Photovoltaic cell panel surface color imaging device based on CIE and sRGB standards

Abstract

The invention discloses a CIE and sRGB standard-based photovoltaic cell panel surface color imaging device, and relates to the field of color imaging, and the device comprises a to-be-detected photovoltaic cell panel, an illumination light source, a color camera, a standard chromaticity conversion model and an sRGB conversion model. The illumination light source irradiates the photovoltaic cell panel to be detected at a set angle; the color camera is used for shooting a color RGB image of a to-be-photovoltaic cell panel and a color RGB image of a calibration color sample; the standard chromaticity conversion model is used for converting a color RGB image of a to-be-photovoltaic cell panel and a color RGB image of a calibration color sample from an RGB space to a CIEXYZ standard space; and the sRGB conversion model is used for converting the image converted to the CIEXYZ standard space to a standard sRGB space. According to the invention, standardized and high-efficiency imaging of the surface color of the photovoltaic panel assembly can be realized.

Description

Technical Field

[0001] The present application relates to the field of color imaging, and in particular to a photovoltaic panel surface color imaging device based on CIE and sRGB standards. Background Art

[0002] The surface color of photovoltaic panels is an important parameter representing the quality of photovoltaic panels. Photovoltaic panels made of different materials, such as monocrystalline silicon photovoltaic panels, polycrystalline silicon photovoltaic panels and amorphous silicon photovoltaic panels, have their own definite color parameter ranges. The change of surface color reflects the performance change of photovoltaic panels to a certain extent, among which the performance change includes the reduction of photoelectric conversion efficiency and the reduction of service life. Therefore, the detection of the surface color of photovoltaic panels is of great significance.

[0003] The detection of the surface color of existing photovoltaic panels usually adopts standard color measurement equipment, such as colorimeter, colorimetric spectrophotometer, spectroradiometer, etc. Although these technologies can measure the standardized color parameters of the surface of photovoltaic panels, such as CIEXYZ and CIE1976Lab of the International Commission on Illumination, they are not competent for large-area online color detection. In addition, the existing color measurement equipment cannot obtain the standardized color image of photovoltaic panels, which is not conducive to the comprehensive analysis and identification of the surface color and quality of photovoltaic panels.

[0004] The Japanese invention patent JP4153755B2 solar panel color difference detection device, although using a color camera to capture color images of solar panels, does not involve the standardization of image colors. Chinese invention patents CN205643190U a solar cell color and appearance defect detection device, CN106327463A solar cell color recognition method, CN203292095U solar cell color difference sorting device, CN216539617U a solar cell color automatic sorting device, CN107185854B photovoltaic cell color difference detection and color classification algorithm based on RGB channels, although all use color cameras to capture the surface color of photovoltaic cells, do not involve the standardization of image colors.

[0005] Therefore, in order to achieve standardized and efficient imaging of the surface color of photovoltaic panel components, thereby improving the comprehensive level of surface quality detection and analysis of photovoltaic panel components, it is urgently necessary to provide a photovoltaic panel surface color imaging device. Summary of the invention

[0006] The purpose of this application is to provide a photovoltaic panel surface color imaging device based on CIE and sRGB standards, which can achieve standardized and efficient imaging of the surface color of photovoltaic panel components.

[0007] To achieve the above objectives, this application provides the following solutions:

[0008] The present application provides a photovoltaic panel surface color imaging device based on CIE and sRGB standards, the photovoltaic panel surface color imaging device based on CIE and sRGB standards comprising: a photovoltaic panel to be tested, an illumination light source, a color camera, a standard chromaticity conversion model and an sRGB conversion model;

[0009] The illumination light source irradiates the photovoltaic cell panel to be tested at a set angle;

[0010] The color camera is used to capture the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample;

[0011] The standard chromaticity conversion model is used to convert the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample from the RGB space to the CIEXYZ standard space;

[0012] The sRGB conversion model is used to convert an image converted to the CIEXYZ standard space to the standard sRGB space.

[0013] Optionally, the photovoltaic cell panels to be tested include: crystalline silicon photovoltaic cell panels and amorphous silicon photovoltaic cell panels.

[0014] Optionally, the color camera includes: a digital RGB color camera, a CMYG color camera or a multi-spectral color camera.

[0015] Optionally, the standard chromaticity conversion model is a nonlinear mapping model that converts the RGB space into the CIEXYZ standard space.

[0016] Optionally, the nonlinear mapping model includes: a multi-term fitting model or an artificial neural network model.

[0017] Optionally, the sRGB conversion model is a mathematical model defined by relevant standards of the International Electrotechnical Commission IEC for converting the RGB space of a standard display into a standard CIEXYZ space.

[0018] Optionally, the photovoltaic panel surface color imaging device based on CIE and sRGB standards further includes: a chromaticity conversion model training system;

[0019] The chromaticity conversion model training system is used to train the model according to the training sample set to obtain a standard chromaticity conversion model.

[0020] Optionally, the chromaticity conversion model training system includes: an illumination light source, a color camera, a standard color card, a training sample collection process, and a model training process.

[0021] Optionally, the standard color card includes: a plurality of color samples; CIExy chromaticity coordinates of the plurality of color samples are evenly distributed in the color gamut space of sRGB.

[0022] Optionally, the model training process includes: polynomial fitting model training or artificial neural network model training.

[0023] According to the specific embodiments provided in this application, this application has the following technical effects:

[0024] The present application provides a photovoltaic panel surface color imaging device based on CIE and sRGB standards, which converts the color RGB image of the photovoltaic panel to be used and the color RGB image of the calibration color sample from the RGB space to the CIEXYZ standard space through the standard chromaticity conversion model; and converts the image converted to the CIEXYZ standard space to the standard sRGB space through the sRGB conversion model, thereby realizing the mutual conversion of RGB, XYZ, sRGB and other spaces, solving the problem of standardized and large-area measurement of the surface chromaticity of the photovoltaic panel, realizing standardized and efficient imaging of the surface color of the photovoltaic panel assembly, and thus being able to improve the work efficiency and level of detection and analysis of the surface characteristics of the photovoltaic panel assembly. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 This is a schematic diagram of the structure of a photovoltaic panel surface color imaging device based on CIE and sRGB standards in one embodiment of the present application;

[0027] Figure 2 This is a schematic diagram of the principle of a photovoltaic panel surface color imaging device using a BP artificial neural network model provided in one embodiment of the present application;

[0028] Figure 3 This is a schematic diagram of the principle of a photovoltaic panel surface color imaging device using a polynomial model provided in one embodiment of the present application;

[0029] Figure 4 A schematic diagram of the chromaticity coordinate distribution of the SG color card training sample provided in one embodiment of the present application;

[0030] Figure 5A schematic diagram of the chromaticity coordinate distribution of the Colorchecker24 color card training sample provided in one embodiment of the present application;

[0031] Figure 6 This is a schematic diagram of the topological structure of the BP artificial neural network model provided in one embodiment of the present application. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0034] In an exemplary embodiment, Figure 1 As shown, a photovoltaic panel surface color imaging device based on CIE and sRGB standards is provided, the device comprising: a photovoltaic panel to be tested, an illumination light source, a color camera, a standard chromaticity conversion model and an sRGB conversion model;

[0035] The illumination light source irradiates the photovoltaic cell panel to be tested at a set angle;

[0036] The color camera is used to capture the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample;

[0037] The standard chromaticity conversion model is used to convert the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample from the RGB space to the CIEXYZ standard space;

[0038] The sRGB conversion model is used to convert an image converted to the CIEXYZ standard space to the standard sRGB space.

[0039] As a specific embodiment, the lighting source includes natural light sources such as sunlight and daytime sky light, and also includes artificial light sources such as halogen lamps and LEDs, which have high color rendering properties and are evenly irradiated on the surface of the photovoltaic panel at a certain light-emitting angle.

[0040] The photovoltaic panels to be tested include but are not limited to: crystalline silicon photovoltaic panels and amorphous silicon photovoltaic panels.

[0041] The color camera includes but is not limited to: a digital RGB color camera, a CMYG color camera or a multi-spectral color camera; and takes color RGB images of the photovoltaic panel and the calibration color sample at a certain position and distance;

[0042] The standard chromaticity conversion model is a nonlinear mapping model for converting RGB space to CIEXYZ standard space. The nonlinear mapping model includes but is not limited to: a multi-term fitting model or an artificial neural network model.

[0043] The sRGB conversion model is a mathematical model defined in relevant standards of the International Electrotechnical Commission (IEC) for converting the RGB space of a standard display into a standard CIEXYZ space.

[0044] The photovoltaic panel surface color imaging device based on CIE and sRGB standards also includes: a chromaticity conversion model training system; the chromaticity conversion model training system is a system for establishing a color conversion model for a specified color camera;

[0045] The chromaticity conversion model training system is used to train the model according to the training sample set to obtain a standard chromaticity conversion model.

[0046] The chromaticity conversion model training system includes: an illumination source, a color camera, a standard color card, a training sample collection process, and a model training process. Among them, the illumination source and the color camera used in the collection process of the RGB image of the training sample are the same as the illumination source and the color camera of the actual shooting system.

[0047] The standard color card includes: a variety of color samples; the CIExy chromaticity coordinates of the various color samples are evenly distributed in the sRGB color gamut space.

[0048] The XYZ data of the training samples are measured using CIE standard illuminants and geometric conditions, including D65 illuminant and de:8° or di:8° geometric conditions.

[0049] The model training process includes: polynomial fitting model training or artificial neural network model training.

[0050] like Figure 1 As shown in FIG. 1 , the illumination light source evenly illuminates the surface of the standard color card at a certain angle, and the standard color card contains n color blocks; the color camera captures the image of the standard color card at a certain angle (for example, the front), and obtains the RGB values ​​of the n color blocks. These n RGB values ​​are the input values ​​of the n training samples of the training sample set (for example, Figure 1On the other hand, the XYZ values ​​of the n color blocks of the standard color card are known. For example, an artificial neural network model or a polynomial model uses these n sets of training samples to train the model, thereby establishing a chromaticity conversion model from RGB space to XYZ space.

[0051] The device provided by this application is described below through two embodiments:

[0052] Embodiment 1

[0053] like Figure 2 As shown, this embodiment provides a photovoltaic panel surface color imaging device using a BP artificial neural network model, including: a photovoltaic panel, a daylight lighting source, an RGB color camera, a BP neural network conversion model, an sRGB conversion model, and a BP neural network training system. The photovoltaic panel presents color under the illumination of daylight; the RGB color camera captures the RGB image of the photovoltaic panel at a certain orientation and angle; the BP neural network converts the RGB image of the color camera into a CIE standard XYZ (CIE1931XYZ standard or CIE964XYZ standard tristimulus value space) image; the sRGB conversion model converts the CIE standard XYZ image into a standard sRGB image. Among them, the sRGB conversion model adopts the relevant standards of the International Electrotechnical Commission IEC, as shown in formula (1):

[0054]

[0055] The BP neural network training system uses the SG color card as a training sample. Under the same daylight lighting environment and shooting geometry conditions, the RGB values ​​of the SG color card are captured by the same RGB color camera, and then combined with the CIE standard XYZ values ​​of the SG color card to form a training sample pair. Figure 4 The CIE1931xy chromaticity coordinate distribution of 96 SG color card training samples is shown, which are measured under D65 illuminant and de:8° geometric conditions.

[0056] Figure 6 It shows the topology of the BP artificial neural network model. Figure 6 In, l n represents the node number of the nth hidden layer; w n (ln,ln+1) represents the link-weight between the nth hidden layer and the n+1th hidden layer; θ (n,ln) Indicates the threshold value of the corresponding node; θ X ,θ Y ,θ Z Indicates the threshold of the corresponding node in the output layer; a (n,ln) Represents the activation value of the corresponding node.

[0057] The BP neural network is trained by using the error back propagation algorithm and a BP neural network chromaticity conversion model is established; wherein, formula (2) represents the calculation formula of the XYZ chromaticity value based on the BP neural network, and formula (3) represents the feedforward process formula of the BP neural network, wherein f(x) represents the node activation function.

[0058]

[0059]

[0060] Embodiment 2

[0061] like Figure 3 As shown, this embodiment provides a photovoltaic panel surface color imaging device using a polynomial conversion model, including: a photovoltaic panel, a daylight lighting source, an RGB color camera, a polynomial conversion model, an sRGB conversion model and a polynomial model training system.

[0062] Photovoltaic panels present colors under sunlight; RGB color cameras take RGB images of photovoltaic panels at certain positions and angles; the polynomial conversion model converts the RGB images of the color camera into XYZ images of the CIE standard; and the sRGB conversion model converts the XYZ images of the CIE standard into standard sRGB images. The sRGB conversion model is shown in formula (1).

[0063] The training system of the polynomial transformation model uses the Colorchecker-24 color card as a training sample. Under the same daylight lighting environment and shooting geometry conditions, the RGB values ​​of the Colorchecker-24 color card are captured by the same RGB color camera, and then combined with the CIE standard XYZ values ​​of the Colorchecker-24 color card to form a training sample pair. Figure 4 The CIE1931xy chromaticity coordinates distribution of 24 Colorchecker-24 color chart training samples are shown, measured under D65 illuminant, de:8° geometry.

[0064] The polynomial transformation model is a training fitting model, and its matrix expression is shown in formula (4).

[0065]

[0066] In formula (4), A is a 3×n matrix, and n is the number of terms in the polynomial. For example, n=9 for a 2nd-order polynomial, n=19 for a 3rd-order polynomial, and n=27 for a 4th-order polynomial.

[0067] According to the known training sample set (R, G, B) → (X, Y, Z), the least square method is used to train formula (4) to determine the transformation matrix A. The calculation formula is as follows (5):

[0068] A=E(PD T )·E(DD T ) -1 (5)

[0069] In formula (5), D represents the random vector composed of the input end of the training sample set: D = (R, G, B, RG, RB, GB, ...), with a total of n elements; P represents the random vector composed of the output end of the training sample set: P = (X, Y, Z), with a total of 3 elements; E represents the mathematical expectation.

[0070] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0071] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A photovoltaic panel surface color imaging device based on CIE and sRGB standards, characterized in that: The photovoltaic panel surface color imaging device based on CIE and sRGB standards includes: a photovoltaic panel to be tested, an illumination light source, a color camera, a standard chromaticity conversion model and an sRGB conversion model; The illumination light source irradiates the photovoltaic cell panel to be tested at a set angle; The color camera is used to capture the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample; The standard chromaticity conversion model is used to convert the color RGB image of the photovoltaic panel to be calibrated and the color RGB image of the color sample from the RGB space to the CIEXYZ standard space; The sRGB conversion model is used to convert an image converted to the CIEXYZ standard space to the standard sRGB space.

2. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 1, characterized in that: The photovoltaic panels to be tested include: crystalline silicon photovoltaic panels and amorphous silicon photovoltaic panels.

3. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 1, characterized in that: The color camera includes: a digital RGB color camera, a CMYG color camera or a multi-spectral color camera.

4. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 1, characterized in that: The standard chromaticity conversion model is a nonlinear mapping model that converts the RGB space into the CIEXYZ standard space.

5. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 4, characterized in that: The nonlinear mapping model includes: a multi-term fitting model or an artificial neural network model.

6. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 1, characterized in that: The sRGB conversion model is a mathematical model defined in relevant standards of the International Electrotechnical Commission (IEC) for converting the RGB space of a standard display into a standard CIEXYZ space.

7. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 1, characterized in that: The photovoltaic panel surface color imaging device based on CIE and sRGB standards also includes: a chromaticity conversion model training system; The chromaticity conversion model training system is used to train the model according to the training sample set to obtain a standard chromaticity conversion model.

8. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 7, characterized in that: The chromaticity conversion model training system includes: an illumination source, a color camera, a standard color card, a training sample collection process and a model training process.

9. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 8, characterized in that: The standard color card includes: a variety of color samples; the CIExy chromaticity coordinates of the various color samples are evenly distributed in the sRGB color gamut space.

10. The photovoltaic panel surface color imaging device based on CIE and sRGB standards according to claim 8, characterized in that: The model training process includes: polynomial fitting model training or artificial neural network model training.

Images