGH are screened out from all the corner positions in the images to be spliced as the corner positions to be fitted.

[0051] Please refer to Figure 5Schematic diagram of the initial stitching line when all the battery cells provided by the embodiments of the present application are powered on; the initial stitching line refers to the boundary line of the battery cells near the stitching position in the image to be stitched, which is a reference line in the stitching process and is used to preliminarily determine the position of the gap between the battery cells. After screening out the corner positions of points E, F, G, and H from all the corner positions in the image to be stitched, a straight line fitting can be performed on the corner positions of any two of the points EF, FH, HG, and / or GE among the corner positions of points E, F, G, and H to obtain the initial stitching line as shown in the figure.

[0052] Step S120: Adjust the position of the initial stitching line according to the target object to obtain the adjusted stitching line. The target object includes: battery cells, gap regions, bus bar regions, and / or segmentation information, and the segmentation information characterizes the relative positional relationship of the image to be stitched.

[0053] It can be understood that there are many purposes for adjusting the position of the initial stitching line, including expanding or shrinking the height of the gap between the battery cells to be stitched for the image, offsetting the gap between the battery cells, or adding a bus bar region, etc., and its adjustment method also varies according to different targets. Therefore, the adjustment method here will be described in detail below.

[0054] Step S130: Stitch the image to be stitched according to the adjusted stitching line to obtain the stitched image.

[0055] In the implementation process of the above solution, the initial stitching line determined from the image to be stitched provides the ability to perform position adjustment, and stitches the image to be stitched according to the adjusted stitching line, improving the problem of image distortion of the photovoltaic module caused by reasons such as image quality. This ability to adjust the position of the stitching line can obtain a better image stitching processing result, making the stitched image closer to the real photovoltaic module, thereby improving the authenticity of the stitched image.

[0056] Please refer to Figure 6 Schematic diagram of adjusting the initial stitching line when all the battery cells provided by the embodiments of the present application are powered on; when all the battery cells are powered on, after determining the initial stitching line from the image to be stitched, the position of the initial stitching line can be adjusted. For example, for the initial stitching line obtained by fitting according to the corner positions F and H in the figure, the stitching line can be moved a preset pixel width (such as half of the gap width) in the direction of the gap on the right side, or the stitching line can be directly moved until it reaches the perpendicular bisector between the corner positions F and I, or the stitching line can be directly moved until it reaches the perpendicular bisector between the corner positions H and K, so as to obtain the moved and adjusted stitching line. The adjusted stitching line refers to the stitching line obtained by adjusting the position of the initial stitching line, and the purpose of the adjustment is to make the image alignment at the stitching seam more accurate and reduce distortion.

[0057] Optionally, when all the solar cells are powered on, the stitching line required in step S130 can also be obtained without moving the initial stitching line. Specifically, for example, after detecting the corner positions F and H in the first row of solar cell arrays EFGH, since the right side of the first row of solar cell arrays EFGH is also the powered-on solar cell array IJKL (please refer to Figure 6 the right side of), therefore, the corner positions I and K of the right-side powered-on solar cell array can be detected. Then, the first perpendicular bisector between the corner position F and the corner position I can be calculated and determined as the stitching line for stitching the to-be-stitched image, or the second perpendicular bisector between the corner position H and the corner position K can be calculated and determined as the stitching line for stitching the to-be-stitched image. The stitching process of the first row is described above, and the stitching processes of other rows are similar. Moreover, since all are powered on and the corner positions in the column direction can be detected, the stitching process in the column direction is similar and will not be elaborated here.

[0058] As an alternative implementation of the above step S110, the above target object includes the segmentation information of the to-be-stitched image; the implementation of determining the initial stitching line from the to-be-stitched image may include:

[0059] Step S111: Determine the initial stitching line from the to-be-stitched image according to the segmentation information of the to-be-stitched image, where the segmentation information represents the relative position relationship of the to-be-stitched image.

[0060] The segmentation information refers to the data representing the relative position relationship of the to-be-stitched image in the complete photovoltaic module. Such segmentation information data can be the data information recorded when the to-be-stitched image is captured. The above segmentation information may include camera parameter information, the order of each local image (such as the order of the first solar cell array, the second solar cell array, etc.), the position (such as the first row and the first column in the first solar cell array), and / or the relationship between this segmented image and other segmented images (such as adjacent or non-adjacent), etc.

[0061] Please refer to Figure 7 the schematic diagram of the to-be-stitched image with some solar cells powered on provided by the embodiment of the present application shown; in some scenarios where only some solar cells in the photovoltaic module are powered on, such as the case of powering on a certain row or a certain column of solar cell arrays in the photovoltaic module. For the convenience of understanding and illustration, the case of only powering on the first row of solar cell arrays will be taken as an example for description below. It can be known from the segmentation information of the to-be-stitched image that the figure shows the first solar cell array of the first row segmented image. The process of determining the initial stitching line from the to-be-stitched image of this solar cell array is similar to the above step S110, that is, first determine the corner positions according to the to-be-stitched image, and then fit the corner positions to obtain the initial stitching line, so it will not be elaborated here.

[0062] As an alternative implementation of the above step S120, the implementation of adjusting the position of the initial splicing line according to the target object may include:

[0063] Step S121: Determine the moving direction and number of pixels of the initial splicing line according to the segmentation information.

[0064] Please refer to Figure 8 the schematic diagram of the initial splicing line of the partially powered battery cells provided by the embodiments of the present application shown; for example, the implementation of the above step S121: the segmentation information of different types of photovoltaic modules may be different, and the moving direction and number of pixels of the initial splicing line determined by different segmentation information may also be different. It can be that a person calculates the moving direction and number of pixels of the initial splicing line through the segmentation information and the model of the photovoltaic module, or the specific values of the segmentation information corresponding to the moving direction and number of pixels of the initial splicing line can be obtained from a configuration file or a database, so that the moving direction and number of pixels of the initial splicing line can be dynamically adjusted. The moving direction and number of pixels in the above configuration file or database can be calculated and set in advance by a person. Specifically, for the battery cell arrays of the segmentation diagrams such as the first row, the second row, and the third row, the spacing pixel value in the row direction is assumed to be expressed as The positive or negative value of this value represents the moving direction of the initial splicing line (for example, a positive value indicates moving downward, and a negative value indicates moving upward), and the numerical value of this value represents the number of moving pixels of the initial splicing line. Similarly, for the battery cell arrays of the segmentation diagrams such as the first column, the second column, and the third column, the spacing pixel value in the column direction is assumed to be expressed as The positive or negative value of this value represents the moving direction of the initial splicing line (for example, a positive value indicates moving to the right, and a negative value indicates moving to the left), and the numerical value of this value represents the number of moving pixels of the initial splicing line.

[0065] Step S122: Move the position of the initial splicing line according to the moving direction and number of pixels of the initial splicing line.

[0066] Please refer to Figure 9 the schematic diagram of adjusting the initial splicing line when the partially powered battery cells provided by the embodiments of the present application are shown; in some scenarios where only some of the battery cells in the photovoltaic module are powered on, some corner positions cannot be detected. For example, when a certain row or column of battery cell arrays in the photovoltaic module is powered on, the initial splicing line can be adjusted in the column direction or the row direction to obtain the spliced line after moving adjustment. For the initial splicing line below the battery cell array of the first row segmentation diagram, the following formula can be used for moving:

[0067]

[0068] Among them, represents the spliced line after being moved below the cell array of the first-row segmented diagram, represents the initial spliced line below the cell array of the first-row segmented diagram, represents the pitch pixel value in the row direction, and respectively represent the row and the column.

[0069] As another alternative implementation of the above step S120, the above target object may include: unpowered cells in the image to be spliced; the implementation of adjusting the position of the initial spliced line according to the target object may include:

[0070] Step S123: Move the position of the spliced line on the side of the initial spliced line close to the unpowered cell towards the side of the unpowered cell by a preset number of pitch pixels to obtain the moved spliced line, and the preset number of pitch pixels is an empirical value set according to the gap width between the cells.

[0071] The above unpowered cells refer to the cells that are not powered on during the acquisition of the electroluminescence (EL) image. Since these cells are not powered on, they appear as dark regions in the image (please refer to Figure 7 or Figure 8 where the dark regions represent unpowered cells and the bright regions represent powered-on cells), lacking obvious feature points. Therefore, it is difficult to find the corner positions of unpowered cells through algorithms, or even if the corner positions found through algorithms are inaccurate.

[0072] Among them, the above preset number of pitch pixels refers to the number of pixels determined according to the actual gap width between the cells. The above preset number of pitch pixels can be an empirical value set according to the gap width between the cells, and this preset number of pitch pixels is used to adjust the position of the spliced line to make it closer to the actual gap position.

[0073] For example, the implementation of the above step S123: For a photovoltaic module with segmented information of three rows and four columns, the initial spliced line below the cell array of the first-row segmented diagram can be moved first using the following formula:

[0074]

[0075] For the cell array of the second-row segmented diagram, use the following formula to move the initial spliced lines above and below the cell array:

[0076]

[0077] For the cell array of the third row segmented diagram, use the following formula to move the initial splicing line above the cell array:

[0078]

[0079] For the cell array of the first column segmented diagram, use the following formula to move the initial splicing line on the right side of the cell array:

[0080]

[0081] For the cell array of the second column segmented diagram, use the following formula to move the initial splicing line on the left side of the cell array:

[0082]

[0083] For the cell array of the third column segmented diagram, use the following formula to move the initial splicing line on the right side of the cell array:

[0084]

[0085] For the cell array of the fourth column segmented diagram, use the following formula to move the initial splicing line on the left side of the cell array:

[0086]

[0087] Among them, for the row direction, respectively represent the moved splicing lines below the cell arrays of the first row and the second row segmented diagrams, respectively represent the moved splicing lines above the cell arrays of the second row and the third row segmented diagrams, and represent the initial splicing lines below the cell arrays of the first row and the second row segmented diagrams, and represent the initial splicing lines above the cell arrays of the first row and the second row segmented diagrams, represents the spacing pixel value of the first move in the row direction, and respectively represent the row and the column;

[0088] For the column direction, represents the moved splicing lines on the left side of the cell arrays of the first column, the second column, the third column and the fourth column segmented diagrams, represents the moved splicing lines on the right side of the cell arrays of the first column, the second column, the third column and the fourth column segmented diagrams, The initial splicing line on the left side of the cell array representing the segmented diagrams of the first column, second column, third column, and fourth column The initial splicing line on the right side of the cell array representing the segmented diagrams of the first column, second column, third column, and fourth column Represents the pitch pixel value for the first movement in the column direction.

[0089] As another alternative implementation of the above step S120, it can be understood that the above implementation of adjusting the position of the initial splicing line can be applied to the appearance image of the photovoltaic module or the electroluminescence (EL) image of the photovoltaic module. The above target object may include a gap area. The implementation of adjusting the splicing line for this gap area may include:

[0090] Step S124a: Move the initial splicing line inward toward the inside of the cell by a first preset distance to retain the gap area between adjacent cells. The first preset distance is less than the size of the cell in the moving direction.

[0091] Please refer to Figure 10 The schematic diagram showing the offset of the splicing line toward the inside of the cell provided by the embodiment of the present application; in the specific practice process, the splicing line in the appearance image of the photovoltaic module can be offset toward the inside of the cell (such as the left side direction in the figure), or the splicing line in the electroluminescence (EL) image of the photovoltaic module can be offset toward the inside of the cell (such as the left side direction in the figure). In order to present a more realistic gap area image in the spliced image (such as the appearance image or electroluminescence image of the photovoltaic module), the initial splicing line of the first cell (such as the splicing line near the right boundary) can be moved inward toward the inside of the first cell (such as the center point direction of the first cell is the left side direction) by a first preset distance, so as to be able to retain the real gap area between adjacent cells in the spliced image after the splicing process is completed, rather than the gap spliced by the two side images. Among them, the first preset distance can be pre-set and less than the size of the cell in the moving direction (such as the width of the first cell on the horizontal axis in the figure), so that the moved splicing line falls inside the first cell, so as to be able to retain the real gap between adjacent cells in the spliced image after the splicing process is completed according to the moved splicing line. In the subsequent defect detection process of the photovoltaic module, it is necessary to perform defect detection based on the real gap between the cells, such as whether the gap width meets the preset width standard, etc.

[0092] Or, the above implementation of adjusting the position of the initial splicing line may include:

[0093] Step S124b: Move the initial splicing line outward from the cell by a second preset distance to reserve the gap area between adjacent cells. The second preset distance is greater than the gap width between cells and less than the sum of the size of an adjacent cell in the moving direction and the gap width.

[0094] Please refer to Figure 11 the schematic diagram showing the offset of the splicing line provided by the embodiment of the present application outward from the cell; in a specific practice process, the splicing line in the appearance image of the photovoltaic module can be offset outward from the cell (such as the right side direction in the figure), or the splicing line in the electroluminescence (EL) image of the photovoltaic module can be offset outward from the cell (such as the right side direction in the figure). In order to present a more realistic image of the gap area in the spliced image (such as the appearance image or the electroluminescence image of the photovoltaic module), the initial splicing line of the first cell (such as the splicing line near the right boundary) can be moved outward from the first cell (i.e., the direction away from the center point of the first cell is called the outward direction) by a second preset distance to reserve the gap area between adjacent cells. Among them, the second preset distance can be a preset distance, which is a distance greater than the gap width between cells and less than the sum of the size of an adjacent cell in the moving direction and the gap width. Let the moved splicing line fall inside the second cell, so that after the splicing process is completed according to the moved splicing line, the real gap area between adjacent cells can be reserved in the spliced image, rather than the gap spliced by the two side images. In the subsequent defect detection process of the photovoltaic module, it is necessary to perform defect detection according to the real gap between cells, such as whether the gap width meets the preset width standard, etc.

[0095] Taking a photovoltaic module with segmented information of three rows and four columns as an example to illustrate the specific process of moving all the moved splicing lines in a preset direction. First, the following formula can be used to move the initial splicing line below the cell array of the first row segmented diagram:

[0096]

[0097] For the cell array of the second row segmented diagram, the following formula is used to move the initial splicing lines above and below the cell array:

[0098]

[0099] For the cell array of the third row segmented diagram, the following formula is used to move the initial splicing line above the cell array:

[0100]

[0101] For the cell array of the first column segmented diagram, use the following formula to move the initial splicing line on the right side of the cell array:

[0102]

[0103] For the cell array of the second column segmented diagram, use the following formula to move the initial splicing line on the left side of the cell array:

[0104]

[0105] For the cell array of the third column segmented diagram, use the following formula to move the initial splicing line on the right side of the cell array:

[0106]

[0107] For the cell array of the fourth column segmented diagram, use the following formula to move the initial splicing line on the left side of the cell array:

[0108]

[0109] Among them, for the row direction, respectively represent the spliced lines after being moved below the cell arrays of the first row and second row segmented diagrams, respectively represent the spliced lines after being moved above the cell arrays of the second row and third row segmented diagrams, and represent the initial spliced lines below the cell arrays of the first row and second row segmented diagrams, and represent the initial spliced lines above the cell arrays of the first row and second row segmented diagrams, represents the preset number of pixel spacings for the secondary movement in the row direction. The positive or negative value of this value indicates the secondary movement direction of the spliced line, and respectively represent the th row and the th column;

[0110] For the column direction, represents the spliced lines after being moved to the left of the cell arrays of the first column, second column, third column, and fourth column segmented diagrams, represents the spliced lines after being moved to the right of the cell arrays of the first column, second column, third column, and fourth column segmented diagrams, represents the initial spliced lines to the left of the cell arrays of the first column, second column, third column, and fourth column segmented diagrams, represents the initial spliced lines to the right of the cell arrays of the first column, second column, third column, and fourth column segmented diagrams, It represents the preset number of pixels of the pitch for the re - movement in the column direction, and the positive or negative value of this value represents the re - movement direction of the splicing line.

[0111] As another alternative implementation of the above - mentioned step S120, the above - mentioned target object may include: the bus bar area in the image to be spliced; the implementation of adjusting the position of the initial splicing line may include:

[0112] Step S125: Move the position of the splicing line on the side of the initial splicing line close to the bus bar area towards the side of the bus bar area to obtain the moved splicing line.

[0113] One implementation of the above - mentioned step S125 is, for example: Please refer to Figure 12 the schematic diagram of the appearance image of the bus bar area in the image to be spliced provided by the embodiment of the present application shown in; the bus bar area refers to the metal bar area in the photovoltaic module for collecting and transmitting current, and the bus bar in the image usually appears as a bright strip. Therefore, the initial splicing line fitted by the right - hand corner points F and H of the second - column cell array can be moved to the right, so that the finally spliced image presents the bus bar area. The determination process of the specific number of pixels of the area width to be moved will be described in detail below. The process of moving the initial splicing line to the right can be expressed by the following formula:

[0114]

[0115] Among them, represents the moved splicing line on the right side of the cell array of the second - column segmented diagram, represents the initial splicing line on the right side of the cell array of the second - column segmented diagram, represents the number of pixels of the area width to be moved.

[0116] Optionally, another implementation of the above - mentioned step S125 is, for example: The initial splicing line fitted by the left - hand corner points I and K of the cell array of the third - column segmented diagram can also be moved to the left, so that the finally spliced image presents the bus bar area. The determination process of the specific number of pixels of the area width to be moved will be described in detail below. The process of moving the initial splicing line to the left can be expressed by the following formula:

[0117]

[0118] Among them, represents the moved splicing line on the left side of the cell array of the second - column segmented diagram, represents the initial splicing line on the left side of the cell array of the second - column segmented diagram, represents the number of pixels of the area width to be moved.

[0119] As an alternative implementation of the above step S125, the implementation of moving the position of the splicing line on the side close to the bus bar area in the initial splicing line towards the side of the bus bar area may include:

[0120] Step S125a: Move the position of the splicing line on the side close to the bus bar area in the initial splicing line towards the side of the bus bar area by the number of pixels of the area width.

[0121] Wherein, the number of pixels of the area width refers to the width of the bus bar area in the image, in pixels. The above number of pixels of the area width may be an empirical value set according to the width of the bus bar area in the image, or the above number of pixels of the area width may be determined from the bus bar area detected from the image to be spliced by a target detection neural network model, or the above number of pixels of the area width may also be determined from the bus bar area segmented from the image to be spliced by an image segmentation model. The target detection neural network model is a computer vision model based on deep learning, used to detect specific targets (such as the bus bar area) from images and output their position and size information. The above target detection neural network model may adopt models such as SSD model, RetinaNet model, EfficinetDet model, CenterNet model, Mask-RCNN model, DETR model, YOLO model or Faster-RCNN model, and the above image segmentation model may adopt models such as U-Net model, DeepLabv3 model, PSPNet model or SegFormer model, etc. Taking the cell array of the second-column segmented diagram as an example, using the following formula Calculating the target box of the bus bar area detected from the image to be spliced by the target detection neural network model can obtain the above number of pixels of the area width. Among them, represents the number of pixels of the area width that needs to be moved, represents the abscissa of the lower right corner of the target box of the bus bar area, represents the abscissa of the lower left corner of the target box of the bus bar area.

[0122] In the implementation process of the above solution, by moving the position of the splicing line on the side close to the bus bar area in the initial splicing line towards the side of the bus bar area, the finally spliced image presents the bus bar area, making the whole image more comprehensive and capable of presenting the complete structural information of the component.

[0123] As an alternative implementation of the above step S130, the implementation of splicing the image to be spliced according to the adjusted splicing line may include:

[0124] Step S131: Align the image to be spliced according to the adjusted splicing line to obtain the aligned image.

[0125] For example, the implementation manner of the above step S131 is as follows: Align the images to be spliced according to the adjusted splicing line to obtain the aligned images. The purpose of alignment here is to make the geometric shapes at the splicing seams of adjacent images match. Optionally, after obtaining the aligned images, geometric perspective transformation (such as translation, rotation) can also be performed on the aligned images.

[0126] Step S132: Perform weighted fusion on the aligned images to obtain the spliced image.

[0127] Please refer to Figure 13 the schematic diagram of the spliced appearance image provided by the embodiment of the present application shown in the figure; For example, the implementation manner of the above step S132 is as follows: By performing weighted fusion on the aligned images to obtain the spliced image, it can effectively eliminate the lateral instability or longitudinal instability of the photovoltaic module during the step-by-step shooting of the three-stage type using a matrix camera, effectively reducing the deviation or misalignment generated at the mapping position, so that the spliced image can more truly reflect the actual defect situation of the photovoltaic module.

[0128] As an optional implementation manner of the above step S132, the implementation manner of performing weighted fusion on the aligned images can include:

[0129] Step S132a: Obtain a weight matrix, which is generated according to a Gaussian weight template or a cosine weight template with a preset fusion width.

[0130] The weight matrix refers to the matrix for performing weighted fusion at the splicing line of the images to be spliced, and its element values represent the weights of each pixel in the overlapping area, which is mainly used for pixel-by-pixel adjustment at the splicing line and the edge area of the images to be spliced to ensure smooth transition in the splicing area.

[0131] The preset fusion width is a parameter in the image splicing process, which determines the smoothness of the transition in the splicing seam area. This fusion width determines the transition range of the splicing seam area. The larger the width, the wider the transition area, and the smoother the change in brightness or texture at the splicing seam. Therefore, by setting an appropriate width (for example, for high-resolution images, the width can be set to 10 - 20 pixels; for low-resolution images, the width can be set to 5 - 10 pixels), the edge traces (such as brightness jumps or geometric misalignments) at the splicing seam can be effectively eliminated.

[0132] The factors considered for the above-mentioned preset fusion width may include: image resolution, lighting conditions, characteristics of the seam area, requirements for stitching accuracy, and computational efficiency, etc. From the perspective of image resolution, high-resolution images require a larger width to achieve smooth transition, while low-resolution images can use a smaller width. From the perspective of lighting conditions, if there is a large difference in lighting between the two side images, a larger width needs to be set to balance the brightness difference. From the perspective of the characteristics of the seam area, if the seam area contains complex textures or details, a larger width needs to be set to avoid unnatural transition. From the perspective of requirements for stitching accuracy, for high-precision stitching (such as quality inspection of photovoltaic modules), a smaller width needs to be set to retain more details. From the perspective of computational efficiency, the larger the width, the higher the computational complexity. Therefore, on the premise of ensuring the stitching effect, a smaller width should be selected as much as possible to improve computational efficiency.

[0133] Step S132b: Use the weight matrix to perform weighted fusion on the aligned images.

[0134] The implementation manner of the above step S132b is, for example: perform weighted fusion on the aligned images using the weight matrix. For example, for each edge area of the image (such as the right side, left side, top, bottom), multiply the weight value in the weight matrix according to the relative position of the pixels in this edge area respectively, so as to adjust the pixel values in the stitching area. Here, weighted fusion refers to performing weighted average on the pixel values according to the preset weight matrix in the overlapping area of the aligned images to achieve the process of smooth transition of the images. Another example: for different stitching areas (such as the right side, left side, top, and bottom) of each image, multiply the pre-computed weight matrix with the image data element by element, so as to update the pixel values of the images to be stitched element by element, and finally complete the fusion of the images.

[0135] As an alternative implementation manner of the above step S132a, the implementation manner of obtaining the weight matrix may include: for the stitching direction of each image to be stitched (up, down, left, right, and the stitching direction is different according to the segmentation information of different segmented graphs), generate a weight template using the preset fusion width, where the weight template can be a linear template, a Gaussian weight template, or a cosine weight template. Then, use the linear template, Gaussian weight template, or cosine weight template to generate the weight matrix. The above-mentioned linear template may include a set of weight values calculated to gradually change from 0 to 1 according to the preset fusion width. Among them, the Gaussian weight template is a weight matrix generated based on the Gaussian function, and its characteristic is that the weight in the center is high and the weight at the edge is low, which can achieve smooth transition. The cosine weight template is a weight matrix generated based on the cosine function or exponential function, and its characteristic is that the weight decreases smoothly from the center to the edge, and is suitable for scenarios that require natural transition.

[0136] Optionally, at the position of the splicing line on the side close to the bus bar area in the initial splicing line, move it towards the side of the bus bar area, so that after the final spliced image shows the bus bar area, it is also possible to perform bus bar recognition on the spliced image to identify whether the bus bars in the bus bar area are offset or missing, etc., for quality inspection operations, thereby effectively improving the quality inspection efficiency of photovoltaic modules.

[0137] Please refer to Figure 14 the schematic structural diagram of the photovoltaic module image splicing device provided by the embodiment of the present application shown in; an embodiment of the present application provides a photovoltaic module image splicing device 200, including:

[0138] A splicing line determination module 210, configured to determine an initial splicing line from the to-be-spliced images, where the to-be-spliced images are partial images of a photovoltaic module, and the initial splicing line represents the boundary of the solar cells in the to-be-spliced images close to the splicing position.

[0139] A splicing line adjustment module 220, configured to adjust the position of the initial splicing line according to the target object to obtain the adjusted splicing line, where the target object includes: solar cells, gap regions, bus bar regions, and / or segmentation information, and the segmentation information represents the relative position relationship of the to-be-spliced images.

[0140] An image splicing module 230, configured to splice the to-be-spliced images according to the adjusted splicing line to obtain the spliced image.

[0141] As an optional implementation manner of the above device, the target object includes segmentation information; the splicing line determination module includes:

[0142] An initial splicing line determination sub-module, configured to determine an initial splicing line from the to-be-spliced images according to the segmentation information of the to-be-spliced images.

[0143] The splicing line adjustment module includes:

[0144] A direction pixel determination sub-module, configured to determine the moving direction and the number of pixels of the initial splicing line according to the segmentation information.

[0145] An initial splicing line moving sub-module, configured to move the position of the initial splicing line according to the moving direction and the number of pixels of the initial splicing line.

[0146] As an optional implementation manner of the above device, the target object includes: the unpowered solar cells in the to-be-spliced images; the above-mentioned splicing line adjustment module includes:

[0147] The first spliced line moving sub-module is used to move the position of the spliced line on the side of the initial spliced line close to the unpowered cell towards the side of the unpowered cell by a preset number of pixel intervals, so as to obtain the moved spliced line. The preset number of pixel intervals is an empirical value set according to the gap width between cells.

[0148] As an alternative implementation of the above device, the target object includes a gap area; the spliced line adjustment module includes:

[0149] The second spliced line moving sub-module is used to move the initial spliced line towards the inner side of the cell by a first preset distance to reserve the gap area between adjacent cells, where the first preset distance is less than the size of the cell in the moving direction; or, move the initial spliced line towards the outer side of the cell by a second preset distance to reserve the gap area between adjacent cells, where the second preset distance is greater than the gap width between cells and less than the sum of the size of the adjacent cell in the moving direction and the gap width.

[0150] As an alternative implementation of the above device, the target object includes: the bus bar area in the image to be spliced; the above-mentioned spliced line adjustment module includes:

[0151] The third spliced line moving sub-module is used to move the position of the spliced line on the side of the initial spliced line close to the bus bar area towards the side of the bus bar area to obtain the moved spliced line.

[0152] As an alternative implementation of the above device, the third spliced line moving sub-module includes:

[0153] The third spliced line moving unit is used to move the position of the spliced line on the side of the initial spliced line close to the bus bar area towards the side of the bus bar area by the number of pixels of the area width; where the number of pixels of the area width is an empirical value set according to the width of the bus bar area in the image, or the number of pixels of the area width is determined by the bus bar area detected from the image to be spliced by the target detection neural network model, or the number of pixels of the area width is determined by the bus bar area segmented from the image to be spliced by the image segmentation model.

[0154] As an alternative implementation of the above device, the image splicing module includes:

[0155] The image alignment sub-module is used to align the image to be spliced according to the adjusted spliced line to obtain the aligned image.

[0156] The weighted fusion sub-module is used to perform weighted fusion on the aligned images to obtain the spliced image.

[0157] As an alternative implementation of the above device, the weighted fusion sub-module includes:

[0158] A weight matrix acquisition unit is used to acquire a weight matrix, where the weight matrix is generated based on a Gaussian weight template or a cosine weight template with a preset fusion width;

[0159] The image weighted fusion unit is used to perform weighted fusion on the aligned images using a weight matrix.

[0160] It should be understood that this device corresponds to the aforementioned photovoltaic module image stitching method embodiment and is capable of executing each of the steps involved in the aforementioned method embodiment. The specific functions of this device can be found in the description above, and a detailed description is omitted here. The device includes at least one software functional module that can be stored in a memory in the form of software or firmware or embedded in the device's operating system (OS).

[0161] An electronic device provided by an embodiment of the present application includes: a processor and a memory, the memory storing machine-readable instructions executable by the processor, and the above method being executed when the machine-readable instructions are executed by the processor.

[0162] The embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to execute the above method. The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0163] An embodiment of the present application further provides a computer program product, including: a computer program or computer instructions, which executes the method described above when the computer program or computer instructions are executed by a processor.

[0164] It should be noted that the various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the similarities between the various embodiments, reference can be made to each other. For the device embodiments, since they are basically similar to the method embodiments, they are described relatively simply. For the relevant parts, reference can be made to the partial descriptions of the method embodiments.

[0165] In several embodiments provided by the embodiments of the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of the devices, methods, and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code. A module, a program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may also occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, which mainly depends on the functions involved.

[0166] In addition, in each of the embodiments of the present application, the various functional modules can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. Moreover, in the description of this specification, the descriptions with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the embodiments of the present application. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples.

[0167] The above description is only an alternative implementation manner of the embodiments of the present application, but the protection scope of the embodiments of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the embodiments of the present application, and all should be covered within the protection scope of the embodiments of the present application.

Claims

1. A method for splicing photovoltaic module images, characterized in that, Including: Determine an initial splicing line from the image to be spliced, where the image to be spliced is a partial image of a photovoltaic module, the initial splicing line characterizes the boundary of the solar cell near the splicing position in the image to be spliced, and the photovoltaic module includes a plurality of solar cell arrays; Adjust the position of the initial splicing line according to a target object to obtain an adjusted splicing line. The target object includes: solar cells, gap regions, bus bar regions, and / or segmentation information, where the segmentation information characterizes the relative position relationship of the image to be spliced, and the segmentation information includes at least one of the following: the order of the solar cell array where the image to be spliced is located, and the position data information of the image to be spliced in the solar cell array; Splice the image to be spliced according to the adjusted splicing line to obtain a spliced image; Among them, determining the initial splicing line from the image to be spliced includes: determining the initial splicing line from the image to be spliced according to the segmentation information; adjusting the position of the initial splicing line according to the target object includes: determining the moving direction and the number of pixels of the initial splicing line according to the segmentation information; moving the position of the initial splicing line according to the moving direction and the number of pixels of the initial splicing line.

2. The method according to claim 1, wherein The target object includes: the unpowered solar cells in the image to be spliced; adjusting the position of the initial splicing line according to the target object includes: Moving the position of the splicing line on the side of the initial splicing line close to the unpowered solar cell by a preset distance in pixels to the side of the unpowered solar cell to obtain a moved splicing line, and the preset distance in pixels is an empirical value set according to the gap width between solar cells.

3. The method according to claim 1, characterized in that The target object includes a gap region; adjusting the position of the initial splicing line according to the target object includes: Moving the initial splicing line inward by a first preset distance in the direction of the solar cell to retain the gap region between adjacent solar cells, where the first preset distance is less than the size of the solar cell in the moving direction; Or, moving the initial splicing line outward by a second preset distance in the direction of the solar cell to retain the gap region between adjacent solar cells, where the second preset distance is greater than the gap width between solar cells and less than the sum of the size of adjacent solar cells and the gap width in the moving direction.

4. The method according to claim 1, wherein The target object includes: the bus bar region in the image to be spliced; adjusting the position of the initial splicing line includes: Moving the position of the splicing line on the side of the initial splicing line close to the bus bar region to the side of the bus bar region to obtain a moved splicing line.

5. The method according to claim 4, characterized in that, The moving the position of the splicing line on the side of the initial splicing line close to the bus bar region to the side of the bus bar region includes: Move the position of the splicing line on the side of the initial splicing line close to the bus bar area towards the side of the bus bar area by the number of pixels of the area width; wherein, the number of pixels of the area width is an empirical value set according to the width of the bus bar area in the image, or the number of pixels of the area width is determined by a target detection neural network model from the bus bar area detected in the image to be spliced, or the number of pixels of the area width is determined by an image segmentation model from the bus bar area segmented from the image to be spliced.

6. The method according to claim 1, wherein The splicing the image to be spliced according to the adjusted splicing line includes: Align the image to be spliced according to the adjusted splicing line to obtain an aligned image; Perform weighted fusion on the aligned image to obtain the spliced image.

7. The method according to claim 6, wherein Performing weighted fusion on the aligned image includes: Obtain a weight matrix, where the weight matrix is generated according to a Gaussian weight template or a cosine weight template of a preset fusion width; Perform weighted fusion on the aligned image using the weight matrix.

8. A photovoltaic module image stitching device, characterized in that, including: A splicing line determination module, configured to determine an initial splicing line from an image to be spliced, where the image to be spliced is a partial image of a photovoltaic module, the initial splicing line represents the boundary of the solar cell close to the splicing position in the image to be spliced, and the photovoltaic module includes a plurality of solar cell arrays; A splicing line adjustment module, configured to adjust the position of the initial splicing line according to a target object to obtain an adjusted splicing line, where the target object includes: a solar cell, a gap area, a bus bar area, and / or segmentation information, and the segmentation information represents the relative position relationship of the image to be spliced, and the segmentation information includes at least one of the following: the order of the solar cell array where the image to be spliced is located, and the position data information of the image to be spliced in the solar cell array; An image splicing module, configured to splice the image to be spliced according to the adjusted splicing line to obtain a spliced image; Wherein, determining the initial splicing line from the image to be spliced includes: determining the initial splicing line from the image to be spliced according to the segmentation information; adjusting the position of the initial splicing line according to the target object includes: determining the moving direction and the number of pixels of the initial splicing line according to the segmentation information; moving the position of the initial splicing line according to the moving direction and the number of pixels of the initial splicing line.

9. An electronic device, characterized in that, including: A processor and a memory, where the memory stores machine-readable instructions executable by the processor, and when the machine-readable instructions are run by the processor, the method according to any one of claims 1 to 7 is executed.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the method according to any one of claims 1 to 7 is executed.

11. A computer program product, characterized in that, including: A computer program or computer instructions, and when the computer program or the computer instructions are run by a processor, the method according to any one of claims 1 to 7 is executed.

Images