Glass introduction risk assessment method, device, storage medium and electronic equipment

Abstract

Disclosed are a method, device, storage medium, and electronic device for assessing the risk of glass introduction. The method includes: obtaining first theoretical image color information corresponding to a first operating condition in which the optical path of a camera-captured image does not introduce glass to be assessed; determining second theoretical image color information corresponding to a second operating condition in which the optical path of the camera-captured image introduces glass to be assessed based on parameter value groups corresponding to multiple visible light wavelengths and the transmittance of the glass to be assessed corresponding to the multiple visible light wavelengths; determining image color change information of the second theoretical image color information relative to the first theoretical image color information; performing color adjustment on a first reference image matching the first operating condition based on the image color change information to obtain a first adjusted image matching the second operating condition; and determining an assessment result of the risk of glass introduction to be assessed based on the first adjusted image. The disclosed embodiments can effectively assess the risk of glass introduction.

Description

Technical Field

[0001] The present disclosure relates to driving technology, and in particular to a glass introduction risk assessment method, device, storage medium, and electronic device. Background Art

[0002] Autonomous driving technology is increasingly being used in mobile devices such as vehicles. This technology often requires cameras to capture images of the actual scene, such as lane markings, traffic lights, and pedestrians. These images are then used by autonomous driving algorithms to make driving decisions. In some cases, there's glass in front of the camera, requiring the camera to capture images through the glass. Summary of the Invention

[0003] There is glass in front of the camera. The introduction of glass will affect the image quality of the image collected by the camera, thereby affecting the safety of autonomous driving. Therefore, it is necessary to evaluate the risks of introducing glass.

[0004] In order to solve the above technical problems, the present disclosure provides a glass introduction risk assessment method, device, storage medium and electronic device to effectively assess the glass introduction risk.

[0005] According to one aspect of an embodiment of the present disclosure, a method for assessing the risk of glass introduction is provided, comprising:

[0006] Acquire first theoretical image color information corresponding to a first working condition in which the glass to be evaluated is not introduced into the optical path of the camera capturing the image;

[0007] Determining, based on parameter value groups corresponding to multiple visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the multiple visible light wavelengths, second theoretical image color information corresponding to a second operating condition of the glass to be evaluated introduced by the optical path of the camera to capture the image; wherein the parameter value group corresponding to any visible light wavelength includes parameter values ​​of multiple basic parameters affecting image color at that visible light wavelength;

[0008] determining image color change information of the second theoretical image color information relative to the first theoretical image color information;

[0009] performing color adjustment on a first reference image matching the first operating condition according to the image color change information to obtain a first adjusted image matching the second operating condition;

[0010] An introduction risk assessment result of the glass to be assessed is determined based on the first adjusted image.

[0011] According to another aspect of an embodiment of the present disclosure, a glass introduction risk assessment device is provided, comprising:

[0012] A first acquisition module is used to acquire first theoretical image color information corresponding to a first working condition in which the glass to be evaluated is not introduced into the optical path of the camera capturing the image;

[0013] A first determination module is configured to determine, based on parameter value groups corresponding to a plurality of visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the plurality of visible light wavelengths, second theoretical image color information corresponding to a second operating condition of the glass to be evaluated when the optical path through which the camera captures the image introduces the image; wherein the parameter value group corresponding to any visible light wavelength includes parameter values ​​of a plurality of basic parameters affecting image color at that visible light wavelength;

[0014] a second determining module, configured to determine image color change information of the second theoretical image color information determined by the first determining module relative to the first theoretical image color information acquired by the first acquiring module;

[0015] a first adjustment module, configured to perform color adjustment on a first reference image matching the first operating condition according to the image color change information determined by the second determination module, to obtain a first adjusted image matching the second operating condition;

[0016] The third determining module is configured to determine an introduction risk assessment result of the glass to be assessed based on the first adjustment image obtained by the first adjustment module.

[0017] According to another aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the above-mentioned glass introduction risk assessment method.

[0018] According to another aspect of the embodiments of the present disclosure, an electronic device is provided, the electronic device including:

[0019] processor;

[0020] a memory for storing instructions executable by the processor;

[0021] The processor is configured to read the executable instructions from the memory and execute the instructions to implement the above-mentioned glass introduction risk assessment method.

[0022] According to yet another aspect of the embodiments of the present disclosure, a computer program product is provided. When instructions in the computer program product are executed by a processor, the above-mentioned glass introduction risk assessment method is performed.

[0023] Based on the glass introduction risk assessment method, device, medium, electronic device and computer program product provided by the above-mentioned embodiments of the present disclosure, it is possible to obtain, according to the imaging principle, through theoretical calculation, the first theoretical image color information corresponding to the first working condition in which the optical path of the camera-captured image does not introduce the glass to be evaluated, and the second theoretical image color information corresponding to the second working condition in which the optical path of the camera-captured image introduces the glass to be evaluated. The image color change information of the second theoretical image color information relative to the first theoretical image color information can effectively characterize the image quality change of the image captured by the camera before and after the introduction of the glass to be evaluated. In this way, according to the image color change information, the first reference image matching the first working condition is color-adjusted to obtain the first adjusted image matching the second working condition. The first adjusted image can simulate the image quality of the image captured by the camera after the glass to be evaluated is introduced. Therefore, the first adjusted image can provide a very effective reference for determining the introduction risk assessment result of the glass to be evaluated, thereby effectively realizing the assessment of the introduction risk of the glass. In addition, in the process of conducting glass introduction risk assessment, the images involved are a first reference image matching the first working condition, and a first adjusted image obtained by color-adjusting the first reference image. That is, in the embodiment of the present disclosure, the camera does not need to perform actual image acquisition under the first working condition and the second working condition respectively, which is conducive to improving the evaluation efficiency and shortening the evaluation cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1-1 It is a grayscale image corresponding to the image captured by the camera when there is no glass in front of the camera in the related art.

[0025] Figure 1-2 It is a grayscale image corresponding to the image captured by the camera when there is glass in front of the camera in the related art.

[0026] Figure 2 4 is a flow chart of a glass introduction risk assessment method provided by some exemplary embodiments of the present disclosure.

[0027] Figure 3-1 Schematic diagram of wavelength-transmittance curves obtained for the glass to be evaluated in some exemplary embodiments of the present disclosure.

[0028] Figure 3-2 It is a partial schematic diagram of an information table in some exemplary embodiments of the present disclosure.

[0029] Figure 3-3 It is a schematic diagram of a relationship curve used to obtain a type of information in an information table in some exemplary embodiments of the present disclosure.

[0030] Figure 3-4 It is a schematic diagram of a relationship curve for obtaining another type of information in the information table in some exemplary embodiments of the present disclosure.

[0031] Figure 4 is a flowchart illustrating a method for determining color information of a second theoretical image in some exemplary embodiments of the present disclosure.

[0032] Figure 5 FIG. 4 is a flowchart illustrating a method of determining a first color influence superposition value to a third color influence superposition value in some exemplary embodiments of the present disclosure.

[0033] Figure 6 4 is a flowchart illustrating a method of determining the second theoretical image color information according to the first color impact superposition value to the third color impact superposition value in some exemplary embodiments of the present disclosure.

[0034] Figure 7 4 is a flowchart illustrating a method for determining color information of a first theoretical image in some exemplary embodiments of the present disclosure.

[0035] Figure 8 4 is a flowchart illustrating a method of determining the fourth color influence superposition value to the sixth color influence superposition value in some exemplary embodiments of the present disclosure.

[0036] Figure 9 FIG. 4 is a flow chart illustrating a method for determining an introduction risk assessment result of a glass to be assessed in some exemplary embodiments of the present disclosure.

[0037] Figure 10 This is a flowchart of a triggering method for performing color adjustment on a first reference image in some exemplary embodiments of the present disclosure.

[0038] Figure 11-1 This is one of the flowcharts of the method of determining the first preset change rate and the second preset change rate in some exemplary embodiments of the present disclosure.

[0039] Figure 11-2 This is the second flowchart of a method for determining the first preset change rate and the second preset change rate in some exemplary embodiments of the present disclosure.

[0040] Figure 12 4 is a flowchart of a method for determining the introduction risk assessment result of the glass to be assessed in some other exemplary embodiments of the present disclosure.

[0041] Figure 13 is a schematic diagram of experimental results in some exemplary embodiments of the present disclosure.

[0042] Figure 14 2 is a schematic structural diagram of a glass introduction risk assessment device provided by some exemplary embodiments of the present disclosure.

[0043] Figure 154 is a schematic diagram of modules involved in triggering an operation of performing color adjustment on a first reference image in some exemplary embodiments of the present disclosure.

[0044] Figure 16 4 is a schematic diagram of modules involved in determining a first preset change rate and a second preset change rate in some exemplary embodiments of the present disclosure.

[0045] Figure 17 is a structural diagram of an electronic device provided by some exemplary embodiments of the present disclosure. DETAILED DESCRIPTION

[0046] To explain the present disclosure, example embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. It should be understood that the present disclosure is not limited to the example embodiments.

[0047] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0048] Application Overview

[0049] When using autonomous driving technology, it is often necessary to capture images of the actual scene through cameras and then make driving decisions through autonomous driving algorithms.

[0050] In some cases, there is glass in front of the camera, for example, there is a windshield, privacy glass, etc. The glass in front of the camera will affect the image quality of the camera. For example, the glass will cause the image collected by the camera to appear yellow. In an optional example, when there is no glass in front of the camera, the grayscale image corresponding to the image collected by the camera can be seen in Figure 1-1 When there is glass in front of the camera, the grayscale image corresponding to the image collected by the camera can be seen in Figure 1-2 .

[0051] To prevent the introduction of glass from affecting the safety of autonomous driving, it is necessary to assess the risks of glass introduction. How to assess the risks of glass introduction is a question worthy of attention for those skilled in the art.

[0052] Exemplary Methods

[0053] Figure 2 4 is a flow chart of a glass introduction risk assessment method provided by some exemplary embodiments of the present disclosure. Figure 2 The method shown can be applied to electronic devices, such as servers, terminal devices, etc. Figure 2The illustrated method may include step 210 , step 220 , step 230 , step 240 , and step 250 .

[0054] Step 210 : obtaining first theoretical image color information corresponding to a first working condition in which the optical path of the camera for capturing the image does not introduce the glass to be evaluated.

[0055] It should be noted that the glass to be evaluated may refer to any glass that requires evaluation of the risks introduced, including but not limited to windshields, privacy glass, etc. The first operating condition in which the optical path of the camera to capture images does not introduce the glass to be evaluated may refer to a condition in which there is no glass to be evaluated in front of the camera, and the camera directly captures images of the visual environment without the need to capture images through the glass to be evaluated. The first theoretical image color information corresponding to the first operating condition may refer to the color information of the image captured by the camera under the first operating condition, obtained through theoretical calculation according to the imaging principle. The determination of the first theoretical image color information may not depend on the image actually captured by the camera.

[0056] Step 220 : Based on the parameter value groups corresponding to the multiple visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the multiple visible light wavelengths, determine the second theoretical image color information corresponding to the second working condition of the glass to be evaluated introduced by the optical path of the camera to capture the image. The parameter value group corresponding to any visible light wavelength includes the parameter values ​​of multiple basic parameters that affect the image color at that visible light wavelength.

[0057] Optionally, multiple visible light wavelengths can be selected from the wavelength range of [400nm, 700nm]. For example, you can start from 400nm in [400nm, 700nm] and select a wavelength every 5nm to obtain multiple visible light wavelengths. In this way, the multiple visible light wavelengths can be: 400nm, 405nm, 410nm, ..., 690nm, 695nm, 700nm. For another example, you can start from 410nm in [400nm, 700nm] and select a wavelength every 10nm to obtain multiple visible light wavelengths. In this way, the multiple visible light wavelengths can be: 410nm, 420nm, 430nm, ..., 680nm, 690nm, 700nm. Of course, the starting point is not limited to 400nm and 410nm, and the interval is not limited to 5nm and 10nm. The specific selection can be determined according to actual conditions, and the present disclosure does not limit this.

[0058] Optionally, for the glass to be evaluated, a curve of the relationship between wavelength and transmittance can be obtained, for example, Figure 3-1 The relationship curve in Figure 3-1 The horizontal axis represents wavelength, and the vertical axis represents transmittance. Figure 3-1The relationship curve in can be used to obtain the transmittance of the glass to be evaluated corresponding to multiple visible light wavelengths.

[0059] Optionally, multiple basic parameters that affect image color may include but are not limited to sensor response (corresponding to Figure 3-2 Sensor RGB response item in), sensor properties (corresponding to Figure 3-2 camera Total item in ), spectral response at a given color temperature (corresponding to Figure 3-2 The D75 color temperature spectrum item in the data, where D75 is a type of light source and can be replaced by other types of light sources, etc. The sensor response can be the overall quantum efficiency (QE) response of the sensor. The sensor response can include a sensor response corresponding to the R channel, a sensor response corresponding to the G channel, and a sensor response corresponding to the B channel. The sensor properties can include a lens (d lens) transmittance property, an infrared (IR) transmittance property, etc. The parameter value group corresponding to any visible light wavelength among multiple visible light wavelengths can include parameter values ​​of multiple basic parameters at the visible light wavelength, for example, a response value of the sensor response, a property value of the sensor property, a response value of the spectral response at a given color temperature, etc.

[0060] Optionally, the parameter value groups corresponding to the multiple visible light wavelengths can be obtained from Figure 3-2 The information in the information table under the item Sensor RGB response (which indicates the response value of the sensor response) can be obtained according to Figure 3-3 The relationship curve in the information table is obtained, and the information under the camera Total item (which represents the attribute value of the sensor attribute) can be obtained based on Figure 3-4 The relationship curve in is obtained; among them, Figure 3-3 The horizontal axis represents the wavelength, and the vertical axis represents the sensor response; Figure 3-4 The horizontal axis represents wavelength, and the vertical axis represents infrared transmittance. Similarly, the information under the D75 color temperature spectrum item in the information table (which represents the spectral response value at a given color temperature) can be obtained based on the relationship curve with wavelength on the horizontal axis and spectral response on the vertical axis.

[0061] It should be noted that the second operating condition, in which the optical path for capturing images from the camera is introduced into the glass to be evaluated, can refer to a condition in which the glass to be evaluated is present in front of the camera, and the camera needs to capture images through the glass to be evaluated. The second theoretical image color information corresponding to the second operating condition can refer to the color information of the image captured by the camera under the second operating condition, obtained through theoretical calculation based on imaging principles. The determination of the second theoretical image color information does not depend on the image actually captured by the camera.

[0062] Step 230: Determine image color change information of the second theoretical image color information relative to the first theoretical image color information.

[0063] In step 230, the second theoretical image color information can be compared with the first theoretical image color information to obtain image color change information. The image color change information can be used to characterize the image quality change of the image captured by the camera before and after the glass to be evaluated is introduced.

[0064] Step 240 : performing color adjustment on the first reference image matching the first working condition according to the image color change information to obtain a first adjusted image matching the second working condition.

[0065] It should be noted that the first reference image matching the first operating condition may be an image used to simulate the image quality of an image captured by the camera under the first operating condition. Alternatively, the first reference image may be an image of a standard color chart, such as a 24-color standard color chart. Of course, the first reference image may also be another image with color, as long as the image quality of the other image meets the requirements.

[0066] Since the image color change information can be used to characterize the change in image quality of the image captured by the camera before and after the introduction of the glass to be evaluated, the first reference image can be used to simulate the image quality of the image captured by the camera under the first working condition. By adjusting the color of the first reference image according to the image color change information, the image quality change of the first reference image can be achieved to obtain an image used to simulate the image quality of the image captured by the camera under the second working condition. This image can be used as the first adjusted image matching the second working condition.

[0067] Step 250 : Determine an introduction risk assessment result of the glass to be assessed based on the first adjusted image.

[0068] Because light sources of different color temperatures emit different spectra, the image quality changes in the images captured by the camera before and after the glass under evaluation are introduced under different color temperature light sources are also different. Therefore, multiple given color temperatures can be used. For each given color temperature, corresponding image color change information and a first adjustment image can be determined, thereby generating multiple first adjustment images. In step 250, the multiple first adjustment images can be combined to determine the introduction risk assessment results for the glass under evaluation, ensuring the accuracy and reliability of the introduction risk assessment results.

[0069] Optionally, there may be two situations for the introduction risk assessment results, namely, introduction risk and no introduction risk; wherein, introduction risk means that the introduction of the glass to be evaluated will affect the safety of autonomous driving, that is, the glass to be evaluated is not suitable for being set in front of the camera; no introduction risk means that the introduction of the glass to be evaluated will not affect the safety of autonomous driving, that is, the glass to be evaluated is suitable for being set in front of the camera.

[0070] Of course, the types of risk assessment results introduced are not limited to this. For example, the risk assessment results can be introduced in the form of a risk level, where a higher risk level indicates a greater impact on the safety of autonomous driving, and the glass is less suitable for placement in front of a camera. Another example is a risk value, where a higher risk value indicates a greater impact on the safety of autonomous driving, and the glass is less suitable for placement in front of a camera.

[0071] In embodiments of the present disclosure, theoretical calculations can be performed based on imaging principles to obtain first theoretical image color information corresponding to a first operating condition where the camera captures an image without the glass to be evaluated, and second theoretical image color information corresponding to a second operating condition where the camera captures an image with the glass to be evaluated. The image color change information of the second theoretical image color information relative to the first theoretical image color information can effectively characterize the change in image quality between the camera capture and the glass to be evaluated. Thus, based on the image color change information, a first reference image matching the first operating condition is color-adjusted to obtain a first adjusted image matching the second operating condition. The first adjusted image simulates the image quality of the camera capture after the glass to be evaluated is introduced. Therefore, the first adjusted image provides a very effective reference for determining the risk assessment results of the glass to be evaluated, thereby effectively assessing the risk of glass introduction. Furthermore, during the glass introduction risk assessment process, the images involved are the first reference image matching the first operating condition and the first adjusted image obtained by color-adjusting the first reference image. In other words, in embodiments of the present disclosure, the camera does not need to perform actual image acquisition under both the first and second operating conditions, which improves assessment efficiency and shortens the assessment cycle.

[0072] In some optional examples, such as Figure 4 As shown, step 220 includes step 2201 and step 2203.

[0073] Step 2201: For each visible light wavelength among a plurality of visible light wavelengths, based on a parameter value group corresponding to the visible light wavelength and the transmittance of the glass to be evaluated corresponding to the visible light wavelength, determine a first color impact superposition value corresponding to the R channel of the visible light wavelength, a second color impact superposition value corresponding to the G channel of the visible light wavelength, and a third color impact superposition value corresponding to the B channel of the visible light wavelength.

[0074] like Figure 5 As shown, the method provided in the embodiment of the present disclosure further includes step 510, step 520, step 530 and step 540. Optionally, the combination of step 510 to step 540 can be used as an optional implementation of step 2201 of the present disclosure.

[0075] Step 510: For each visible light wavelength among the multiple visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel.

[0076] Optionally, the multiple basic parameters affecting image color may include a sensor response corresponding to the R channel, a sensor response corresponding to the G channel, a sensor response corresponding to the B channel, sensor properties, and a spectral response at a given color temperature. Then, the sensor properties and the spectral response at a given color temperature may serve as basic parameters shared by the R, G, and B channels. The sensor response corresponding to the R channel may serve as a basic parameter exclusive to the R channel, the sensor response corresponding to the G channel may serve as a basic parameter exclusive to the G channel, and the sensor response corresponding to the B channel may serve as a basic parameter exclusive to the B channel. On this basis, the parameter value groups corresponding to each visible light wavelength may be divided into a first parameter value, a second parameter value, a third parameter value, and a fourth parameter value.

[0077] Assumptions Figure 3-2 Assuming that 400nm is one of the multiple visible light wavelengths, there are two first parameter values ​​corresponding to 400nm, namely 3.666 and 101.929, the second parameter value corresponding to 400nm is 0.7426, the third parameter value corresponding to 400nm is 0.791, and the fourth parameter value corresponding to 400nm is 3.81141. Figure 3-2 430nm is used as another visible light wavelength among the multiple visible light wavelengths. There are two first parameter values ​​corresponding to 430nm, which are 77.33 and 103.092 respectively. The second parameter value corresponding to 430nm is 0.24333. The third parameter value corresponding to 430nm is 0.3369. The fourth parameter value corresponding to 430nm is 5.25091.

[0078] Step 520 : Determine a first color impact superposition value corresponding to the R channel of the visible light wavelength based on the first parameter value, the second parameter value, and a first product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength.

[0079] Step 530 : Determine a second color impact superposition value corresponding to the G channel of the visible light wavelength based on the first parameter value, the third parameter value, and a second product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength.

[0080] Step 540 : Determine a third color impact superposition value corresponding to the B channel of the visible light wavelength based on the first parameter value, the fourth parameter value, and a third product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength.

[0081] Optionally, a first product of the first parameter value, the second parameter value, and the transmittance of the glass to be evaluated corresponding to the visible light wavelength can be used as a first color impact superposition value of the visible light wavelength corresponding to the R channel, a second product of the first parameter value, the third parameter value, and the transmittance of the glass to be evaluated corresponding to the visible light wavelength can be used as a second color impact superposition value of the visible light wavelength corresponding to the G channel, and a third product of the first parameter value, the fourth parameter value, and the transmittance of the glass to be evaluated corresponding to the visible light wavelength can be used as a third color impact superposition value of the visible light wavelength corresponding to the B channel.

[0082] Assumptions Figure 3-2 400 nm is taken as one of the multiple visible light wavelengths, and the transmittance of the glass to be evaluated corresponding to the visible light wavelength is expressed as K. Then, the first color impact superposition value of the visible light wavelength corresponding to the R channel can be 3.666×101.929×0.7426×K, the second color impact superposition value of the visible light wavelength corresponding to the G channel can be 3.666×101.929×0.791×K, and the third color impact superposition value of the visible light wavelength corresponding to the B channel can be 3.666×101.929×3.81141×K.

[0083] In some embodiments, in order to simplify the calculation, the first product, the second product and the third product can also be rounded (for example, rounded up), and the rounded value corresponding to the first product is used as the first color impact superposition value corresponding to the R channel of the visible light wavelength, the rounded value corresponding to the second product is used as the second color impact superposition value corresponding to the G channel of the visible light wavelength, and the rounded value corresponding to the third product is used as the third color impact superposition value corresponding to the B channel of the visible light wavelength.

[0084] In this way, using Figure 5In the illustrated embodiment, for each of multiple visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided according to certain rules. Then, combined with operation logic such as multiplication, the color impact superposition value corresponding to the R channel, G channel, and B channel of the visible light wavelength can be efficiently and reliably determined. The color impact superposition value can represent the impact of the superposition of different optical properties on the color of the corresponding channel of the image when the glass to be evaluated is introduced.

[0085] Step 2203 : Determine second theoretical image color information corresponding to the second working condition based on the first color impact superposition value, the second color impact superposition value, and the third color impact superposition value corresponding to each of the multiple visible light wavelengths.

[0086] like Figure 6 As shown, the method provided in the embodiment of the present disclosure further includes step 610, step 620, step 630, step 640, step 650 and step 660. Optionally, the combination of step 610 to step 660 can be used as an optional implementation of step 2203 of the present disclosure.

[0087] Step 610: Determine a first sum of first color impact superposition values ​​corresponding to a plurality of visible light wavelengths.

[0088] Step 620 : Determine a second sum of second color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths.

[0089] Step 630 : Determine a third sum of the third color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths.

[0090] Assume that the multiple visible light wavelengths are N visible light wavelengths, and the first color influence superposition values ​​corresponding to the N visible light wavelengths are x1, x2, x3, ..., x N－1 、x N , the second color influence superposition values ​​corresponding to the N visible light wavelengths are y1, y2, y3, ..., y N－1 、y N , the third color influence superposition values ​​corresponding to the N visible light wavelengths are z1, z2, z3, ..., z N－1 、z N , then:

[0091] First sum = x1+x2+x3+…+x N－1 +x N

[0092] Second sum = y1+y2+y3+…+y N－1 +y N

[0093] The third sum = z1+z2+z3+…+zN－1 +z N

[0094] Step 640 : Determine a theoretical ratio of the pixel value of the G channel to the pixel value of the R channel under the second working condition based on the ratio of the second sum to the first sum.

[0095] Alternatively, the ratio of the second sum to the first sum can be used as the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel under the second working condition. Assuming that under the second working condition, the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel is expressed as Rgain 后 , then:

[0096] Rgain 后 = Second sum / First sum

[0097] Step 650 : Determine a theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition based on the ratio of the second sum to the third sum.

[0098] Optionally, the ratio of the second sum to the third sum can be used as the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition. Assuming that under the second working condition, the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel is expressed as Bgain 后 , then:

[0099] Bgain 后 = Second sum / Third sum

[0100] Step 660 , based on the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel and the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition, determine the second theoretical image color information corresponding to the second working condition.

[0101] Optionally, the second theoretical image color information may include the Rgain determined in step 640. 后 and Bgain determined in step 650 后 In some embodiments, the second theoretical image color information may include the Rgain determined in step 640. 后 The reciprocal of and Bgain determined in step 650 后 The reciprocal of .

[0102] In this way, using Figure 6The illustrated embodiment can efficiently and reliably determine the second theoretical image color information based on the first color influence superposition value, the second color influence superposition value, and the third color influence superposition value corresponding to each of multiple visible light wavelengths, combined with operation logic such as addition and division operations. Since the color influence superposition value can characterize the impact of the superposition of different optical properties on the color of the corresponding channel of the image when the glass to be evaluated is introduced, the second theoretical image color information obtained by combining the various first color influence superposition values, the various second color influence superposition values, and the various third color influence superposition values ​​can comprehensively reflect the impact of the superposition of different optical properties on the overall color of the image. Therefore, the second theoretical image color information can effectively characterize the color information of the image captured by the camera under the second working condition.

[0103] In some embodiments, after obtaining the first sum, the second sum, and the third sum, the ratio of the first sum to the second sum can be used as the theoretical ratio of the pixel value of the R channel to the pixel value of the G channel, and the ratio of the first sum to the third sum can be used as the theoretical ratio of the pixel value of the R channel to the pixel value of the B channel under the second working condition. The second theoretical image color information may include the theoretical ratio of the pixel value of the R channel to the pixel value of the G channel under the second working condition, and the theoretical ratio of the pixel value of the R channel to the pixel value of the B channel under the second working condition.

[0104] In the embodiments of the present disclosure, based on the parameter value groups corresponding to multiple visible light wavelengths and the transmittance of the glass to be evaluated corresponding to multiple visible light wavelengths, the effects of different optical properties on color can be superimposed according to the imaging principle. As a result, the second theoretical image color information can be obtained efficiently and reliably through theoretical calculation without actual image acquisition.

[0105] In some optional examples, such as Figure 7 As shown, step 210 includes step 2101 and step 2103.

[0106] Step 2101: For each visible light wavelength among a plurality of visible light wavelengths, based on a parameter value group corresponding to the visible light wavelength, determine a fourth color impact superposition value corresponding to the R channel of the visible light wavelength, a fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and a sixth color impact superposition value corresponding to the B channel of the visible light wavelength.

[0107] like Figure 8 As shown, the method provided in the embodiment of the present disclosure further includes step 810, step 820, step 830 and step 840. Optionally, the combination of step 810 to step 840 can be used as an optional implementation of step 2101 of the present disclosure.

[0108] Step 810: For each visible light wavelength among the multiple visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel.

[0109] It should be noted that the specific implementation of step 801 can be referred to the above description of step 510 and will not be repeated here.

[0110] Step 820 : Determine a fourth color impact superposition value corresponding to the R channel of the visible light wavelength based on a fourth product of the first parameter value and the second parameter value.

[0111] Step 830 : Determine a fifth color impact superposition value corresponding to the G channel of the visible light wavelength based on a fifth product of the first parameter value and the third parameter value.

[0112] Step 840 : Determine a sixth color impact superposition value corresponding to the B channel of the visible light wavelength based on a sixth product of the first parameter value and the fourth parameter value.

[0113] Optionally, the fourth product of the first parameter value and the second parameter value can be used as the fourth color impact superposition value corresponding to the R channel of the visible light wavelength, the fifth product of the first parameter value and the third parameter value can be used as the fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and the sixth product of the first parameter value and the fourth parameter value can be used as the sixth color impact superposition value corresponding to the B channel of the visible light wavelength.

[0114] Assumptions Figure 3-2 400nm is taken as one of the multiple visible light wavelengths, then the fourth color impact superposition value of this visible light wavelength corresponding to the R channel is 3.666×101.929×0.7426, the fifth color impact superposition value of this visible light wavelength corresponding to the G channel is 3.666×101.929×0.791, and the sixth color impact superposition value of this visible light wavelength corresponding to the B channel is 3.666×101.929×3.81141.

[0115] In some embodiments, in order to simplify the calculation, the fourth product, the fifth product and the sixth product can also be rounded (for example, rounded up), and the rounded value corresponding to the fourth product is used as the fourth color impact superposition value corresponding to the R channel of the visible light wavelength, the rounded value corresponding to the fifth product is used as the fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and the rounded value corresponding to the sixth product is used as the sixth color impact superposition value corresponding to the B channel of the visible light wavelength.

[0116] In this way, using Figure 8 In the illustrated embodiment, for each of multiple visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided according to certain rules. Then, combined with operation logic such as multiplication, the color impact superposition value corresponding to the R channel, G channel, and B channel of the visible light wavelength can be efficiently and reliably determined. The color impact superposition value can represent the impact of the superposition of different optical properties on the color of the corresponding channel of the image when the glass to be evaluated is not introduced.

[0117] Step 2103 : Determine first theoretical image color information corresponding to the first working condition based on the fourth color impact superposition value, the fifth color impact superposition value, and the sixth color impact superposition value corresponding to each of the plurality of visible light wavelengths.

[0118] It should be noted that the specific implementation of step 2103 is similar to that of step 2203. For details, please refer to the above description of step 2203 and will not be repeated here. 前 and Bgain 前 , Rgain 前 With Rgain 后 The properties are the same, both represent the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel, the only difference is Rgain 前 Corresponding to the first working condition, Rgain 后 Corresponding to the second working condition. Similarly, Bgain 前 With Bgain 后 The properties are the same, both represent the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel. The only difference is Bgain 前 Corresponding to the first working condition, Bgain 后 Corresponding to the second working condition.

[0119] In the embodiments of the present disclosure, based on the parameter value groups corresponding to multiple visible light wavelengths, the effects of different optical properties on color can be superimposed according to the imaging principle. As a result, the first theoretical image color information can be obtained efficiently and reliably through theoretical calculation without actual image acquisition.

[0120] In some optional examples, such as Figure 9 As shown, step 250 includes step 2501, step 2503, step 2505 and step 2507.

[0121] Step 2501: Based on the first adjusted image, a first task execution result is obtained via a neural network model for performing a predetermined machine vision task.

[0122] It should be noted that the neural network model involved in the embodiments of the present disclosure may refer to a neural network model that has been pre-trained using a large amount of sample data, for example, a convolutional neural network model (CNN). The predetermined machine vision tasks may include but are not limited to semantic segmentation tasks, obstacle detection tasks, lane line detection tasks, etc.

[0123] Assuming that the predetermined machine vision task is an obstacle detection task, and the first adjusted image is an image representing the environment in which the vehicle is located, then in step 2501, the neural network model can be used to perform obstacle detection on the first adjusted image to determine the position, category and other information of the obstacle in the first adjusted image, and the position of the obstacle is marked in the first adjusted image by a 2D detection frame, and the category of the obstacle is marked at a predetermined position of the 2D detection frame (for example, above, below, left, right, etc.). Afterwards, the first adjusted image with the added 2D detection frame and the corresponding category can be used as the result of executing the first task.

[0124] Of course, in step 2501, the first adjusted image may be subjected to image enhancement, filtering and other processing first, and then obstacle detection may be performed on the processed first adjusted image using a neural network model.

[0125] Step 2503: Based on the first reference image, obtain the second task execution result through the neural network model.

[0126] It should be noted that the specific implementation of step 2503 can be referred to the relevant introduction to step 2501 above, and will not be repeated here.

[0127] Step 2505: Determine a first error between the first task execution result and the second task execution result.

[0128] Optionally, a loss function may be used to calculate the first error between the first task execution result and the second task execution result. The loss function may include, but is not limited to, a mean absolute error loss function (L1 loss function), a mean square error loss function (L2 loss function), and the like.

[0129] Step 2507: Determine the introduction risk assessment result of the glass to be assessed based on the first error.

[0130] Optionally, the first error may be in numerical form. In step 2507, the first error may be compared with a preset error also in numerical form, and the introduction risk assessment result of the glass to be assessed is determined based on the comparison result.

[0131] As described above, there can be multiple given color temperatures and multiple first adjustment images. Consequently, the multiple first adjustment images can be associated with multiple first errors, and each of the multiple first errors can be compared with a preset error to obtain multiple comparison results. If all of the multiple first errors are less than the preset error, a risk assessment result indicating that the glass under evaluation has no risk can be determined. If at least some of the multiple first errors are greater than or equal to the preset error, a risk assessment result indicating that the glass under evaluation has a risk can be determined.

[0132] In some embodiments, the ratios of multiple first errors to preset errors can also be calculated. If the ratios of multiple first errors to the preset errors are all smaller than the preset ratio (for example, 1), an introduction risk assessment result can be determined to indicate that the glass to be evaluated has no introduction risk. If the ratios of at least some of the multiple first errors to the preset errors are greater than or equal to the preset ratio, an introduction risk assessment result can be determined to indicate that the glass to be evaluated has an introduction risk.

[0133] In other embodiments, multiple error ranges may be pre-set, and the multiple error ranges may correspond to multiple introduction risk levels. The introduction risk assessment result may be determined based on the introduction risk levels corresponding to the error ranges to which the multiple first errors respectively belong.

[0134] In an embodiment of the present disclosure, both the first adjustment image and the first reference image can be used for a predetermined machine vision task. By referring to the error between the task execution results corresponding to the first adjustment image and the first reference image, the image quality of the first adjustment image can be evaluated, thereby providing a very effective reference for determining the introduction risk assessment result, so as to ensure the accuracy and reliability of the determined introduction risk assessment result.

[0135] In some optional examples, such as Figure 10 As shown, the method provided in the embodiment of the present disclosure further includes step 1010 , step 1020 , step 1030 and step 1040 .

[0136] Step 1010, based on the image color change information, determine a first theoretical change rate of the first category ratio value and a second theoretical change rate of the second category ratio value; wherein the first category ratio value is the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel, and the second category ratio value is the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel.

[0137] As described above, the first theoretical image color information may include Rgain 前 and Bgain 前 , the second theoretical image color information may include Rgain 后 and Bgain 后The image color change information may include a first theoretical change rate and a second theoretical change rate. Thus, in step 1010, the first theoretical change rate and the second theoretical change rate may be directly extracted from the image color change information.

[0138] Assuming that the first theoretical rate of change is expressed as ΔRgain and the second theoretical rate of change is expressed as ΔBgain, we have:

[0139] ΔRgain=(Rgain 后 -Rgain 前 ) / Rgain 前

[0140] ΔBgain=(Bgain 后 -Bgain 前 ) / Bgain 前

[0141] Step 1020 : Compare the first theoretical change rate with the first preset change rate corresponding to the first analog ratio value to obtain a first comparison result.

[0142] Optionally, the first preset change rate corresponding to the first analog value can be set to 3%, 5%, 7%, etc., which are not listed here one by one. The first comparison result can be used to represent the magnitude relationship between the first theoretical change rate and the first preset change rate.

[0143] Step 1030 : Compare the second theoretical change rate with the second preset change rate corresponding to the second analog ratio to obtain a second comparison result.

[0144] Optionally, the second preset change rate corresponding to the second analog value can be set to 3%, 5%, 7%, etc., which are not listed here. The second preset change rate can be the same as or different from the first preset change rate. The second comparison result can be used to represent the relationship between the second theoretical change rate and the second preset change rate.

[0145] Step 1040 , determining whether the first comparison result and the second comparison result satisfy a preset color adjustment condition; in response to the first comparison result and the second comparison result satisfying the preset color adjustment condition, executing step 240 .

[0146] If the first comparison result indicates that the first theoretical change rate is less than the first preset change rate, and the second comparison result indicates that the second theoretical change rate is less than the second preset change rate, it can be considered that the change in image quality of the image captured by the camera caused by the introduction of the glass to be evaluated is within an allowable range. Then, it can be determined that the first comparison result and the second comparison result meet the preset color adjustment conditions. Next, the first reference image can be color adjusted to obtain a first adjusted image, and based on the first adjusted image, the introduction risk assessment result of the glass to be evaluated can be determined.

[0147] If the first comparison result indicates that the first theoretical change rate is greater than or equal to the first preset change rate, and / or the second comparison result indicates that the second theoretical change rate is greater than or equal to the second preset change rate, it can be considered that the introduction of the glass to be evaluated causes the change in image quality of the image captured by the camera to exceed the allowable range. In this case, it can be determined that the first comparison result and the second comparison result do not meet the preset color adjustment conditions. In this case, there is no need to determine the first adjusted image, and it can be directly determined that the glass to be evaluated has an introduction risk.

[0148] In an embodiment of the present disclosure, by determining a first theoretical change rate and a second theoretical change rate, and referring to the size relationship between the first theoretical change rate and the first preset change rate, as well as the size relationship between the second theoretical change rate and the second preset change rate, it is possible to efficiently and reliably judge whether the change in image quality of the image captured by the camera caused by the introduction of the glass to be evaluated is within an allowable range. Based on the judgment result, it can be decided whether to execute subsequent steps 240 and 250, without having to execute subsequent steps 240 and 250 in all cases. This is conducive to saving power consumption and avoiding unnecessary waste of resources.

[0149] In some optional examples, such as Figure 11-1 As shown, the method provided by the embodiment of the present disclosure also includes step 1110, step 1120, step 1130, step 1140, step 1150, step 1160 and step 1170.

[0150] Step 1110 , determining a first reference change rate corresponding to the first type of ratio value and a second reference change rate corresponding to the second type of ratio value.

[0151] Optionally, corresponding reference change rates may be pre-set for the first and second types of ratios. The reference change rate set for the first type of ratio may serve as the initial first reference change rate, and the reference change rate set for the second type of ratio may serve as the initial second reference change rate. The first and second reference change rates may be the same or different. For example, both the first and second reference change rates may be 3%.

[0152] Step 1120 : Perform color adjustment on the second reference image matching the first working condition according to the current first reference change rate and the current second reference change rate to obtain a second adjusted image matching the second working condition.

[0153] It should be noted that the second reference image matching the first operating condition may be an image used to simulate the image quality of the image captured by the camera under the first operating condition. Alternatively, the second reference image may be an image of a standard color chart or another image with color. The second reference image may be the same image as the first reference image or a different image.

[0154] The following formula is introduced above:

[0155] ΔRgain=(Rgain 后 -Rgain 前 ) / Rgain 前

[0156] ΔBgain=(Bgain 后 -Bgain 前 ) / Bgain 前

[0157] In step 1120, the current first reference change rate can be substituted into the above formula as ΔRgain, and the current second reference change rate can be substituted into the above formula as ΔBgain. The pixel value of each pixel in the second reference image matching the first working condition can be considered to be known, that is, for each pixel in the second reference image, Rgain in the above formula 前 and Bgain 前 All of them can be considered known. Thus, according to the above formula, the Rgain corresponding to each pixel in the second reference image can be obtained: 后 and Bgain 后 , and for each pixel in the second reference image, the color adjustment may not change the pixel value corresponding to the G channel of the pixel, combined with the Rgain corresponding to the pixel 后 and Bgain 后 , the pixel value corresponding to the R channel and the pixel value corresponding to the B channel of the pixel point after color adjustment can be calculated, thereby obtaining the second adjusted image.

[0158] In an optional example, the pixel point with pixel coordinates (U, V) in the second reference image corresponds to a pixel value of 100 for the R, G, and B channels, ΔRgain is 1%, and ΔBgain is 2%. Assuming that the pixel values ​​corresponding to the R, G, and B channels after color adjustment are a1, 100, and a2, respectively, then according to the above formula:

[0159] 1% = (100 / a1 - 100) / 100

[0160] 2% = (100 / a2 - 100) / 100

[0161] Through calculation, we know that a1=99, a2=98. Therefore, the pixel point with pixel coordinates (U, V) in the second adjusted image corresponds to the pixel values ​​of the R channel, G channel, and B channel of 99, 100, and 98 respectively.

[0162] According to the above-described method, the pixel values ​​of each pixel point in the second adjusted image corresponding to the R channel, the G channel, and the B channel can be obtained, and the second adjusted image can be generated based on these pixel values.

[0163] Step 1130: Based on the second adjusted image, obtain a third task execution result through a neural network model.

[0164] Step 1140: Based on the second reference image, obtain the fourth task execution result through the neural network model.

[0165] Step 1150 : Determine a second error between the execution result of the third task and the execution result of the fourth task.

[0166] It should be noted that the specific implementation of step 1130 and step 1140 can refer to the relevant introduction to step 2501 above, and the specific implementation of step 1150 can refer to the relevant introduction to step 2505 above, which will not be repeated here.

[0167] Step 1160 : In response to the second error being greater than or equal to the preset error, determining a first preset change rate based on the current first reference change rate, and determining a second preset change rate based on the current second reference change rate.

[0168] Step 1170 : In response to the second error being smaller than the preset error, amplify the current first reference change rate and / or the current second reference change rate, and return to step 240 .

[0169] If the second error is greater than or equal to the preset error, it can be considered that the current first reference change rate and the current second reference change rate exceed the change rate range acceptable to the neural network model. In this case, the current first reference change rate can be determined as the first preset change rate, and the current second reference change rate can be determined as the second preset change rate. Alternatively, the first preset change rate can be smaller than the current first reference change rate, and the second preset change amount can be smaller than the current second reference change rate.

[0170] If the second error is less than the preset error, it can be considered that the current first reference change rate and the current second reference change rate are within the change rate range acceptable to the neural network model. Then, the current first reference change rate and / or the current second reference change rate can be amplified. For example, a change rate that is 1% or 2% greater than the current first reference change rate can be used as the new first reference change rate, and a change rate that is 1% or 2% greater than the current second reference change rate can be used as the new second reference change rate. The process then returns to step 240 to subsequently verify whether the new first reference change rate and the new second reference change rate are within the change rate range acceptable to the neural network model.

[0171] In an embodiment of the present disclosure, the second reference image matching the first working condition can be color-adjusted according to the current first reference change rate and the current second reference change rate to obtain a second adjusted image, and the second adjusted image and the second reference image can both be used for a predetermined machine vision task. With reference to the error between the task execution results corresponding to the second adjustment image and the second reference image, it can be evaluated whether the current first reference change rate and the current second reference change rate are within the change rate range acceptable to the neural network model. Based on the evaluation result, it can be determined whether to use the current reference change rate as the preset change rate or to amplify the current reference change rate. This is conducive to reversely searching for a suitable change rate as the preset change rate within the change rate range acceptable to the neural network model.

[0172] In an alternative example, Figure 11-2 As shown, the initial ΔRgain and ΔBgain (equivalent to the initial first reference change rate and the initial second reference change rate mentioned above) can be determined for the introduction of the glass to be evaluated. The initial ΔRgain and ΔBgain are then substituted into the image debugging system. The image debugging system can perform color adjustment on the second reference image through image simulation to obtain a second adjusted image, after which the image quality of the second adjusted image can be confirmed. If the image quality of the second adjusted image meets the requirements (corresponding to the case where the second error is less than the preset error mentioned above), the search for acceptable ΔRgain and ΔBgain can be continued based on the initial ΔRgain and ΔBgain until the maximum ΔRgain and ΔBgain within the change rate range allowed by the neural network model are obtained. In this way, the appropriate first preset change rate and second preset change rate can be reversely formulated.

[0173] In some optional examples, such as Figure 12 As shown, step 250 includes step 2509 , step 2511 and step 2513 .

[0174] Step 2509: Determine a fifth parameter value of the image quality assessment parameter of the first adjusted image.

[0175] Optionally, the image quality assessment parameters of the first adjusted image may include but are not limited to image saturation, image color deviation, etc.

[0176] Step 2511: Determine distribution information of the fifth parameter value relative to a preset quality assessment parameter value range.

[0177] Optionally, the distribution information of the fifth parameter value relative to the preset quality assessment parameter value range may be used to indicate whether the fifth parameter value is within the preset quality assessment parameter value range or exceeds the preset quality assessment parameter value range.

[0178] Step 2513: Determine the introduction risk assessment result of the glass to be assessed based on the distribution information.

[0179] Optionally, the image quality assessment parameter may only include image color deviation.

[0180] As described above, there can be multiple given color temperatures and multiple first adjustment images. A fifth parameter value indicating image color deviation can be determined for each of the multiple first adjustment images, thereby obtaining multiple fifth parameter values. If all of the multiple fifth parameter values ​​are within a preset quality assessment parameter value range, an introduction risk assessment result indicating that the glass under evaluation has no introduction risk can be determined. If at least some of the multiple fifth parameter values ​​are outside the preset quality assessment parameter value range, an introduction risk assessment result indicating that the glass under evaluation has an introduction risk can be determined.

[0181] In some embodiments, the distribution information and the first error described above can also be combined to determine an introduction risk assessment result for the glass to be evaluated. For example, if multiple fifth parameter values ​​are all within a preset quality assessment parameter value range, and multiple first errors are all less than the preset error, an introduction risk assessment result can be determined indicating that the glass to be evaluated has no introduction risk.

[0182] In an embodiment of the present disclosure, the distribution information of the fifth parameter value of the image quality assessment parameter of the first adjustment image relative to the preset quality assessment parameter value range can be used to characterize the quality of the first adjustment image, thereby providing an effective reference for determining the introduction risk assessment result of the glass to be evaluated, thereby ensuring the accuracy and reliability of the determined introduction risk assessment result.

[0183] In some optional examples, experiments can be conducted to verify the rationality of using the first theoretical image color and the second theoretical image color obtained by theoretical calculation according to the imaging principle and used for the risk assessment of glass introduction. The experimental results can be referred to Figure 13 .

[0184] Figure 13 Device 1, device 2, and device 3 in the figure can be respectively regarded as a glass to be evaluated. Figure 13 The D75, D65, TL85, and A can be used as a light source respectively, and each light source corresponds to a different color temperature.

[0185] Taking device 1 as an example of the case where the glass to be evaluated is used, for each color temperature among multiple color temperatures, actual image acquisition can be performed through the camera with and without the glass to be evaluated to obtain two images corresponding to the color temperature. Based on the two images obtained, ΔRgain and ΔBgain can be calculated. The calculated ΔRgain and ΔBgain are the ΔRgain and ΔBgain obtained through practice. In addition, for this color temperature, ΔRgain and ΔBgain can also be determined according to the theoretical calculation method introduced above. Afterwards, the ΔRgain and ΔBgain determined by theoretical calculation can be subtracted from the ΔRgain and ΔBgain obtained through practice to obtain the difference Δ. By Figure 13 As can be seen, at the color temperature corresponding to D75, the difference Δ corresponding to device 1 ranges from 0.33% to -0.81%; at the color temperature corresponding to D65, the difference Δ corresponding to device 1 ranges from 0.58% to -0.09%; at the color temperature corresponding to TL85, the difference Δ corresponding to device 1 ranges from -0.65% to -0.30%; and at the color temperature corresponding to A, the difference Δ corresponding to device 1 ranges from -0.17% to -0.9%. It is easy to see that the difference Δ corresponding to device 1 is very small at various color temperatures, demonstrating that the data obtained through practical application is essentially consistent with the data obtained through theoretical calculations. Therefore, using the embodiments of the present disclosure, it is possible to effectively assess the risk of glass introduction without actual image acquisition, thereby improving assessment efficiency and shortening the assessment cycle.

[0186] Exemplary devices

[0187] Figure 14 It is a schematic structural diagram of a glass introduction risk assessment method and apparatus provided by some exemplary embodiments of the present disclosure. Figure 14 The apparatus shown comprises:

[0188] The first acquisition module 1410 is configured to acquire first theoretical image color information corresponding to a first working condition in which the optical path of the camera capturing the image does not introduce the glass to be evaluated;

[0189] The first determination module 1420 is configured to determine, based on parameter value groups corresponding to multiple visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the multiple visible light wavelengths, second theoretical image color information corresponding to a second operating condition of the glass to be evaluated introduced by the optical path of the camera to capture the image. The parameter value group corresponding to any visible light wavelength includes parameter values ​​of multiple basic parameters that affect image color at that visible light wavelength.

[0190] The second determining module 1430 is configured to determine image color change information of the second theoretical image color information determined by the first determining module 1420 relative to the first theoretical image color information acquired by the first acquiring module 1410;

[0191] A first adjustment module 1440 is configured to perform color adjustment on the first reference image matching the first working condition according to the image color change information determined by the second determination module 1430 to obtain a first adjusted image matching the second working condition;

[0192] The third determining module 1450 is configured to determine an introduction risk assessment result of the glass to be assessed based on the first adjusted image obtained by the first adjusting module 1440 .

[0193] In some optional examples, the first determining module 1420 includes:

[0194] a first determining submodule configured to determine, for each visible light wavelength among a plurality of visible light wavelengths, a first color impact superposition value corresponding to the R channel of the visible light wavelength, a second color impact superposition value corresponding to the G channel of the visible light wavelength, and a third color impact superposition value corresponding to the B channel of the visible light wavelength based on a parameter value group corresponding to the visible light wavelength and a transmittance of the glass to be evaluated corresponding to the visible light wavelength;

[0195] The second determining submodule is configured to determine second theoretical image color information corresponding to a second working condition based on the first color impact superposition value, the second color impact superposition value, and the third color impact superposition value corresponding to each of the multiple visible light wavelengths determined by the first determining submodule.

[0196] In some optional examples, the first determining submodule includes:

[0197] a first determining unit, configured to divide, for each visible light wavelength among the plurality of visible light wavelengths, a parameter value group corresponding to the visible light wavelength into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel;

[0198] a second determining unit, configured to determine a first color impact superposition value corresponding to the R channel of the visible light wavelength based on the first parameter value determined by the first determining unit, the second parameter value determined by the first determining unit, and a first product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength;

[0199] a third determining unit, configured to determine a second color impact superposition value corresponding to the G channel of the visible light wavelength based on the first parameter value determined by the first determining unit, the third parameter value determined by the first determining unit, and a second product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength;

[0200] and a fourth determining unit, configured to determine a third color impact superposition value corresponding to the B channel for the visible light wavelength based on the first parameter value determined by the first determining unit, the fourth parameter value determined by the first determining unit, and a third product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength.

[0201] In some optional examples, the second determining submodule includes:

[0202] a fifth determining unit, configured to determine a first sum of first color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths determined by the first determining submodule;

[0203] a sixth determining unit, configured to determine a second sum of the second color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths determined by the first determining submodule;

[0204] a seventh determining unit, configured to determine a third sum of the third color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths determined by the first determining submodule;

[0205] an eighth determining unit, configured to determine a theoretical ratio of the pixel value of the G channel to the pixel value of the R channel under the second working condition based on the ratio of the second sum determined by the sixth determining unit to the first sum determined by the fifth determining unit;

[0206] a ninth determining unit, configured to determine a theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition based on the ratio of the second sum determined by the sixth determining unit to the third sum determined by the seventh determining unit;

[0207] The tenth determination unit is used to determine the second theoretical image color information corresponding to the second working condition based on the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel under the second working condition determined by the eighth determination unit, and the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel determined by the ninth determination unit.

[0208] In some optional examples, obtain the module, including:

[0209] a third determining submodule, configured to determine, for each visible light wavelength among the plurality of visible light wavelengths, based on the parameter value group corresponding to the visible light wavelength, a fourth color impact superposition value corresponding to the R channel of the visible light wavelength, a fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and a sixth color impact superposition value corresponding to the B channel of the visible light wavelength;

[0210] The fourth determining submodule is configured to determine first theoretical image color information corresponding to the first working condition based on the fourth color impact superposition value, the fifth color impact superposition value, and the sixth color impact superposition value corresponding to each of the multiple visible light wavelengths determined by the third determining submodule.

[0211] In some optional examples, the third determining submodule includes:

[0212] an eleventh determining unit, configured to divide, for each visible light wavelength among the plurality of visible light wavelengths, a parameter value group corresponding to the visible light wavelength into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel;

[0213] a twelfth determining unit, configured to determine a fourth color impact superposition value corresponding to the R channel of the visible light wavelength based on a fourth product of the first parameter value determined by the eleventh determining unit and the second parameter value determined by the eleventh determining unit;

[0214] a thirteenth determining unit, configured to determine a fifth color impact superposition value corresponding to the G channel of the visible light wavelength based on a fifth product of the first parameter value determined by the eleventh determining unit and the third parameter value determined by the eleventh determining unit;

[0215] The fourteenth determining unit is configured to determine a sixth color impact superposition value corresponding to the B channel of the visible light wavelength based on a sixth product of the first parameter value determined by the eleventh determining unit and the fourth parameter value determined by the eleventh determining unit.

[0216] In some optional examples, the third determining module 1450 includes:

[0217] A first acquisition submodule, configured to obtain a first task execution result based on the first adjusted image obtained by the first adjustment module 1440 via a neural network model for performing a predetermined machine vision task;

[0218] A second acquisition submodule is configured to obtain a second task execution result based on the first reference image via a neural network model;

[0219] a fifth determining submodule, configured to determine a first error between the first task execution result obtained by the first acquiring submodule and the second task execution result obtained by the second acquiring submodule;

[0220] The sixth determining submodule is configured to determine an introduction risk assessment result of the glass to be assessed based on the first error determined by the fifth determining submodule.

[0221] In some optional examples, such as Figure 15As shown, the apparatus provided by the embodiment of the present disclosure further includes:

[0222] a fourth determining module 1510 configured to determine, based on the image color change information determined by the second determining module 1430, a first theoretical change rate of a first-category ratio value and a second theoretical change rate of a second-category ratio value; wherein the first-category ratio value is a theoretical ratio of a pixel value of the G channel to a pixel value of the R channel, and the second-category ratio value is a theoretical ratio of a pixel value of the G channel to a pixel value of the B channel;

[0223] a first comparison module 1520 for comparing the first theoretical change rate determined by the fourth determination module 1510 with a first preset change rate corresponding to the first analog ratio value to obtain a first comparison result;

[0224] a second comparing module 1530 for comparing the second theoretical change rate determined by the fourth determining module 1510 with a second preset change rate corresponding to the second analog ratio to obtain a second comparison result;

[0225] The triggering module 1540 is configured to trigger the first adjustment module 1440 in response to the first comparison result obtained by the first comparison module 1520 and the second comparison result obtained by the second comparison module 1530 satisfying a preset color adjustment condition.

[0226] In some optional examples, such as Figure 16 As shown, the apparatus provided by the embodiment of the present disclosure further includes:

[0227] A fifth determining module 1610 is configured to determine a first reference change rate corresponding to the first type of ratio and a second reference change rate corresponding to the second type of ratio;

[0228] A second adjustment module 1620 is configured to perform color adjustment on a second reference image matching the first operating condition according to a current first reference change rate and a current second reference change rate to obtain a second adjusted image matching the second operating condition;

[0229] A second acquisition module 1630 is configured to obtain a third task execution result through a neural network model based on the second adjusted image obtained by the second adjustment module 1620;

[0230] A third acquisition module 1640 is configured to obtain a fourth task execution result based on the second reference image via a neural network model;

[0231] A sixth determining module 1650 is configured to determine a second error between the third task execution result obtained by the second obtaining module 1630 and the fourth task execution result obtained by the third obtaining module 1640;

[0232] a seventh determining module 1660 for determining, in response to the sixth determining module 1650 determining that the second error is greater than or equal to the preset error, a first preset change rate based on the current first reference change rate, and determining a second preset change rate based on the current second reference change rate;

[0233] The processing module 1670 is configured to amplify the current first reference change rate and / or the current second reference change rate in response to the second error determined by the sixth determining module 1650 being less than the preset error, and trigger the second adjusting module 1620 .

[0234] In some optional examples, the third determining module 1450 includes:

[0235] a seventh determining submodule, configured to determine a fifth parameter value of the image quality assessment parameter of the first adjusted image obtained by the first adjusting module 1440;

[0236] an eighth determining submodule, configured to determine distribution information of a fifth parameter value of the image quality assessment parameter determined by the seventh determining submodule relative to a preset quality assessment parameter value range;

[0237] The ninth determining submodule is configured to determine an introduction risk assessment result of the glass to be assessed based on the distribution information determined by the eighth determining submodule.

[0238] In the device of the present disclosure, the various optional embodiments, optional implementation methods and optional examples disclosed above can be flexibly selected and combined as needed to achieve corresponding functions and effects, and the present disclosure does not list them one by one.

[0239] Exemplary electronic devices

[0240] Figure 17 17 is a block diagram of an electronic device according to an embodiment of the present disclosure. The electronic device 1700 includes one or more processors 1710 and a memory 1720 .

[0241] The processor 1710 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 1700 to perform desired functions.

[0242] The memory 1720 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 1710 may execute the one or more computer program instructions to implement the methods of the various embodiments of the present disclosure described above and / or other desired functions.

[0243] In one example, the electronic device 1700 may further include an input device 1730 and an output device 1740 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0244] The input device 1730 may also include, for example, a keyboard, a mouse, and the like.

[0245] The output device 1740 can output various information to the outside, and may include, for example, a display, a speaker, a printer, a communication network and its connected remote output device, etc.

[0246] Of course, to simplify, Figure 17 Only some of the components related to the present disclosure in the electronic device 1700 are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, according to specific application scenarios, the electronic device 1700 may further include any other appropriate components.

[0247] Exemplary computer program products and computer-readable storage media

[0248] In addition to the above-mentioned methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions that, when executed by a processor, enable the processor to perform the steps of the method according to various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section of this specification.

[0249] The computer program product may be written in any combination of one or more programming languages ​​to implement the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0250] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, causes the processor to execute the steps of the method according to various embodiments of the present disclosure described in the above “Exemplary Method” section of this specification.

[0251] The computer-readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0252] The basic principles of the present disclosure have been described above in conjunction with specific embodiments. However, the advantages, strengths, and effects mentioned in this disclosure are merely illustrative and non-limiting, and should not be construed as necessarily possessed by each embodiment of the present disclosure. The specific details disclosed above are provided for illustrative purposes and to facilitate understanding, and are not intended to be limiting. The above details do not limit the present disclosure to necessarily being implemented using the above specific details.

[0253] Those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include these modifications and variations.

Claims

1. A glass introduction risk assessment method comprising: Acquire first theoretical image color information corresponding to a first working condition in which the glass to be evaluated is not introduced into the optical path of the camera capturing the image; Determining, based on parameter value groups corresponding to multiple visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the multiple visible light wavelengths, second theoretical image color information corresponding to a second operating condition of the glass to be evaluated introduced by the optical path of the camera to capture the image; wherein the parameter value group corresponding to any visible light wavelength includes parameter values ​​of multiple basic parameters affecting image color at that visible light wavelength; determining image color change information of the second theoretical image color information relative to the first theoretical image color information; performing color adjustment on a first reference image matching the first operating condition according to the image color change information to obtain a first adjusted image matching the second operating condition; An introduction risk assessment result of the glass to be assessed is determined based on the first adjusted image.

2. The method according to claim 1, wherein The determining, based on the parameter value groups corresponding to the plurality of visible light wavelengths and the transmittances of the glass to be evaluated corresponding to the plurality of visible light wavelengths, second theoretical image color information corresponding to the second working condition of the glass to be evaluated introduced by the optical path for capturing the image by the camera includes: For each visible light wavelength in the plurality of visible light wavelengths, based on the parameter value group corresponding to the visible light wavelength and the transmittance of the glass to be evaluated corresponding to the visible light wavelength, determine a first color impact superposition value corresponding to the R channel of the visible light wavelength, a second color impact superposition value corresponding to the G channel of the visible light wavelength, and a third color impact superposition value corresponding to the B channel of the visible light wavelength; Second theoretical image color information corresponding to the second working condition is determined based on the first color impact superposition value, the second color impact superposition value, and the third color impact superposition value corresponding to each of the multiple visible light wavelengths.

3. The method according to claim 2, wherein: The determining, for each visible light wavelength among the plurality of visible light wavelengths, a first color impact superposition value corresponding to the R channel, a second color impact superposition value corresponding to the G channel, and a third color impact superposition value corresponding to the B channel of the visible light wavelength based on the parameter value group corresponding to the visible light wavelength and the transmittance of the glass to be evaluated corresponding to the visible light wavelength, comprises: For each visible light wavelength in the plurality of visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel; Determine a first color impact superposition value corresponding to the R channel of the visible light wavelength based on a first product of the first parameter value, the second parameter value, and the transmittance of the glass to be evaluated corresponding to the visible light wavelength; Determine a second color impact superposition value corresponding to the G channel of the visible light wavelength based on the first parameter value, the third parameter value, and a second product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength; A third color impact superposition value corresponding to the B channel of the visible light wavelength is determined based on the first parameter value, the fourth parameter value, and a third product of the transmittance of the glass to be evaluated corresponding to the visible light wavelength.

4. The method according to claim 2, wherein: The determining, based on the first color impact superposition value, the second color impact superposition value, and the third color impact superposition value corresponding to each of the plurality of visible light wavelengths, second theoretical image color information corresponding to the second working condition includes: determining a first sum of the first color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths; determining a second sum of the second color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths; determining a third sum of the third color impact superposition values ​​corresponding to each of the plurality of visible light wavelengths; Determining a theoretical ratio of the pixel value of the G channel to the pixel value of the R channel under the second working condition based on the ratio of the second sum to the first sum; Determining a theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition based on the ratio of the second sum to the third sum; Based on the theoretical ratio of the pixel value of the G channel to the pixel value of the R channel and the theoretical ratio of the pixel value of the G channel to the pixel value of the B channel under the second working condition, the second theoretical image color information corresponding to the second working condition is determined.

5. The method according to claim 1, wherein The method of obtaining the first theoretical image color information corresponding to the first working condition of the glass to be evaluated without introducing the optical path of the camera to capture the image includes: determining, for each visible light wavelength in the plurality of visible light wavelengths, a fourth color impact superposition value corresponding to the R channel of the visible light wavelength, a fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and a sixth color impact superposition value corresponding to the B channel of the visible light wavelength based on the parameter value group corresponding to the visible light wavelength; First theoretical image color information corresponding to the first working condition is determined based on the fourth color impact superposition value, the fifth color impact superposition value, and the sixth color impact superposition value corresponding to each of the multiple visible light wavelengths.

6. The method according to claim 5, wherein: The determining, for each visible light wavelength among the plurality of visible light wavelengths, based on the parameter value group corresponding to the visible light wavelength, a fourth color impact superposition value corresponding to the R channel of the visible light wavelength, a fifth color impact superposition value corresponding to the G channel of the visible light wavelength, and a sixth color impact superposition value corresponding to the B channel of the visible light wavelength includes: For each visible light wavelength in the plurality of visible light wavelengths, the parameter value group corresponding to the visible light wavelength is divided into a first parameter value of a basic parameter shared by the R channel, the G channel, and the B channel, a second parameter value of the basic parameter exclusively for the R channel, a third parameter value of the basic parameter exclusively for the G channel, and a fourth parameter value of the basic parameter exclusively for the B channel; determining a fourth color impact superposition value corresponding to the R channel of the visible light wavelength based on a fourth product of the first parameter value and the second parameter value; determining a fifth color impact superposition value corresponding to the G channel of the visible light wavelength based on a fifth product of the first parameter value and the third parameter value; A sixth color impact superposition value corresponding to the B channel of the visible light wavelength is determined based on a sixth product of the first parameter value and the fourth parameter value.

7. The method according to claim 1, wherein The step of determining the introduction risk assessment result of the glass to be assessed based on the first adjusted image includes: Obtaining a first task execution result based on the first adjusted image via a neural network model for performing a predetermined machine vision task; Based on the first reference image, obtaining a second task execution result through the neural network model; determining a first error between the first task execution result and the second task execution result; Based on the first error, an introduction risk assessment result of the glass to be assessed is determined.

8. The method according to claim 7, further comprising: Determining, based on the image color change information, a first theoretical change rate of a first-category ratio value and a second theoretical change rate of a second-category ratio value; wherein the first-category ratio value is a theoretical ratio of a pixel value of a G channel to a pixel value of an R channel, and the second-category ratio value is a theoretical ratio of a pixel value of a G channel to a pixel value of a B channel; comparing the first theoretical change rate with a first preset change rate corresponding to the first analog ratio to obtain a first comparison result; comparing the second theoretical change rate with a second preset change rate corresponding to the second analog ratio to obtain a second comparison result; In response to the first comparison result and the second comparison result satisfying the preset color adjustment condition, the step of performing color adjustment on the first reference image matching the first working condition according to the image color change information to obtain a first adjusted image matching the second working condition is executed.

9. The method according to claim 8, further comprising: determining a first reference change rate corresponding to the first-class ratio and a second reference change rate corresponding to the second-class ratio; performing color adjustment on a second reference image matching the first operating condition according to the current first reference change rate and the current second reference change rate to obtain a second adjusted image matching the second operating condition; Based on the second adjusted image, obtaining a third task execution result through the neural network model; Based on the second reference image, obtaining a fourth task execution result through the neural network model; determining a second error between the third task execution result and the fourth task execution result; In response to the second error being greater than or equal to a preset error, determining the first preset change rate based on the current first reference change rate, and determining the second preset change rate based on the current second reference change rate; In response to the second error being less than the preset error, the current first reference change rate and / or the current second reference change rate are amplified, and the step of performing color adjustment on the second reference image matching the first operating condition according to the current first reference change rate and the current second reference change rate is returned to obtain a second adjusted image matching the second operating condition.

10. The method according to claim 1, wherein The step of determining the introduction risk assessment result of the glass to be assessed based on the first adjusted image includes: determining a fifth parameter value of an image quality assessment parameter of the first adjusted image; Determining distribution information of the parameter value relative to a preset quality assessment parameter value range; Based on the distribution information, an introduction risk assessment result of the glass to be assessed is determined.

11. A glass introduction risk assessment device comprising: A first acquisition module is used to acquire first theoretical image color information corresponding to a first working condition in which the glass to be evaluated is not introduced into the optical path of the camera capturing the image; A first determination module is configured to determine, based on parameter value groups corresponding to a plurality of visible light wavelengths and the transmittance of the glass to be evaluated corresponding to the plurality of visible light wavelengths, second theoretical image color information corresponding to a second operating condition of the glass to be evaluated when the optical path through which the camera captures the image introduces the image; wherein the parameter value group corresponding to any visible light wavelength includes parameter values ​​of a plurality of basic parameters affecting image color at that visible light wavelength; a second determining module, configured to determine image color change information of the second theoretical image color information determined by the first determining module relative to the first theoretical image color information acquired by the first acquiring module; a first adjustment module, configured to perform color adjustment on a first reference image matching the first operating condition according to the image color change information determined by the second determination module, to obtain a first adjusted image matching the second operating condition; The third determining module is configured to determine an introduction risk assessment result of the glass to be assessed based on the first adjustment image obtained by the first adjustment module. 12 . A computer-readable storage medium storing a computer program, wherein the computer program is used to execute the glass introduction risk assessment method according to claim 1 .

13. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the glass introduction risk assessment method according to any one of claims 1 to 10.

Images