Photovoltaic module image panoramic stitching method, storage medium and electronic device

Abstract

The application discloses a kind of photovoltaic module image panorama splicing method, storage medium and electronic equipment, it is related to image processing technical field, the method includes: obtaining multiple target images for target photovoltaic module;The joint point cloud information of the target image obtained is generated;Texture feature map of the target image obtained is extracted;The texture feature map and the joint point cloud information are registered, and the pixel value of the texture feature map is assigned to the orthographic projection image of the joint point cloud information based on registration result, form splicing image, to improve image splicing accuracy.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method for panoramic image stitching of photovoltaic modules, a storage medium, and an electronic device. Background Technology

[0002] Due to its numerous advantages such as being pollution-free, widely distributed, and renewable, solar energy is widely considered one of the most promising clean energy sources. In recent years, with the development of drone technology, the identification of component faults has increasingly relied on automated methods such as AI image recognition. However, in some large-scale centralized power plants, to ensure the safe flight of drones and the accurate location of photovoltaic module faults, it is necessary to perform full-site modeling of the entire power plant, i.e., to establish a digital map of the entire plant.

[0003] To achieve full-site modeling of a photovoltaic power plant, multiple consecutively captured images need to be stitched together using algorithms to create a single overall image of the plant; this is called panoramic stitching. Generally, some third-party platforms offer panoramic stitching software, but these are primarily based on pixel-level stitching. In complex geographical conditions, such as deserts and plateaus, the results are unsatisfactory, prone to jagged edges and misalignments, and the stitching accuracy is low. Summary of the Invention

[0004] This invention aims to at least partially address one of the technical problems in related technologies. Therefore, one objective of this invention is to provide a panoramic image stitching method, storage medium, and electronic device for photovoltaic modules to improve image stitching accuracy.

[0005] According to a first aspect of the present invention, a panoramic image stitching method for photovoltaic modules is provided, the method comprising:

[0006] Acquire multiple target images of the target photovoltaic module;

[0007] Generate joint point cloud information of the acquired target image;

[0008] Extract the texture feature map of the acquired target image;

[0009] The texture feature map and the joint point cloud information are registered, and the pixel values ​​of the texture feature map are assigned to the orthophoto image of the joint point cloud information based on the registration result to form a stitched image.

[0010] Optionally, the extraction of the texture feature map of the acquired target image includes:

[0011] Obtain texture feature points from each target image;

[0012] The depth map and normal map of the target image are determined based on the texture feature points;

[0013] A texture feature map is generated based on the depth map and normal map.

[0014] Optionally, obtaining texture feature points in each target image includes:

[0015] The scene partition to which each target image belongs is determined based on the image features in each target image;

[0016] The target images in each scene partition are merged, and texture feature points are obtained based on the merged image.

[0017] Optionally, after forming the stitched image, the method further includes:

[0018] Identify the target straight line in the stitched image;

[0019] Based on the preset length data of the photovoltaic panel, edge lines are selected from the target straight lines;

[0020] The slope of the edge line is adjusted based on the average slope of the edge line.

[0021] Optionally, generating the joint point cloud information of the acquired target image includes:

[0022] Extract image feature points from each target image;

[0023] The estimated camera parameters of each target image are obtained, and joint point cloud information is generated based on the image feature points and the estimated camera parameters.

[0024] Optionally, generating the joint point cloud information of the acquired target image includes:

[0025] Obtain pixels with consistent luminance in each target image;

[0026] Depth estimation information is generated based on the acquired pixels, and joint point cloud information is generated using the depth estimation information.

[0027] Optionally, the orthophoto image is obtained in the following manner:

[0028] The joint point cloud information is rasterized based on the digital elevation model to obtain the point cloud raster data corresponding to the joint point cloud information.

[0029] Convert the point cloud raster data into triangular mesh data;

[0030] The triangular mesh data is orthophoto-projected to obtain the orthophoto-projected image.

[0031] Optionally, converting the point cloud raster data into triangular mesh data includes:

[0032] Based on the finite element method, the reconstructed Poisson surface of the point cloud raster data is obtained;

[0033] The reconstructed Poisson surface is split to obtain triangular mesh data.

[0034] According to a second aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, wherein when the computer program is executed by a processor, the above-described panoramic image stitching method for photovoltaic modules is implemented.

[0035] According to a third aspect of the present invention, an electronic device is provided, comprising: a memory and a processor; the memory stores a computer program, which, when executed by the processor, implements the above-described panoramic image stitching method for photovoltaic modules.

[0036] As described above, the solution provided by the embodiments of the present invention does not directly stitch together the individual target images. Instead, it calculates joint point cloud information to map the scenes in multiple images onto the same point cloud, thereby achieving matching of corresponding positions between multiple target images. As a result, the orthophoto image generated by the joint point cloud information projection is panoramic. Furthermore, through the above matching process, the positions of the photovoltaic modules in the stitched image also have a more accurate relative relationship, thereby improving the stitching accuracy.

[0037] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating a panoramic image stitching method for photovoltaic modules provided in an embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0040] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0041] In one embodiment of the present invention, see Figure 1 A method for panoramic image stitching of photovoltaic modules is provided, including the following steps S101-S104.

[0042] S101: Acquire multiple target images of the target photovoltaic module;

[0043] S102: Generate joint point cloud information of the acquired target image;

[0044] S103: Extract the texture feature map of the acquired target image;

[0045] S104: Register the texture feature map and the joint point cloud information, and assign pixel values ​​of the texture feature map to the orthophoto image of the joint point cloud information based on the registration result to form a stitched image.

[0046] The target image can be a 2D image of the target photovoltaic module captured by a drone.

[0047] After acquiring the target image, it can be determined whether the size of the target image exceeds a preset size threshold. If it does, the target image that exceeds the threshold is segmented to obtain two or more sub-images below the size threshold.

[0048] In step S102, the step of generating joint point cloud information can be performed using OpenSfM or OpenMVS software. The following examples illustrate two ways of generating joint point cloud information.

[0049] In the first implementation, image feature points of each target image can be extracted; estimated camera parameters of each target image can be obtained; and joint point cloud information can be generated based on the image feature points and estimated camera parameters.

[0050] Specifically, the aforementioned image feature points can be feature points extracted by algorithms such as SIFT (Scale-Invariant Feature Transform), SURF (Speeded Up Robust Features), and ORB (OrientedFAST and Rotated BRIEF).

[0051] The estimated camera parameters are calculated based on the optical flow field or pixel gradient between each target image.

[0052] In this way, given the feature points and estimated camera parameters, matching feature point pairs can be found among the feature points of multiple target images. For each pair of feature points, the camera intrinsic parameters and relative camera pose in the estimated camera parameters are obtained, the normalized camera coordinates are calculated, and then the camera coordinates are transformed into three-dimensional coordinates of the point cloud through a triangulation algorithm, thereby constructing the joint point cloud information of each target image.

[0053] In one embodiment, the optimal camera intrinsic and extrinsic parameters can be generated using bundle adjustment to determine the aforementioned camera intrinsic parameters and camera relative pose.

[0054] In the second implementation, pixels with photometric consistency in each target image can be obtained; depth estimation information can be generated based on the obtained pixels, and joint point cloud information can be generated using the depth estimation information.

[0055] The above steps are based on the assumption of photometric consistency: the same spatial point should have the same photometric intensity (grayscale value or grayscale gradient) in images taken from different viewpoints. For example, in two target images of the same object taken at different times or angles, the brightness change of corresponding points should be as small as possible. Therefore, by setting a brightness change threshold, pixels with brightness changes less than the threshold can be selected as pixels with photometric consistency.

[0056] The depth of the aforementioned pixels is obtained by searching for matching points of the pixels in the reference image in the same row of the target image based on epipolar constraints, calculating the photometric error between the two, and thus determining the depth, thereby obtaining depth estimation information containing the depth.

[0057] Joint point cloud information can be established in the following way:

[0058] For each pixel in the depth estimation information, the pixel's coordinates are transformed to the world coordinate system by capturing the camera pose of the target image to which the pixel belongs. After processing each pixel in this way, a point cloud in the world coordinate system is formed, which is the joint point cloud information.

[0059] The number of points in the joint point cloud information obtained by the first implementation method is usually smaller than that of the second implementation method, so it is called sparse point cloud information. In contrast, the joint point cloud information obtained by the second implementation method is called dense point cloud information.

[0060] After generating the joint point cloud information, in one embodiment of the present invention, the joint point cloud information can be processed by an FPCFilter (Fast Point Cloud Filter) to remove outliers and noise points in the joint point cloud information.

[0061] In step S104, the registration process includes: setting the location information of the ground control point (GCP); when the target image contains the preset ground control point, the generated point cloud information contains the point cloud of the corresponding GCP. In this way, the registration matrix can be calculated based on the preset location information and the coordinates of the point cloud of the corresponding GCP, and based on the registration matrix, the texture feature map and the joint point cloud information are transformed to the same coordinate system to achieve registration, so that the texture of the texture feature map can be correctly assigned to the orthophoto image.

[0062] In one embodiment of the present invention, the orthophoto image is obtained in the following manner:

[0063] The joint point cloud information is rasterized based on the digital elevation model to obtain the point cloud raster data corresponding to the joint point cloud information; the point cloud raster data is converted into triangular mesh data; and the triangular mesh data is orthophoto-projected to obtain the orthophoto-projected image.

[0064] Specifically, the boundary range is calculated based on the input joint point cloud information, and the width and height of the joint point cloud information are calculated according to the specified resolution to form a digital elevation model (DEM) corresponding to the joint point cloud information. Each cell in the model contains a value that represents the ground elevation represented by that cell, corresponding to the point cloud height under different terrains in the joint point cloud information.

[0065] In this case, the DEM can be divided into multiple tiles according to the preset number of segments, all generated tiles can be verified and virtual grids can be created and merged to obtain point cloud grid data, and the grids of the DEM can be filled with gaps and smoothed.

[0066] The triangular mesh data is orthophotographed onto a two-dimensional plane to form an orthophoto image.

[0067] Accordingly, after determining the above two-dimensional plane, the registered texture feature map is transformed into the coordinate system of the joint point cloud information and projected onto the above two-dimensional plane. The color value of each triangle included in the texture feature map is assigned to each pixel of the orthophoto image in the two-dimensional plane by interpolation. After filling in the color information, the stitched image is finally formed.

[0068] As described above, the solution provided by the embodiments of the present invention does not directly stitch together the individual target images. Instead, it calculates joint point cloud information to map the scenes in multiple images onto the same point cloud, thereby achieving matching of corresponding positions between multiple target images. As a result, the orthophoto image generated by the joint point cloud information projection is panoramic. Furthermore, through the above matching process, the positions of the photovoltaic modules in the stitched image also have a more accurate relative relationship, thereby improving the stitching accuracy.

[0069] In one embodiment of the present invention, the texture feature map in S103 can be obtained in the following manner: obtaining texture feature points in each target image; determining the depth map and normal map of the target image based on the texture feature points; and generating the texture feature map based on the depth map and normal map.

[0070] Texture feature points are corner points and points representing edge positions extracted from each target image using the MVS-Texturing algorithm.

[0071] The following explains how to obtain a depth map.

[0072] During the initialization phase: a depth value is randomly generated for each pixel in the texture feature points, and iterative optimization is performed.

[0073] Iterative optimization: An improved PatchMatch scheme is used to optimize the depth value of each pixel to form the final depth map.

[0074] After obtaining the depth map, the depth values ​​within the window are used to fit a local plane, and the normal is calculated through the fitted local plane to form a normal map.

[0075] Edge detection is performed on the depth map and normal map. Since the depth map and normal map are not affected by lighting, edge points can be detected, forming a texture feature map.

[0076] Since there are multiple target images, a texture feature map can be obtained for each target image. Therefore, for each texture feature map, registration is performed separately with the joint point cloud information. During registration, the position of the preset texture features possessed by the GCP in each texture feature map can be detected, thereby obtaining the registration matrix.

[0077] Observation reveals that the images of photovoltaic panels exhibit a clear collinearity characteristic; that is, several adjacent photovoltaic panels lie on the same straight line, or the components within the same photovoltaic panel should also be on a straight line. Therefore, this prior knowledge can be used to optimize the stitching algorithm by correcting the images of non-collinear photovoltaic components and adjusting the slope of the photovoltaic panels' edges in the image.

[0078] Specifically, after the stitched image is formed, it can be processed as follows:

[0079] Identify target straight lines in the stitched image; filter out edge straight lines from the target straight lines based on the preset length data of the photovoltaic panels; adjust the slope of the edge straight lines based on the average slope of the edge straight lines.

[0080] First, after obtaining the stitched image, the image is corrected to ensure it is generally level. Most drone-captured images are already level during flight path planning; the few that do not meet this requirement can be manually adjusted.

[0081] Then, the LSD (Line Segment Detector) algorithm is used to identify the straight lines in the image as the target straight lines.

[0082] Based on the preset length data of the photovoltaic panel, the straight lines whose approximate length represented by the target straight line and the length data are within the preset range are retained as edge lines. When capturing the target image, different shooting heights can be set, and different empirical length data values ​​can be set at different shooting heights.

[0083] In this way, the selected edge lines can accurately represent the edges of the photovoltaic panel.

[0084] When adjusting the slope, the stitched image can be divided into c rows, with a row height of 50 pixels. For each of the c rows, calculate the slope of each line selected in step 3. Take the average slope. For lines with a large error between the slope and the average (based on experimental results, a threshold of 3%), adjust the slope of the corresponding component to the average.

[0085] This method fully utilizes the collinearity of photovoltaic panels, effectively improving the stitching accuracy of existing panoramic image stitching algorithms. It has strong applicability and good scalability, and can be widely used in application scenarios with collinearity, such as road, pipeline, and urban building inspection applications.

[0086] In one embodiment, texture feature points in each target image can be obtained in the following manner:

[0087] The scene partition to which each target image belongs is determined based on the image features in each target image;

[0088] The target images in each scene partition are merged, and texture feature points are obtained based on the merged image.

[0089] Regarding the determination of scene zoning, in the case of continuously capturing target images,

[0090] In two target images captured at adjacent times, the grayscale difference or color histogram similarity can be used as the basis for classifying them into different scene partitions.

[0091] During the merging process, two adjacent target images are merged when the difference in the mean gray level of two regions is less than a threshold. This reduces the processing of duplicate image information.

[0092] In the foregoing embodiments, point cloud raster data can be converted into triangular mesh data in the following manner:

[0093] Based on the finite element method, the reconstructed Poisson surface of the point cloud raster data is obtained; the reconstructed Poisson surface is then decomposed to obtain triangular mesh data.

[0094] For example, a vector field can be constructed from the normal information of point clouds in point cloud raster data, the Poisson equation can be solved to obtain continuous implicit functions, and then isosurfaces can be extracted to generate a surface model. The finite element method discretizes the continuous problem into finite elements and solves the Poisson square to obtain an approximate solution, thus enabling the discretization of the Poisson equation solution process.

[0095] After the Poisson surface is reconstructed, the initial mesh contained therein may lack detail, and the model quality needs to be improved by splitting and refining it.

[0096] The subdivision method can be based on curvature, such as prioritizing the subdivision of high curvature regions (e.g., edges) and reducing the subdivision of low curvature regions (e.g., planes); the subdivision method can also be Delaunay triangulation, etc., and the embodiments of the present invention do not limit this.

[0097] In one embodiment of the present invention, a computer-readable storage medium is also provided, on which a computer program is stored, wherein when the computer program is executed by a processor, the photovoltaic module image panoramic stitching method of any of the above embodiments is implemented.

[0098] In one embodiment of the present invention, an electronic device is also provided, including: a memory and a processor; the memory stores a computer program, and when the computer program is executed by the processor, it implements the panoramic image stitching method for photovoltaic modules of any of the above embodiments.

[0099] Figure 2 This is a structural block diagram of an electronic device according to an embodiment of the present invention.

[0100] like Figure 2 As shown, the electronic device 200 includes a processor 201 and a memory 203. The processor 201 and the memory 203 are connected, for example, via a bus 202. Optionally, the electronic device 200 may also include a transceiver 204. It should be noted that in practical applications, the transceiver 204 is not limited to one type, and the structure of this electronic device 200 does not constitute a limitation on the embodiments of the present invention.

[0101] Processor 201 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 201 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0102] Bus 202 may include a pathway for transmitting information between the aforementioned components. Bus 202 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 202 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 2 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0103] The memory 203 stores a computer program corresponding to the photovoltaic module image panoramic stitching method of the above embodiments of the present invention. This computer program is controlled and executed by the processor 201. The processor 201 executes the computer program stored in the memory 203 to implement the content shown in the aforementioned method embodiments.

[0104] Among them, electronic devices 200 include, but are not limited to: mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and in-vehicle terminals (such as in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 2 The electronic device 200 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0105] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0106] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0107] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0108] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0109] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0110] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0111] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0112] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for panoramic stitching of photovoltaic module images, characterized in that, The method comprises: acquiring a plurality of target images of a target photovoltaic module; generating joint point cloud information of the acquired target images; extracting texture feature maps of the acquired target images; registering the texture feature maps and the joint point cloud information, and assigning pixel values of the texture feature maps to orthographic projection images of the joint point cloud information based on registration results to form a stitched image; the generating of the joint point cloud information of the acquired target images comprises: extracting image feature points of each target image; acquiring estimated camera parameters of each target image, and generating joint point cloud information based on the image feature points and the estimated camera parameters; or, acquiring pixel points with consistent luminosity in each target image; generating depth estimation information based on the acquired pixel points, and generating joint point cloud information using the depth estimation information.

2. The method of claim 1, wherein, The extracting of the texture feature maps of the acquired target images comprises: acquiring texture feature points in each target image; the texture feature points are corner points and edge position points extracted from each target image using an MVS-Texturing algorithm; determining a depth map and a normal map of the target image according to the texture feature points; generating a texture feature map based on the depth map and the normal map.

3. The method of claim 2, wherein, The acquiring of the texture feature points in each target image comprises: determining scene partitions to which each target image belongs according to image features in each target image; merging target images in each scene partition, and acquiring texture feature points based on the merged images.

4. The method of claim 1, wherein, After the formation of the stitched image, the method further comprises: identifying target straight lines in the stitched image; screening edge straight lines from the target straight lines according to preset length data of photovoltaic panels; adjusting slopes of the edge straight lines based on an average value of the slopes.

5. The method of claim 1, wherein, The orthographic projection image is obtained in the following manner: performing rasterization processing on the joint point cloud information based on a digital elevation model to obtain point cloud raster data corresponding to the joint point cloud information; converting the point cloud raster data into triangular mesh data; performing orthographic projection on the triangular mesh data to obtain the orthographic projection image.

6. The method of claim 5, wherein, The conversion of the point cloud raster data into triangular mesh data comprises: obtaining a reconstructed Poisson surface of the point cloud raster data based on a finite element method; splitting the reconstructed Poisson surface to obtain triangular mesh data.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that The computer program is executed by a processor to implement the photovoltaic module image panoramic stitching method according to any one of claims 1-6.

8. An electronic device, comprising: comprises: a memory and a processor; a computer program is stored on the memory, and the computer program is executed by the processor to implement the photovoltaic module image panoramic stitching method according to any one of claims 1-6.

Images