Intelligent detection device and method based on charge-coupled device

Abstract

The invention provides an intelligent detection device and method based on a charge-coupled device, and the method comprises the steps: collecting an initial detection image through a charge-coupled device camera, carrying out the feature matching according to a comparison image sequence, and obtaining a plurality of comparison image defect features corresponding to the initial image defect features; based on the initial image defect feature, a plurality of contrast image defect features corresponding to the initial image defect feature and the change node of the environment temperature, carrying out credibility analysis to obtain an image defect credibility score corresponding to the initial image defect feature; and carrying out image position marking on the initial image defect features of which the image defect credibility scores are higher than a credibility threshold. According to the method, defect feature comparison can be carried out through the comparison images to obtain the credible score of each defect feature, so that the interference of random shadow noise on defect detection is reduced, and the accuracy of defect detection of the photovoltaic solar cell panel is improved.

Description

Technical Field

[0001] The present application relates to the field of image detection technology, and more specifically, to an intelligent detection device and method based on a charge coupled device. Background Art

[0002] With the vigorous development of the photovoltaic industry, related technical issues have also emerged. Photovoltaic solar panels are important components of photovoltaic power generation. Their quality is directly related to many aspects such as the photoelectric conversion rate and power generation life of photovoltaic power generation. Solar panels may also easily produce defects due to improper operation during production and installation, such as cracks, broken grids, and solid black. Solar panel defects not only affect the power generation efficiency of solar cells, but also pose serious safety hazards. Therefore, solar cell defect detection plays a vital role in the production and power generation process.

[0003] The existing technology mainly collects photovoltaic solar panel images and performs defect detection based on the panel images. However, the panel images are easily interfered with by environmental factors during the acquisition process, especially by light and shadow noise. In the process of defect detection on the panel images, light and shadow noise is easily mistaken for defective parts of the panel, which greatly affects the accuracy of photovoltaic solar panel defect detection. Summary of the Invention

[0004] The present application provides an intelligent detection device and method based on a charge-coupled device, which can obtain a confidence score for each defect feature by comparing defect features through image comparison, thereby reducing the interference of random light and shadow noise on defect detection and improving the accuracy of defect detection on photovoltaic solar panels.

[0005] In a first aspect, the present application provides an intelligent detection method based on a charge-coupled device, which can be executed by a network device, or can also be executed by a chip configured in the network device, and the present application does not limit this.

[0006] Specifically, the method includes:

[0007] After applying a test current to the photovoltaic solar cell panel, an initial test image is collected by a charge-coupled device camera, and defects are extracted from the initial test image to obtain an initial image defect feature set;

[0008] Adjusting the ambient temperature, and collecting a comparison image sequence according to the change nodes of the ambient temperature;

[0009] For any one of the initial image defect features in the initial image defect feature set, feature matching is performed according to the comparison image sequence to obtain a plurality of comparison image defect features corresponding to the initial image defect feature;

[0010] Performing a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then determining image defect credibility scores corresponding to other initial image defect features in the same manner;

[0011] A credible threshold is determined according to the grayscale rising gradient of each contrast image in the contrast image sequence, and image positions of initial image defect features having an image defect credibility score higher than the credible threshold are marked.

[0012] In conjunction with the first aspect, in certain implementations of the first aspect, performing defect extraction on the initial detection image to obtain the initial image defect feature set specifically includes:

[0013] Acquire the initial detection image, perform edge detection on the initial detection image, perform background segmentation based on the extracted edge feature points of the solar panel object, and obtain a solar panel object image;

[0014] A threshold segmentation algorithm is used to extract defect features from the panel object image to obtain a plurality of initial image defect features, wherein each initial image defect feature includes defect position coordinate information, defect boundary perimeter information, and defect boundary shape information, and all initial image defect features are combined into the initial image defect feature set.

[0015] In conjunction with the first aspect, in certain implementations of the first aspect, performing feature matching according to the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature specifically includes:

[0016] Acquire each comparison image in the comparison image sequence, perform defect detection on any comparison image, obtain multiple defect areas of the comparison image, and perform feature matching with each defect area according to the defect position coordinate information, defect boundary perimeter information and defect boundary shape information in the defect feature of the initial image, and obtain the defect area with the highest matching score as the comparison image defect feature corresponding to the comparison image.

[0017] In conjunction with the first aspect, in certain implementations of the first aspect, performing a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the ambient temperature change node to obtain an image defect credibility score corresponding to the initial image defect feature specifically includes:

[0018] Obtaining the boundary regularity of the defect feature of the initial image;

[0019] Obtaining boundary regularity corresponding to a plurality of comparison image defect features corresponding to the initial image defect feature;

[0020] The shape regularity of the initial image defect feature is used as the first sequence element, and the shape regularities corresponding to the multiple comparative image defect features corresponding to the initial image defect feature are used as subsequent sequence elements to construct an electroluminescent feature defect sequence;

[0021] The correlation coefficient between the electroluminescent characteristic defect sequence and the corresponding ambient temperature change node is obtained as the image defect credibility score.

[0022] In conjunction with the first aspect, in certain implementations of the first aspect, determining the credible threshold according to the grayscale rising gradient of each contrast image in the contrast image sequence specifically includes:

[0023] Obtaining the grayscale mean corresponding to each contrast image in the contrast image sequence;

[0024] The grayscale rising gradient is determined according to the grayscale mean values ​​corresponding to each contrast image, and the credible threshold is obtained by mapping the interval range where the grayscale rising gradient is located.

[0025] In combination with the first aspect, in certain implementations of the first aspect, after marking the image position of the initial image defect feature whose image defect credibility score is higher than the credibility threshold, the method further includes: using the marked position as the defect feature of the photovoltaic solar panel, classifying the defect feature of the photovoltaic solar panel, and uploading it to a cloud storage server.

[0026] In combination with the first aspect, in some implementations of the first aspect, after collecting the initial detection image through the charge coupled device camera, it also includes: after grayscale processing the initial detection image, using median filtering to filter noise on the initial detection image.

[0027] In a second aspect, the present application provides an intelligent detection device based on a charge coupled device, comprising:

[0028] An image acquisition module is used to acquire an initial detection image through a charge-coupled device camera after applying a detection current to the photovoltaic solar cell panel, and to extract defects from the initial detection image to obtain a defect feature set of the initial image;

[0029] The image acquisition module is further used to adjust the ambient temperature and acquire a comparison image sequence according to the change nodes of the ambient temperature;

[0030] an image processing module, configured to perform feature matching on any one of the initial image defect feature sets according to the comparison image sequence, to obtain a plurality of comparison image defect features corresponding to the initial image defect feature;

[0031] The image processing module is further configured to perform a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then determine image defect credibility scores corresponding to other initial image defect features in the same manner;

[0032] The image position marking module is used to determine a credible threshold according to the grayscale rising gradient of each contrast image in the contrast image sequence, and mark the image position of the initial image defect feature whose image defect credible score is higher than the credible threshold.

[0033] In a third aspect, the present application provides a computer terminal device, which includes a memory and a processor, wherein the memory stores a code, and the processor is configured to obtain the code and execute the above-mentioned intelligent detection method based on a charge-coupled device.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium, which stores at least one computer program, and the computer program is loaded and executed by a processor to implement the operations performed by the above-mentioned intelligent detection method based on a charge-coupled device.

[0035] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0036] The present application provides an intelligent detection device and method based on a charge-coupled device. First, after applying a detection current to a photovoltaic solar cell panel, an initial detection image is captured by a charge-coupled device camera, and defects are extracted from the initial detection image to obtain an initial image defect feature set; the ambient temperature is adjusted, and a comparison image sequence is captured according to the change nodes of the ambient temperature; for any initial image defect feature in the initial image defect feature set, feature matching is performed according to the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature; a credibility analysis is performed based on the initial image defect feature, the multiple comparison image defect features corresponding to the initial image defect feature, and the change nodes of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then the image defect credibility scores corresponding to other initial image defect features are determined in the same manner; a credibility threshold is determined according to the grayscale rising gradient of each comparison image in the comparison image sequence, and the image position of the initial image defect feature with an image defect credibility score higher than the credibility threshold is marked.

[0037] It can be seen that the present application can simulate the electroluminescence effect of photovoltaic solar panels under different temperature conditions by adjusting the ambient temperature and collecting comparison image sequences at different temperature nodes, thereby providing more information for subsequent defect feature matching and helping to identify light and shadow interference caused by environmental changes. According to the physical characteristics of electroluminescence, the electroluminescence effect of photovoltaic solar panels increases with changes in temperature, and the boundary regularity of the defect position shows a certain regularity as the temperature rises, indicating the authenticity of the defect, rather than interference caused by ambient light and shadow noise. The present application matches similar comparison image defect features at multiple temperature nodes based on the defect features of the initial image. Based on the matching results between the initial image defect features and the comparison image defect features, and combined with the temperature change nodes, the credibility score of each defect feature is determined, thereby effectively distinguishing between false matches caused by environmental noise (such as light and shadow changes) and true defect features.

[0038] In summary, this application calculates the credibility score of each defect feature based on the matching results between the defect features of the initial image and the defect features of the comparison image, and combines the temperature change nodes. It can obtain the credibility score of each defect feature by comparing the defect features of the comparison image, thereby reducing the interference of random light and shadow noise on defect detection and improving the accuracy of defect detection of photovoltaic solar panels. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is an exemplary flow chart of an intelligent detection method based on a charge coupled device according to some embodiments of the present application;

[0040] Figure 2is an exemplary flow chart for determining an image defect confidence score in some embodiments of the present application;

[0041] Figure 3 is a schematic structural diagram of an intelligent detection device based on a charge coupled device according to some embodiments of the present application;

[0042] Figure 4 It is a structural diagram of a computer terminal device that implements an intelligent detection method based on a charge-coupled device according to some embodiments of the present application. DETAILED DESCRIPTION

[0043] The present application applies a detection current to a photovoltaic solar panel, then captures an initial detection image using a charge-coupled device camera, performs defect extraction on the initial detection image, and obtains an initial image defect feature set; adjusts the ambient temperature, and captures a comparison image sequence based on the ambient temperature change nodes; performs feature matching on any initial image defect feature in the initial image defect feature set based on the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature; performs a credibility analysis based on the initial image defect feature, the multiple comparison image defect features corresponding to the initial image defect feature, and the ambient temperature change nodes to obtain an image defect credibility score corresponding to the initial image defect feature, and then uses the same method to determine the image defect credibility scores corresponding to other initial image defect features; determines a credibility threshold based on the grayscale rising gradient of each comparison image in the comparison image sequence, and marks the image position of the initial image defect features whose image defect credibility scores are higher than the credibility threshold. The present application can obtain a credibility score for each defect feature by performing defect feature comparison on the comparison images, thereby reducing the interference of random light and shadow noise on defect detection and improving the accuracy of photovoltaic solar panel defect detection.

[0044] In order to better understand the above technical solution, the following will be combined with the accompanying drawings and specific implementation methods to describe the above technical solution in detail. Figure 1 , which is an exemplary flow chart of an intelligent detection method based on a charge coupled device according to some embodiments of the present application. The intelligent detection method 100 based on a charge coupled device mainly includes the following steps:

[0045] In step S101, after applying a detection current to the photovoltaic solar cell panel, an initial detection image is collected by a charge coupled device camera, and defects are extracted from the initial detection image to obtain an initial image defect feature set.

[0046] It should be noted that when a solar panel is injected with external current, the battery material will emit light of a certain wavelength. This phenomenon is called electroluminescence. Healthy battery materials will emit light evenly, while defective areas (such as cracks, uneven doping, poor contact, etc.) will appear as abnormal changes in brightness. In some embodiments, the initial detection image described in this application is the electroluminescence image of the photovoltaic solar panel. Collecting electroluminescence (EL) images is a commonly used technology for defect detection of photovoltaic solar panels. The acquisition of EL defect images of solar cells is mainly based on the principle of electroluminescence and is used to detect and analyze tiny defects in solar cells. In specific implementation, first, ensure that the surface of the panel to be inspected is clean and free of obvious contamination. By applying an electric field, the photovoltaic peak wavelength between the energy bands of the battery components reaches about 1150nm, stimulating the electroluminescence effect in the solar panel. Finally, the spectrum is collected by a CCD camera to obtain the EL image of the solar cell.

[0047] The charge coupled device camera described in the present application is a camera that uses a charge coupled device (CCD) as a photosensitive element, and collects an initial detection image through a camera with a charge coupled device (CCD).

[0048] Optionally, in some embodiments, after acquiring the initial detection image by the charge coupled device camera, the method further includes: performing image preprocessing on the initial detection image;

[0049] Optionally, in some embodiments, performing image preprocessing on the initial detection image specifically includes:

[0050] After grayscale processing is performed on the initial detection image, median filtering is used to perform noise filtering on the initial detection image.

[0051] Optionally, in some embodiments, performing defect extraction on the initial detection image to obtain the initial image defect feature set specifically includes:

[0052] Acquire the initial detection image, perform edge detection on the initial detection image, perform background segmentation based on the extracted edge feature points of the solar panel object, and obtain a solar panel object image;

[0053] A threshold segmentation algorithm is used to extract defect features from the panel object image to obtain a plurality of initial image defect features, wherein each initial image defect feature includes defect position coordinate information, defect boundary perimeter information, and defect boundary shape information, and all initial image defect features are combined into the initial image defect feature set.

[0054] In a specific implementation, the Canny edge detection algorithm can be used to perform edge detection on the initial detection image, so as to extract the edge feature points of the panel object and use them for background segmentation to obtain the panel object image. Then, the Otsu algorithm is used to obtain the global threshold of the panel object image for threshold segmentation. The pixels in the panel object image that are below the global threshold are marked as defective pixels. After the contours of each defect area in the image are extracted through connected area analysis of the image and each is used as a separate initial image defect feature, the defect position coordinate information, defect boundary perimeter information, and defect boundary shape information in each initial image defect feature are recorded.

[0055] In step S102, the ambient temperature is adjusted, and a comparison image sequence is collected according to the change nodes of the ambient temperature.

[0056] It should be noted that temperature changes can affect the electroluminescent performance of photovoltaic solar panels. Specifically, the band gap of the semiconductor materials that make up photovoltaic solar panels typically decreases with increasing temperature because thermal vibrations in the crystal lattice affect the interatomic distances and alter the distribution of electronic energy levels. This decrease in band gap reduces the efficiency of radiative recombination, as some energy is converted into thermal vibrations rather than photon emission. Furthermore, temperature changes reduce carrier mobility, thereby reducing the probability that injected carriers can effectively participate in radiative recombination. This effect leads to a decrease in electroluminescent efficiency. Furthermore, temperature affects the competitive relationship between carriers in the non-radiative recombination process, thereby affecting electroluminescent efficiency.

[0057] In this application, by adjusting the ambient temperature and collecting comparative image sequences at different temperature nodes, the electroluminescence effect of photovoltaic solar panels under different temperature conditions can be simulated. This multi-temperature image acquisition provides more information for subsequent defect feature matching. It not only reflects the physical characteristics of the defect (such as the electroluminescence effect), but also helps identify light and shadow interference caused by environmental changes. In addition, the images collected at multiple temperature nodes have different lighting and temperature effects, which can help the system better distinguish between real defects and noise caused by lighting or temperature changes, making defect judgment through feature analysis such as regularity more accurate.

[0058] It should be noted that under normal temperature conditions, as the temperature increases, the electroluminescent effect increases. In some specific embodiments of the present application, a temperature adjustment range in which the electroluminescent effect increases can be selected, and the ambient temperature can be linearly adjusted using a fresh air temperature adjustment system. The electroluminescent images of the photovoltaic solar panel are collected at several equally spaced change nodes of the ambient temperature as comparison images, thereby obtaining a set of comparison image sequences for comparison with the image defect features in the initial detection image, thereby improving the accuracy of panel defect detection.

[0059] In step S103, for any one of the initial image defect feature sets, feature matching is performed according to the comparison image sequence to obtain a plurality of comparison image defect features corresponding to the initial image defect feature;

[0060] Optionally, in some embodiments, performing feature matching according to the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature specifically includes:

[0061] Acquire each comparison image in the comparison image sequence, perform defect detection on any comparison image, obtain multiple defect areas of the comparison image, and perform feature matching with each defect area according to the defect position coordinate information, defect boundary perimeter information and defect boundary shape information in the defect feature of the initial image, and obtain the defect area with the highest matching score as the comparison image defect feature corresponding to the comparison image.

[0062] In specific implementation, in each comparison image, the same method as the initial image defect detection can be used to extract the defect area (through threshold segmentation and contour detection) to obtain multiple defect areas in the image, and then for the defect area in each comparison image, the following matching is performed with the initial image defect feature: position matching: calculate the Euclidean distance between the initial defect position and the current defect position as the position matching value, perimeter matching: compare the difference in boundary perimeter as the perimeter matching value, shape matching: obtain the area ratio between the defect features as the shape matching value, and then perform weighted fusion of the position matching value, perimeter matching value and shape matching value according to the preset position weight, perimeter weight and shape weight to obtain the final matching score, and use the defect area with the highest matching score as the comparison image defect feature corresponding to the comparison image.

[0063] In step S104, a credibility analysis is performed based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then the image defect credibility scores corresponding to other initial image defect features are determined in the same way.

[0064] It should be noted that as the ambient temperature increases, the electroluminescent effect corresponding to the initial image defect feature at the same position increases. At this time, due to the greater difference in grayscale values ​​between the defective area of ​​the panel and the surrounding area, the boundary of the defective area of ​​the panel is clearer in the image, which improves the regularity of the defect feature of the comparative image obtained by image acquisition at the defect position. Therefore, the credibility score of the defect can be judged by analyzing the correlation between the regularity sequence and the temperature node. If the regularity increases regularly with increasing temperature, it indicates that the authenticity of the defect is very high. The probability of misjudgment of the defect at this location due to random light and shadow noise is very low, and the credibility score of the defect at this location is high. Therefore, in the process of defect detection of the panel, by performing credibility analysis on the defect features of the initial image, the interference of random light and shadow noise is reduced, and the accuracy of defect detection of the panel is improved.

[0065] Optionally, in some embodiments, reference Figure 2 As shown in FIG. 1 , this figure is an exemplary flow chart for determining an image defect credibility score in some embodiments of the present application. A credibility analysis is performed based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the ambient temperature change node to obtain an image defect credibility score corresponding to the initial image defect feature. Specifically, the process includes:

[0066] In step S1041, the boundary regularity of the initial image defect feature is obtained;

[0067] In step S1042, the boundary regularity corresponding to the plurality of comparative image defect features corresponding to the initial image defect feature is obtained;

[0068] In step S1043, the shape regularity of the initial image defect feature is used as the first sequence element, and the shape regularities of the multiple comparative image defect features corresponding to the initial image defect feature are used as subsequent sequence elements to construct an electroluminescent feature defect sequence;

[0069] In step S1044 , the correlation coefficient between the electroluminescent characteristic defect sequence and the corresponding ambient temperature change node is obtained as the image defect credibility score.

[0070] In specific implementation, the boundary regularity of the initial image defect feature can be calculated by the following formula: boundary regularity = defect boundary perimeter of the initial image defect feature^2 / (4×pi×defect boundary area of ​​the initial image defect feature), where pi is the natural pi. The boundary regularity corresponding to the comparison image defect feature can be calculated and obtained in the same way.

[0071] Optionally, in some embodiments, the ratio of the Pearson correlation coefficient between the electroluminescent characteristic defect sequence and the corresponding ambient temperature change node temperature value to a preset standard correlation coefficient is used as the image defect credibility score.

[0072] It should be noted that due to uneven illumination or changes in ambient light, false defects caused by changes in light and shadow may appear in the image, especially in low-contrast or high-reflection areas. Random light and shadow noise often has no fixed pattern and sometimes fluctuates with temperature changes. Based on the matching results between the initial image defect features and the comparison image defect features, and combined with the temperature change nodes, the credibility score of each defect feature is calculated. This application can effectively identify these random noises through the correlation analysis between regularity changes and temperature nodes. For example, if the regularity of a defect is significantly improved at multiple temperature nodes, and this change conforms to the expected pattern (such as light intensity changes uniformly with increasing temperature), then the defect feature is likely to be real, while random light and shadow noise cannot show this consistency. Therefore, by comparing the credibility analysis of image defect features, errors caused by illumination changes and environmental interference can be effectively filtered out.

[0073] In step S105 , a credible threshold is determined according to the grayscale rising gradient of each contrast image in the contrast image sequence, and image positions of initial image defect features having image defect credible scores higher than the credible threshold are marked.

[0074] It should be noted that in the contrast image sequence, the change in the grayscale rising gradient can reflect the change in the light intensity in the image. The grayscale rising gradient refers to the rate at which the average grayscale value (brightness) of a defect area changes with the increase in ambient temperature in the contrast image sequence. The larger the gradient, the more obvious the electroluminescent effect changes with temperature, and the more significant the brightness improvement of the defect area. When the grayscale rising gradient is large, the brightness contrast of the defect area is significantly enhanced, resulting in the defect boundary in the image becoming clearer and smoother. The improvement in regularity indicates that the characteristic performance of the defect is more stable, so its credibility is relatively higher. The defect with a large grayscale rising gradient indicates that its physical response is stronger and its credibility is higher. Therefore, a higher credibility threshold can be set for it, requiring the correlation between regularity and temperature node to be more significant. Conversely, for defects with a smaller grayscale rising gradient, the credibility threshold is relatively low. This application determines the credibility threshold based on the grayscale rising gradient of each contrast image in the contrast image sequence, so that as the grayscale rising gradient changes, the credibility threshold can be flexibly adjusted to better adapt to the characteristic performance of different types of defects.

[0075] Optionally, in some embodiments, determining the credible threshold according to the grayscale rising gradient of each contrast image in the contrast image sequence specifically includes:

[0076] Obtaining the grayscale mean corresponding to each contrast image in the contrast image sequence;

[0077] The grayscale rising gradient is determined according to the grayscale mean values ​​corresponding to each contrast image, and the credible threshold is obtained by mapping the interval range where the grayscale rising gradient is located.

[0078] In specific implementation, the interval range of the grayscale rising gradient can be mapped according to a preset mapping table to obtain the corresponding credible threshold. The mapping table can be calibrated according to historical experience, and this application does not limit this.

[0079] Optionally, in some embodiments, after marking the image position of the initial image defect feature whose image defect credibility score is higher than the credibility threshold, the method further includes: using the marked position as the defect feature of the photovoltaic solar panel, classifying the defect feature of the photovoltaic solar panel, and uploading it to a cloud storage server.

[0080] In a specific implementation, defect feature classification of the defect features of the photovoltaic solar cell panel specifically includes: using a support vector machine to learn and classify the defect features, thereby dividing each defect feature into different defect categories, and specific defect categories include: crack defects, hole defects, surface stain defects and material defects.

[0081] Optionally, in some embodiments, during the process of classifying the defect features of the photovoltaic solar panel and uploading the defect features to a cloud storage server, the defect feature classification results are stored as a JSON file.

[0082] In addition, in another aspect of the present application, in some embodiments, the present application provides an intelligent detection device based on a charge coupled device, referring to Figure 3 , which is a schematic structural diagram of a charge-coupled device-based intelligent detection device according to some embodiments of the present application. The charge-coupled device-based intelligent detection device 200 includes: an image acquisition module 201, an image processing module 202, and an image position marking module 203, which are described as follows:

[0083] The image acquisition module 201 is used in some embodiments of the present application to acquire an initial detection image through a charge-coupled device camera after applying a detection current to the photovoltaic solar panel, and to perform defect extraction on the initial detection image to obtain an initial image defect feature set;

[0084] The image acquisition module 201 is further configured to adjust the ambient temperature in some embodiments of the present application, and to acquire a comparison image sequence according to a change in the ambient temperature;

[0085] The image processing module 202 is configured in some embodiments of the present application to perform feature matching on any one of the initial image defect feature sets according to the comparison image sequence to obtain a plurality of comparison image defect features corresponding to the initial image defect feature;

[0086] In some embodiments of the present application, the image processing module 202 is further configured to perform a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the ambient temperature change node to obtain an image defect credibility score corresponding to the initial image defect feature, and then determine image defect credibility scores corresponding to other initial image defect features in the same manner.

[0087] The image position marking module 203 is used in some embodiments of the present application to determine a credible threshold based on the grayscale rising gradient of each contrast image in the contrast image sequence, and to mark the image position of the initial image defect features whose image defect credible score is higher than the credible threshold.

[0088] The above describes in detail an example of an intelligent detection device and method based on a charge-coupled device provided in an embodiment of the present application. It can be understood that in order to achieve the above functions, the corresponding device includes a hardware structure and / or software module corresponding to performing each function.

[0089] Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function in the application is executed in hardware or in a computer software-driven hardware manner depends on the specific application and design constraints of the technical solution. Therefore, professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0090] In addition, the present application also provides a computer terminal device, which includes a memory and a processor, the memory stores a code, and the processor is configured to obtain the code and execute the above-mentioned intelligent detection method based on a charge-coupled device.

[0091] In some embodiments, reference Figure 4 , which is a schematic diagram of the structure of a computer terminal device that implements an intelligent detection method based on a charge coupled device according to some embodiments of the present application. In the above embodiment, an intelligent detection method based on a charge coupled device can be implemented by Figure 4The computer terminal device 300 shown in FIG. 1 is implemented as shown in FIG. 1 , and the computer terminal device 300 includes at least one communication bus 301 , a communication interface 302 , a processor 303 and a memory 304 .

[0092] The processor 303 can be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more processors for controlling the execution of a charge-coupled device-based intelligent detection method in the present application.

[0093] The communication bus 301 may include a path for transmitting information between the aforementioned components.

[0094] The memory 304 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compact disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 304 may exist independently and be connected to the processor 303 via the communication bus 301. The memory 304 may also be integrated with the processor 303.

[0095] Memory 304 is used to store program code for implementing the solution of the present application, and is controlled by processor 303 for execution. Processor 303 is used to execute the program code stored in memory 304. The program code may include one or more software modules. In the above embodiment, the determination of the image defect credibility score can be implemented by processor 303 and one or more software modules in the program code in memory 304.

[0096] The communication interface 302 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc.

[0097] Optionally, the computer terminal device 300 may further include a power supply 305 for providing power to various devices or circuits in the real-time computer terminal device.

[0098] In a specific implementation, as an embodiment, a computer terminal device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. The processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0099] The computer terminal device can be a general-purpose computer terminal device or a dedicated computer terminal device. In a specific implementation, the computer terminal device can be a desktop computer, a portable computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of computer terminal device.

[0100] In addition, in other aspects of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores at least one computer program, and the computer program is loaded and executed by a processor to implement the operations performed by the above-mentioned intelligent detection method based on a charge-coupled device.

[0101] In summary, in an intelligent detection device and method based on a charge-coupled device disclosed in an embodiment of the present application, an initial detection image is first captured by a charge-coupled device camera, and defects are extracted from the initial detection image to obtain an initial image defect feature set; the ambient temperature is adjusted, and a comparison image sequence is captured according to the change nodes of the ambient temperature; for any initial image defect feature in the initial image defect feature set, feature matching is performed according to the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature; a trustworthy analysis is performed based on the initial image defect feature, the multiple comparison image defect features corresponding to the initial image defect feature, and the change nodes of the ambient temperature to obtain an image defect trustworthy score corresponding to the initial image defect feature, and then the image defect trustworthy scores corresponding to other initial image defect features are determined in the same manner; a trustworthy threshold is determined based on the grayscale rising gradient of each comparison image in the comparison image sequence, and the image position of the initial image defect feature with an image defect trustworthy score higher than the trustworthy threshold is marked. The present application can obtain a trustworthy score for each defect feature by comparing defect features with comparison images, thereby reducing the interference of random light and shadow noise on defect detection and improving the accuracy of defect detection of photovoltaic solar panels.

[0102] The above description is merely an embodiment of the present application. Common knowledge such as the specific technical solutions or features of the solutions is not described in detail herein. It should be noted that those skilled in the art may make various modifications and improvements without departing from the technical solution of the present application, and these modifications and improvements should also be considered within the scope of protection of the present application. These modifications and improvements will not affect the effectiveness of the implementation of the present application or the practical application of the patent.

[0103] The scope of protection claimed by this application shall be determined by the content of the claims. The specific embodiments and other descriptions in the specification may be used to interpret the content of the claims. Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of the invention. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application is intended to include such modifications and variations.

Claims

1. An intelligent detection method based on a charge coupled device, characterized in that: include: After applying a test current to the photovoltaic solar cell panel, an initial test image is collected by a charge-coupled device camera, and defects are extracted from the initial test image to obtain an initial image defect feature set; Adjusting the ambient temperature, and collecting a comparison image sequence according to the change nodes of the ambient temperature; For any one of the initial image defect features in the initial image defect feature set, feature matching is performed according to the comparison image sequence to obtain a plurality of comparison image defect features corresponding to the initial image defect feature; Performing a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then determining image defect credibility scores corresponding to other initial image defect features in the same manner; A credible threshold is determined according to the grayscale rising gradient of each contrast image in the contrast image sequence, and image positions of initial image defect features having an image defect credibility score higher than the credible threshold are marked.

2. The method according to claim 1, wherein Defect extraction is performed on the initial detection image to obtain the initial image defect feature set, which specifically includes: Acquire the initial detection image, perform edge detection on the initial detection image, perform background segmentation based on the extracted edge feature points of the solar panel object, and obtain a solar panel object image; A threshold segmentation algorithm is used to extract defect features from the panel object image to obtain a plurality of initial image defect features, wherein each initial image defect feature includes defect position coordinate information, defect boundary perimeter information, and defect boundary shape information, and all initial image defect features are combined into the initial image defect feature set.

3. The method according to claim 1, wherein Performing feature matching according to the comparison image sequence to obtain multiple comparison image defect features corresponding to the initial image defect feature specifically includes: Acquire each comparison image in the comparison image sequence, perform defect detection on any comparison image to obtain multiple defect areas of the comparison image, and perform feature matching with each defect area based on the defect position coordinate information, defect boundary perimeter information and defect boundary shape information in the defect feature of the initial image, and obtain the defect area with the highest matching score as the comparison image defect feature corresponding to the comparison image.

4. The method according to claim 1, wherein A credibility analysis is performed based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature. Specifically, the credibility score includes: Obtaining the boundary regularity of the defect feature of the initial image; Obtaining boundary regularity corresponding to a plurality of comparison image defect features corresponding to the initial image defect feature; The shape regularity of the initial image defect feature is used as the first sequence element, and the shape regularities corresponding to the multiple comparative image defect features corresponding to the initial image defect feature are used as subsequent sequence elements to construct an electroluminescent feature defect sequence; The correlation coefficient between the electroluminescent characteristic defect sequence and the corresponding ambient temperature change node is obtained as the image defect credibility score.

5. The method according to claim 1, wherein Determining the credible threshold according to the grayscale rising gradient of each contrast image in the contrast image sequence specifically includes: Obtaining the grayscale mean corresponding to each contrast image in the contrast image sequence; The grayscale rising gradient is determined according to the grayscale mean values ​​corresponding to each contrast image, and the credible threshold is obtained by mapping the interval range where the grayscale rising gradient is located.

6. The method according to claim 1, wherein After marking the image position of the initial image defect features whose image defect credibility score is higher than the credibility threshold, the method further includes: using the marked position as the defect feature of the photovoltaic solar panel, classifying the defect feature of the photovoltaic solar panel, and uploading it to a cloud storage server.

7. The method according to claim 1, wherein After the initial detection image is collected by the charge coupled device camera, the method further includes: performing grayscale processing on the initial detection image, and then performing noise filtering on the initial detection image by using a median filter.

8. An intelligent detection device based on a charge coupled device, characterized in that: include: An image acquisition module is used to acquire an initial detection image through a charge-coupled device camera after applying a detection current to the photovoltaic solar cell panel, and to extract defects from the initial detection image to obtain a defect feature set of the initial image; The image acquisition module is further used to adjust the ambient temperature and acquire a comparison image sequence according to the change nodes of the ambient temperature; an image processing module, configured to perform feature matching on any one of the initial image defect feature sets according to the comparison image sequence, to obtain a plurality of comparison image defect features corresponding to the initial image defect feature; The image processing module is further configured to perform a credibility analysis based on the initial image defect feature, multiple comparative image defect features corresponding to the initial image defect feature, and the change node of the ambient temperature to obtain an image defect credibility score corresponding to the initial image defect feature, and then determine image defect credibility scores corresponding to other initial image defect features in the same manner; The image position marking module is used to determine a credible threshold according to the grayscale rising gradient of each contrast image in the contrast image sequence, and mark the image position of the initial image defect feature whose image defect credible score is higher than the credible threshold.

9. A computer terminal device, characterized in that: The computer terminal device includes a memory and a processor, the memory stores codes, and the processor is configured to obtain the codes and execute the intelligent detection method based on a charge coupled device according to any one of claims 1 to 7.

10. A computer-readable storage medium storing at least one computer program, characterized in that: The computer program is loaded and executed by a processor to implement the operations performed by the intelligent detection method based on a charge coupled device according to any one of claims 1 to 7.

Images