Image processing model generation method, device, equipment and storage medium

Abstract

The present application provides a method, device, equipment and storage medium for generating an image processing model, the method comprising: obtaining an initial ISP parameter set, the initial ISP parameter set including at least one initial ISP parameter; the initial ISP parameter set indicates an initial scene corresponding to an initial image; screening at least one initial ISP parameter to obtain at least one target ISP parameter; based on a sample scene, adjusting the initial value of each target ISP parameter in at least one target ISP parameter in the initial ISP parameter set to a corresponding target value to obtain a first sample ISP parameter set, the first sample ISP parameter set indicating a sample scene corresponding to a sample image; inputting the first sample ISP parameter set into an image processing model to be trained to obtain an output image; calculating the output image to obtain a first loss function value. The technical solution of the present application can save NPU computing power.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to a method, device, equipment and storage medium for generating an image processing model. Background Art

[0002] In a chip system, especially a high-resolution ISP (Image Signal Processor) system with an NPU (Neural Processing Unit), NPU computing power is a very important resource. Therefore, it is necessary to improve the efficiency of using NPU computing power.

[0003] In related technologies, the AI ​​HDR (Artificial Intelligence High-Dynamic Range) model uses the limited NPU computing power to enhance the sRGB (standard Red Green Blue) image obtained after the chip ISP processing to a high dynamic range image. The AI ​​HDR model is trained based on randomly simulated paired training data.

[0004] However, in related technologies, the training data used in training the AI ​​HDR model has a high degree of randomness, which results in the AI ​​HDR model being unable to solve practical problems in a targeted manner, thereby wasting NPU computing power. Summary of the invention

[0005] The embodiments of the present application provide a method, device, equipment and storage medium for generating an image processing model, which can obtain an AI HDR model that solves targeted problems, thereby avoiding the waste of NPU computing power. The technical solution is as follows:

[0006] According to a first aspect of an embodiment of the present application, a method for generating an image processing model is provided, the method comprising:

[0007] Acquire an initial ISP parameter set, the initial ISP parameter set including at least one initial ISP parameter, each of the at least one initial ISP parameter being set with an initial value; the initial ISP parameter set indicates an initial scene corresponding to an initial image, and the initial ISP parameter indicates an image effect of the initial image;

[0008] Screening the at least one initial ISP parameter to obtain at least one target ISP parameter;

[0009] Based on the sample scene, adjusting the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value to obtain a first sample ISP parameter set, where the first sample ISP parameter set indicates the sample scene corresponding to the sample image;

[0010] Inputting the first sample ISP parameter set into the image processing model to be trained to obtain an output image;

[0011] The output image is calculated to obtain a first loss function value, and based on the first loss function value, the model parameters of the image processing model to be trained are updated.

[0012] In a possible implementation, after obtaining the initial ISP parameter set, the following is further included:

[0013] Each of the at least one initial ISP parameter is adjusted at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter.

[0014] In a possible implementation, adjusting each of the at least one initial ISP parameter at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter includes:

[0015] For one adjustment, one of the at least one initial ISP parameter is used as the ISP parameter to be adjusted;

[0016] Adjusting the initial value of the ISP parameter to be adjusted to a corresponding adjustment value to obtain the adjusted ISP parameter;

[0017] Based on the adjusted ISP parameter and the remaining initial ISP parameters in the at least one initial ISP parameter, an image effect of the adjusted image is obtained.

[0018] In a possible implementation, at least one of the initial ISP parameters is screened to obtain at least one target ISP parameter, including:

[0019] The at least one initial ISP parameter is screened based on the image effect of the adjusted image to obtain the at least one target ISP parameter.

[0020] In a possible implementation, based on the sample scenario, adjusting the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value includes:

[0021] Adjusting an initial value of each target ISP parameter of the at least one target ISP parameter to a corresponding first candidate value to obtain a corresponding first candidate image;

[0022] Based on the sample scene, obtaining a comparison result between the image effect of the first candidate image and the image effect of the initial image;

[0023] If the comparison result meets the preset conditions, the first candidate value of each target ISP parameter in the at least one target ISP parameter is used as the target value of each target ISP parameter in the at least one target ISP parameter; or if the comparison result does not meet the preset conditions, the first candidate value of each target ISP parameter in the at least one target ISP parameter is continued to be adjusted until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained.

[0024] In a possible implementation, before inputting the first sample ISP parameter set into the image processing model to be trained to obtain the output image, the method further includes:

[0025] Acquire first environment data; the first environment data indicates the environment corresponding to the first sample ISP parameter set.

[0026] In a possible implementation, the first sample ISP parameter set is input into the image processing model to be trained to obtain an output image, including:

[0027] The first environmental data and the first sample ISP parameter set are input into the image processing model to be trained to obtain the output image.

[0028] According to a first aspect of an embodiment of the present application, a device for generating an image processing model is provided, the device comprising:

[0029] A first acquisition module is used to acquire an initial ISP parameter set, where the initial ISP parameter set includes at least one initial ISP parameter, and each of the at least one initial ISP parameter is set with an initial value; the initial ISP parameter set indicates an initial scene corresponding to an initial image, and the initial ISP parameter indicates an image effect of the initial image;

[0030] A screening module, used for screening the at least one initial ISP parameter to obtain at least one target ISP parameter;

[0031] A first adjustment module is used to adjust the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value based on the sample scene, so as to obtain a first sample ISP parameter set, where the first sample ISP parameter set indicates the sample scene corresponding to the sample image;

[0032] A training module, used for inputting the first sample ISP parameter set into the image processing model to be trained to obtain an output image;

[0033] The updating module is used to calculate the output image to obtain a first loss function value, and based on the first loss function value, update the model parameters of the image processing model to be trained.

[0034] In a possible implementation, the device further includes:

[0035] The second adjustment module is used to adjust each of the at least one initial ISP parameter at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter.

[0036] In a possible implementation, the second adjustment module includes:

[0037] A first determining unit is configured to use, for one adjustment, one of the at least one initial ISP parameter as the ISP parameter to be adjusted;

[0038] A first adjustment unit, used for adjusting the initial value of the ISP parameter to be adjusted to a corresponding adjustment value to obtain the adjusted ISP parameter;

[0039] The first acquisition unit is used to acquire an image effect of the adjusted image based on the adjusted ISP parameter and the remaining initial ISP parameters in the at least one initial ISP parameter.

[0040] In a possible implementation, the screening module includes:

[0041] The screening unit is used to screen the at least one initial ISP parameter based on the image effect of the adjusted image to obtain the at least one target ISP parameter.

[0042] In a possible implementation, the adjustment module includes:

[0043] A second adjustment unit, configured to adjust an initial value of each target ISP parameter of the at least one target ISP parameter to a corresponding first candidate value, so as to obtain a corresponding first candidate image;

[0044] A second acquisition unit, configured to acquire, based on the sample scene, a comparison result between an image effect of the first candidate image and an image effect of the initial image;

[0045] A determination unit is configured to, if the comparison result meets a preset condition, use the first candidate value of each target ISP parameter in the at least one target ISP parameter as the target value of each target ISP parameter in the at least one target ISP parameter; or, if the comparison result does not meet the preset condition, continue to adjust the first candidate value of each target ISP parameter in the at least one target ISP parameter until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained.

[0046] In a possible implementation, the device further includes:

[0047] The second acquisition module is used to acquire first environment data; the first environment data indicates the environment corresponding to the first sample ISP parameter set.

[0048] In a possible implementation, the training module includes:

[0049] An input unit is used to input the first environmental data and the first sample ISP parameter set into the image processing model to be trained to obtain the output image.

[0050] According to a third aspect of an embodiment of the present application, a computer device is provided, comprising a processor and a memory, wherein the memory is used to store at least one program, and the at least one program is loaded by the processor and executes the above-mentioned method for generating an image processing model.

[0051] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, in which at least one program is stored. The at least one program is loaded and executed by a processor to implement the above-mentioned method for generating an image processing model.

[0052] According to a second aspect of an embodiment of the present application, there is provided a device comprising:

[0053] The beneficial effects of the technical solution provided by the embodiment of the present application are:

[0054] In an embodiment of the present application, an embodiment of the present application provides a method for generating an image processing model, determines a first sample ISP parameter according to a required sample scene, and then trains the AI ​​HDR model based on the first sample ISP parameter, and associates the ISP with the AI ​​HDR, so that the trained AI HDR module is associated with the actual scene, which is targeted when processing actual problems, thereby saving NPU computing power. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0056] Figure 1 is a schematic diagram of an implementation environment provided according to an embodiment of the present application;

[0057] Figure 2 It is a flowchart of a method for generating an image processing model provided in an embodiment of the present application;

[0058] Figure 3 is a structural schematic diagram of a device for generating an image processing model according to an embodiment of the present application;

[0059] Figure 4 is a schematic diagram of the structure of a terminal provided according to an embodiment of the present application;

[0060] Figure 5 It is a structural diagram of a server provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0062] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0063] In the present application, the terms "first" and "second" are used to distinguish the same or similar items with substantially the same effects and functions. It should be understood that there is no logical or temporal dependency between "first", "second" and "nth", nor is there a limitation on the quantity and execution order. It should also be understood that although the following description uses the terms first, second, etc. to describe various elements, these elements should not be limited by the terms.

[0064] These terms are only used to distinguish one element from another element. For example, without departing from the scope of various examples, a first action can be referred to as a second action, and similarly, a second action can also be referred to as a first action. Both the first action and the second action can be actions, and in some cases, can be separate and different actions.

[0065] Here, at least one means one or more than one, for example, at least one action can be one action, two actions, three actions, or any other action that is an integer greater than or equal to one. And multiple means two or more than two, for example, multiple actions can be two actions, three actions, or any other action that is an integer greater than or equal to two.

[0066] Figure 1 1 is a schematic diagram of an implementation environment provided according to an embodiment of the present application, the implementation environment may include a terminal 100, and a chip 101 of the terminal 100 includes a high-resolution ISP 1011 with an NPU and an AI HDR model 1012. When an image is processed based on the chip 101, the image is processed by the ISP 1011 and the AI ​​HDR model 1012 in the chip, and the processed image is output, wherein the AI ​​HDR model 1012 can be set at a preset position in the chip 101 as needed.

[0067] In some embodiments, the terminal 100 further includes a camera component, which is used to acquire sample image data.

[0068] In some embodiments, ISP1011 includes multiple modules, and different modules correspond to different ISP parameters, for example, a white balance module, a contrast module, a sharpening module, and a brightness module. Among them, the white balance module is provided with white balance ISP parameters, the contrast module is provided with contrast ISP parameters, the sharpening module is provided with sharpening ISP parameters, and the brightness module is provided with brightness ISP parameters.

[0069] Optionally, taking the brightness ISP parameter as an example, when the brightness ISP parameter is adjusted, the brightness performance of the ISP can be changed. For example, the brightness of an image processed by the ISP after the brightness ISP parameter is changed becomes darker or brighter compared to the brightness of an image processed by the ISP without changing the brightness ISP parameter.

[0070] Optionally, when ISP1011 is closer to the input end of terminal 100 than AI HDR model 1012, the input end of the terminal obtains sample image data, sends the sample image data to ISP1011, and ISP1011 processes the sample image data based on the ISP parameters to obtain a first intermediate image. ISP1011 inputs the first intermediate image to AI HDR model 1012, and AI HDR model 1012 processes the first intermediate image to obtain a first target image.

[0071] Optionally, when ISP1011 is farther away from the input end of terminal 100 than AI HDR model 1012, the input end of terminal 100 obtains sample image data, sends the sample image data to AI HDR model, AI HDR model 1012 processes the sample image data to obtain a second intermediate image. The second intermediate image is sent to ISP1011, and ISP1011 processes the second intermediate image based on the ISP parameters to obtain a second target image.

[0072] In some embodiments, terminal 100 is connected to a wired or wireless network.

[0073] In some embodiments, the wireless network or wired network uses standard communication technology and / or protocol. The network is usually the Internet, but it can also be any network, including but not limited to any combination of local area network (LAN), metropolitan area network (MAN), wide area network (WAN), mobile, wired or wireless network, private network or virtual private network. In some embodiments, the data exchanged through the network is represented by technology and / or format including HyperText Mark-up Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec) can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technology can also be used to replace or supplement the above data communication technology.

[0074] Figure 2is a flow chart of a method for generating an image processing model according to an embodiment of the present application, such as Figure 2 As shown, in the embodiment of the present application, an example is taken as an example of being applied to a terminal with a high-resolution ISP and an AI HDR model provided with an NPU. The method comprises the following steps:

[0075] In step 201, the terminal obtains an initial ISP parameter set, which includes at least one initial ISP parameter, each of which is set with an initial value; the initial ISP parameter set indicates an initial scene corresponding to an initial image, and the initial ISP parameter indicates an image effect of the initial image.

[0076] In some embodiments, the image processing model may be an AI HDR model, which is used to enhance a low dynamic range image into a high dynamic range image.

[0077] In some embodiments, the initial ISP parameter set includes various initial ISP parameters that affect the image effect. For example, the initial ISP parameter may be any one of white balance, brightness, contrast, and sharpening. The image effect of the initial image indicates the clarity of the initial image and the degree of restoration of the corresponding initial scene. The initial scene and the sample scene may be any one of a city night scene, a community entrance, and a suburban sky. The initial value is a range, for example: 10 to 230.

[0078] In some embodiments, the initial ISP parameters may be obtained as follows:

[0079] The first method is to reuse existing training data. For example, the training data of the last training of the AI ​​HDR model is used as the initial ISP parameters to save the cost of training the AI ​​HDR model and shorten the training time of the AI ​​HDR model.

[0080] The second method: initialize the terminal (chip), input the preset image into the chip, obtain the first processed image after being processed by ISP and the AI ​​HDR model to be trained, adjust the original ISP parameter set corresponding to the preset image in a targeted manner according to the image effect of the first processed image, obtain the adjusted first original ISP parameter set, input the first processed image into the chip, obtain the second processed image after being processed by ISP based on the first original ISP parameter set and the AI ​​HDR model to be trained, if the image effect of the second processed image meets the preset image effect, then use the first original ISP parameter set as the initial ISP parameter set, or if the image effect of the second processed image does not meet the preset image effect, then continue to adjust the first original ISP parameter set to obtain the second original ISP parameter, etc., until the initial ISP parameter set is obtained. Optionally, the image effect can be determined from the following aspects of the image: color, brightness, contrast, clarity, and noise performance, etc., where color indicates whether the color of the image is bright, natural, and the color reproduction. Brightness indicates the brightness of the image. Contrast indicates the brightness and darkness contrast of the image and the detail performance of the bright and dark parts of the image. The clarity indicates the detail performance of the image, which may include edge clarity and texture clarity. The noise performance indicates the noise level in the image, which may include noise points, etc.

[0081] In some embodiments, in order to restore the corresponding scene to the greatest extent possible, different scenes have different concerns about ISP parameters, that is, the initial value of each initial ISP parameter in at least one initial ISP parameter set corresponding to different initial scenes is different.

[0082] Optionally, when the initial scene is a city night scene, the initial ISP parameter (brightness) should highlight the characteristics and details of the city night scene, such as lights, buildings, and people, and at the same time, the overall light and dark distribution and color contrast of the image should also be considered. The initial ISP parameter (contrast) needs to highlight the layering and three-dimensionality of the image.

[0083] Optionally, when the initial scene is a residential gate, the required brightness should be able to clearly display the residential gate, access control system, and security booth, etc., to highlight the characteristics and atmosphere of the residential district. That is, the initial values ​​of the initial ISP parameters (brightness) corresponding to the residential gate and the urban night scene are different.

[0084] In some embodiments, after step 201, the following step 201A is also included: the terminal adjusts each of the at least one initial ISP parameter at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter.

[0085] In some embodiments, since the number of target ISP parameters in the initial ISP parameter set cannot be determined, when adjusting the initial ISP parameter set, one initial ISP parameter in the initial ISP parameter set is used as the ISP parameter to be adjusted, or multiple initial ISP parameters in the initial ISP parameter set are used as the ISP parameters to be adjusted at the same time.

[0086] Optionally, when adjusting the ISP parameter to be adjusted, the ISP parameter to be adjusted may be adjusted at least once.

[0087] Optionally, when multiple initial ISP parameters in the initial ISP parameter set are simultaneously used as ISP parameters to be adjusted, for example, two initial ISP parameters in the initial ISP parameter set can be simultaneously used as ISP parameters to be adjusted, and when the initial ISP parameter set is adjusted, each initial ISP parameter in a group of data is adjusted each time. For example, the initial ISP parameter (brightness) and the initial ISP parameter (contrast) in the initial ISP parameter set are simultaneously used as ISP parameters to be adjusted.

[0088] In some embodiments, the above-mentioned step 201A includes, for one adjustment, using one of the at least one initial ISP parameters as the ISP parameter to be adjusted; adjusting the initial value of the ISP parameter to be adjusted to the corresponding adjustment value to obtain the adjusted ISP parameter; and obtaining the image effect of the adjusted image based on the adjusted ISP parameter and the remaining initial ISP parameters of the at least one initial ISP parameter.

[0089] In some embodiments, an example of adjusting one initial ISP parameter in the initial ISP parameter set each time is used for explanation. Optionally, when the initial value of the initial ISP parameter (brightness) is adjusted for the first time, the initial value of the initial ISP parameter (brightness) ranges from 10 to 230, and the initial value of the initial ISP parameter (brightness) and the initial image are input into the parameter debugging tool, and based on the adjustment of the initial value of the initial ISP parameter (brightness) by the parameter debugging tool, the adjustment value of the initial ISP parameter (brightness) is 50 to 200. The parameter debugging tool adjusts the initial image based on the adjustment value of the initial ISP parameter (brightness) being 50 to 200, and obtains a brightness-adjusted image and an image effect of the brightness-adjusted image.

[0090] Optionally, when the initial value of the initial ISP parameter (contrast) is adjusted for the second time, the initial value of the initial ISP parameter (contrast) and the initial image are input into the parameter debugging tool, and based on the adjustment of the initial value of the initial ISP parameter (contrast) by the parameter debugging tool, an adjusted value of the initial ISP parameter (contrast) is obtained. The parameter debugging tool adjusts the initial image based on the adjusted value of the initial ISP parameter (contrast) to obtain a contrast-adjusted image and an image effect of the contrast-adjusted image.

[0091] Similarly, the image effect of the white balance adjusted image corresponding to the initial ISP parameter (white balance) and the image effect of the sharpening adjusted image corresponding to the initial ISP parameter (sharpening) can be obtained.

[0092] Optionally, when determining the target ISP parameters, the image effects of the brightness adjusted image, the image effects of the contrast adjusted image, the image effects of the white balance adjusted image, and the image effects of the sharpening adjusted image can be compared. If the image effect of the contrast adjusted image is the best, the initial ISP parameters (contrast) are used as the target ISP parameters, thereby obtaining a target ISP initial parameter.

[0093] Similarly, two or three target ISP parameters may also be obtained, which will not be described in detail in the embodiments of the present application.

[0094] In step 202, the terminal screens the at least one initial ISP parameter to obtain at least one target ISP parameter.

[0095] In some embodiments, a sample scene is determined according to actual needs. After the sample scene is determined, at least one target ISP parameter affecting the image effect of the sample image is determined based on the sample scene.

[0096] In some embodiments, the above step 202 includes step 202A: the terminal screens the at least one initial ISP parameter based on the image effect of the adjusted image to obtain the at least one target ISP parameter.

[0097] Optionally, at least one initial ISP parameter includes M initial ISP parameters as an example for illustration, where M is an integer greater than or equal to 1. First, N target ISP parameters (that is, N valid ISP parameters) in the at least one initial ISP parameter are determined, where N is an integer greater than or less than 1 and less than or equal to M.

[0098] Optionally, the target ISP parameter can be understood as adjusting the initial image based on the adjusted target ISP parameter to obtain a first adjusted image, and the first adjusted image can better restore the sample scene than the initial image, that is, in the sample scene, the image effect of the first adjusted image is better than the image effect of the initial image. Each of the remaining (MN) initial ISP parameters in the at least one initial ISP parameter is used as an invalid ISP parameter.

[0099] Optionally, for invalid ISP parameters, it can be understood that the initial image is adjusted based on the adjusted invalid ISP parameters to obtain a second adjusted image. The second adjusted image cannot restore the sample scene better than the initial image, that is, in the sample scene, the image effect of the second adjusted image is worse than the image effect of the initial image.

[0100] Optionally, in order to select N target ISP parameters from M initial ISP parameters, the M initial ISP parameters and the initial image are input into a parameter debugging tool, the parameter debugging tool adjusts the initial value of each of the M initial ISP parameters based on the parameter debugging tool to obtain an adjustment value corresponding to each of the M initial ISP parameters, and the parameter debugging tool adjusts the initial image based on the adjustment value corresponding to each of the M initial ISP parameters to obtain a third adjusted image. The parameter debugging tool displays the adjustment value corresponding to each of the M initial ISP parameters and the corresponding image effect in the form of a document, and the document includes the initial value of each of the M initial ISP parameters and the corresponding adjustment value, the initial image, and the third adjusted image, so as to determine whether to use each of the M initial ISP parameters as a target ISP parameter by comparing the image effect of the third adjusted image with the image effect of the initial image.

[0101] It should be noted that the parameter debugging tool involved in the embodiments of the present application is a conventional tool. Therefore, the embodiments of the present application no longer make specific limitations on the parameter debugging tool.

[0102] In the embodiment of the present application, the initial ISP parameter set is screened through the above method to preliminarily determine N target ISP parameters, so that in the subsequent adjustment process, only the N target ISP parameters need to be adjusted, and the remaining (MN) initial ISP parameters in the initial ISP parameter set are no longer adjusted, thereby reducing the number of initial ISP parameters that need to be adjusted, thereby improving the training efficiency of the AI ​​HDR model and saving NPU computing power.

[0103] In step 203, based on the sample scene, the terminal adjusts the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to the corresponding target value to obtain a first sample ISP parameter set, where the first sample ISP parameter set indicates the sample scene corresponding to the sample image.

[0104] In some embodiments, it can be known from the analysis of the above steps 201 and 202 that different initial scenes have different corresponding initial ISP parameter sets, that is, the initial value of each initial ISP parameter in at least one initial ISP parameter of different initial scenes is different. Therefore, in order to obtain a sample image by adjusting the initial image so that the sample image can better restore the sample scene, it is necessary to adjust each target ISP parameter in at least one target ISP parameter, that is, adjust the initial value of each target ISP parameter in at least one target ISP parameter to the target value, so as to obtain a first sample ISP parameter set that can better restore the sample scene.

[0105] In some embodiments, the above step 203 includes: adjusting the initial value of each target ISP parameter in the at least one target ISP parameter to the corresponding first candidate value to obtain the corresponding first candidate image. Based on the sample scene, obtaining a comparison result of the image effect of the first candidate image and the image effect of the initial image. If the comparison result meets the preset conditions, the first candidate value of each target ISP parameter in the at least one target ISP parameter is used as the target value of each target ISP parameter in the at least one target ISP parameter; or if the comparison result does not meet the preset conditions, the first candidate value of each target ISP parameter in the at least one target ISP parameter continues to be adjusted until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained.

[0106] Optionally, when obtaining the comparison result of the image effect of the first candidate image and the image effect of the initial image based on the sample scene, the image effect of the first candidate image and the image effect of the initial image can be first manually determined, and then the image effect of the first candidate image and the image effect of the initial image are manually compared to obtain the comparison result. Finally, the terminal obtains the comparison result.

[0107] In some embodiments, according to the required sample scene, at least one target ISP parameter and the initial image obtained in step 202 are input into a parameter debugging tool, the parameter debugging tool adjusts the initial value of each target ISP parameter in the at least one target ISP parameter respectively to obtain a second candidate adjustment value of each target ISP parameter in the at least one target ISP parameter, the initial image is adjusted based on the adjustment value of each target ISP parameter in the at least one target ISP parameter to obtain a second candidate image, and according to the image effect of the fourth adjusted image and the image effect of the initial image, it is determined whether to determine the second candidate adjustment value of each target ISP parameter in the at least one target ISP parameter as the corresponding target value.

[0108] Optionally, taking N=1 as an example, the range of the initial value of the target ISP parameter (brightness) is 10-230, and the initial value of the target ISP parameter (brightness) is adjusted based on the parameter debugging tool, and the third candidate adjustment value of the target ISP parameter (brightness) is 50-200, and the initial image is adjusted based on the first adjustment value of the target ISP parameter (brightness) to obtain a third candidate image. If the image effect of the fifth adjusted image does not meet the preset conditions, the third candidate adjustment value of the target ISP parameter (brightness) is adjusted based on the parameter debugging tool to obtain a fourth candidate adjustment value of 60-180, and the third candidate image is adjusted based on the fourth candidate adjustment value to obtain a fourth candidate image, and if the image effect of the fourth candidate image meets the preset conditions, the fourth candidate adjustment value is determined as the target value of the target ISP parameter (brightness).

[0109] Optionally, taking the initial ISP parameter set including brightness, contrast, white balance and sharpening as an example, after obtaining the target value of the target ISP parameter (brightness), the target value of the target ISP parameter (brightness), the initial value of the contrast, the initial value of the white balance and the initial value of the sharpening are used as the first sample ISP parameter set.

[0110] In step 204, the terminal inputs the first sample ISP parameter set into the image processing model to be trained to obtain an output image.

[0111] In some embodiments, before step 204, the terminal may also acquire first environment data, where the first environment data indicates an environment corresponding to the first sample ISP parameter set, for example, an environment corresponding to any one of a sunny day, indoors, and outdoors.

[0112] In some embodiments, the terminal inputs the first sample ISP parameters and the sample image data into the ISP, the ISP processes the sample image data based on the first sample ISP parameters to obtain a third intermediate image, the ISP inputs the third intermediate image into the AI ​​HDR model to be trained, the AI ​​HDR model to be trained processes the third intermediate image to obtain an output image. The sample image data is obtained based on the camera component, and the sample image data includes information such as the sample scene, color, sharpening, white balance, and contrast corresponding to the sample image.

[0113] In some embodiments, step 204 includes: inputting the first environmental data and the first sample ISP parameter set into the image processing model to be trained to obtain the output image.

[0114] In step 205, the terminal calculates the output image to obtain a first loss function value, and based on the first loss function value, updates the model parameters of the image processing model to be trained.

[0115] In some embodiments, a preset loss function is obtained, and a preset label prediction and an output image are calculated based on the preset loss function to obtain a first loss function value.

[0116] In some embodiments, after step 205, the following may be included: deploying the trained AI HDR model in the chip, the chip obtaining the image data to be processed, inputting the image data to be processed into the ISP, the ISP processing the image data to be processed based on the ISP parameters to obtain a fourth intermediate image, inputting the fourth intermediate image into the trained AI HDR model, and processing the fourth intermediate image based on the AI ​​HDR model to obtain a target image. The image data to be processed is obtained based on the camera component, and the image data to be processed includes information such as the actual scene, color, sharpening, white balance, and contrast corresponding to the image to be processed.

[0117] Optionally, after obtaining the target image, the image effect of the target image is obtained to determine whether the image effect of the target image meets the preset target effect. If the image effect of the target image does not meet the preset target effect, the above steps 201 to 205 are repeated until the image effect of the target image meets the preset target effect. By iteratively training the AI ​​HDR model to be trained, the trained AI HDR model can solve the highlight edge problem in a targeted manner.

[0118] The above-mentioned embodiments of the present application are described by taking application in a terminal as an example. In addition, the embodiments of the present application can also be applied to a server.

[0119] In an embodiment of the present application, a first sample ISP parameter is determined according to the required sample scene, and then the AI ​​HDR model is trained based on the first sample ISP parameter to associate the ISP parameter with the AI ​​HDR, so that the trained AI HDR module is associated with the actual scene, which is targeted when dealing with actual problems, thereby saving NPU computing power.

[0120] Figure 3 3 is a schematic diagram of a device 300 for generating an image processing model according to an embodiment of the present application, the device comprising:

[0121] The first acquisition module 301 is used to acquire an initial ISP parameter set, where the initial ISP parameter set includes at least one initial ISP parameter, and each of the at least one initial ISP parameter is set with an initial value; the initial ISP parameter set indicates an initial scene corresponding to an initial image, and the initial ISP parameter indicates an image effect of the initial image.

[0122] The screening module 302 is used to screen the at least one initial ISP parameter to obtain at least one target ISP parameter.

[0123] The first adjustment module 303 is used to adjust the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value based on the sample scene, so as to obtain a first sample ISP parameter set, wherein the first sample ISP parameter set indicates the sample scene corresponding to the sample image.

[0124] The training module 304 is used to input the first sample ISP parameter set into the image processing model to be trained to obtain an output image.

[0125] The updating module 305 is used to calculate the output image to obtain a first loss function value, and based on the first loss function value, update the model parameters of the image processing model to be trained.

[0126] In a possible implementation, the device further includes:

[0127] The second adjustment module is used to adjust each of the at least one initial ISP parameter at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter.

[0128] In a possible implementation, the second adjustment module includes:

[0129] The first determining unit is configured to use, for one adjustment, one of the at least one initial ISP parameter as the ISP parameter to be adjusted.

[0130] The first adjustment unit is used to adjust the initial value of the ISP parameter to be adjusted to a corresponding adjustment value to obtain the adjusted ISP parameter.

[0131] An acquisition unit is used to acquire an image effect of the adjusted image based on the adjusted ISP parameter and the remaining initial ISP parameters in the at least one initial ISP parameter.

[0132] In a possible implementation, the screening module includes:

[0133] The screening unit is used to screen the at least one initial ISP parameter based on the image effect of the adjusted image to obtain the at least one target ISP parameter.

[0134] In a possible implementation, the first adjustment module 303 includes:

[0135] A second adjustment unit, configured to adjust an initial value of each target ISP parameter of the at least one target ISP parameter to a corresponding first candidate value, so as to obtain a corresponding first candidate image;

[0136] The second acquisition unit is used to acquire a comparison result between the image effect of the first candidate image and the image effect of the initial image based on the sample scene.

[0137] A determination unit is configured to, if the comparison result meets a preset condition, use the first candidate value of each target ISP parameter in the at least one target ISP parameter as the target value of each target ISP parameter in the at least one target ISP parameter; or, if the comparison result does not meet the preset condition, continue to adjust the first candidate value of each target ISP parameter in the at least one target ISP parameter until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained.

[0138] In a possible implementation, the device further includes:

[0139] The second acquisition module is used to acquire first environment data; the first environment data indicates the environment corresponding to the first sample ISP parameter set.

[0140] In a possible implementation, the training module 304 includes:

[0141] An input unit is used to input the first environmental data and the first sample ISP parameter set into the image processing model to be trained to obtain the output image.

[0142] It should be noted that: when the image processing model generation device provided in the above embodiment performs the corresponding steps, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the image processing model generation device provided in the above embodiment and the image processing model generation method embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0143] In an embodiment of the present application, a first sample ISP parameter is determined according to the required sample scene, and then the AI ​​HDR model is trained based on the first sample ISP parameter, and the ISP is associated with the AI ​​HDR, so that the trained AI HDR module is associated with the actual scene, which is targeted when processing actual problems, thereby saving NPU computing power.

[0144] An embodiment of the present application also provides a computer device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the above-mentioned method for generating an image processing model when executing the computer program.

[0145] Taking computer equipment as the terminal as an example, Figure 4 This is a schematic diagram of the structure of a terminal provided in an embodiment of the present application, see Figure 4 The terminal 400 may be a smart phone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer or a desktop computer. The terminal 400 may also be called a user device, a portable terminal, a laptop terminal, a desktop terminal or other names.

[0146] Typically, the terminal 400 includes a processor 401 and a memory 402 .

[0147] The processor 401 may include one or more processing cores, such as a 4-core processor, a 4-core processor, etc. The processor 401 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 401 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 401 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 401 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0148] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 402 is used to store at least one program code, which is used to be executed by the processor 401 to implement the process executed by the terminal in the method for generating an image processing model provided in the method embodiment of the present application.

[0149] In some embodiments, the terminal 400 may further optionally include: a peripheral device interface 403 and at least one peripheral device. The processor 401, the memory 402 and the peripheral device interface 403 may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface 403 via a bus, a signal line or a circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 404, a display screen 405, a camera assembly 406, an audio circuit 407 and a power supply 408.

[0150] The peripheral device interface 403 can be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 401 and the memory 402. In some embodiments, the processor 401, the memory 402 and the peripheral device interface 403 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 401, the memory 402 and the peripheral device interface 403 can be implemented on a separate chip or circuit board, which is not limited in the embodiments of the present application.

[0151] The radio frequency circuit 404 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 404 communicates with the communication network and other communication devices through electromagnetic signals. The radio frequency circuit 404 converts the electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. In some embodiments, the radio frequency circuit 404 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, and the like. The radio frequency circuit 404 can communicate with other terminals through at least one wireless communication protocol. The wireless communication protocol includes, but is not limited to: a metropolitan area network, various generations of mobile communication networks (2G, 3G, 4G and 5G), a wireless local area network and / or a WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 404 may also include circuits related to NFC (Near Field Communication), which is not limited in this application.

[0152] The display screen 405 is used to display the UI (User Interface). The UI may include graphics, text, icons, videos, and any combination thereof. When the display screen 405 is a touch display screen, the display screen 405 also has the ability to collect touch signals on the surface or above the surface of the display screen 405. The touch signal can be input to the processor 401 as a control signal for processing. At this time, the display screen 405 can also be used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, the display screen 405 can be one, set on the front panel of the terminal 400; in other embodiments, the display screen 405 can be at least two, respectively set on different surfaces of the terminal 400 or in a folding design; in other embodiments, the display screen 405 can be a flexible display screen, set on the curved surface or folding surface of the terminal 400. Even, the display screen 405 can also be set to a non-rectangular irregular shape, that is, a special-shaped screen. The display screen 405 can be made of materials such as LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode, organic light-emitting diode).

[0153] The camera assembly 406 is used to capture images or videos. In some embodiments, the camera assembly 406 includes a front camera and a rear camera. Typically, the front camera is disposed on the front panel of the terminal, and the rear camera is disposed on the back of the terminal. In some embodiments, there are at least two rear cameras, which are any one of a main camera, a depth of field camera, a wide-angle camera, and a telephoto camera, so as to realize the fusion of the main camera and the depth of field camera to realize the background blur function, the fusion of the main camera and the wide-angle camera to realize the panoramic shooting and VR (Virtual Reality) shooting function or other fusion shooting functions. In some embodiments, the camera assembly 406 may also include a flash. The flash may be a monochrome temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm light flash and a cold light flash, which can be used for light compensation at different color temperatures.

[0154] The audio circuit 407 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, and convert the sound waves into electrical signals and input them into the processor 401 for processing, or input them into the radio frequency circuit 404 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the terminal 400. The microphone may also be an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 401 or the radio frequency circuit 404 into sound waves. The speaker may be a traditional film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 407 may also include a headphone jack.

[0155] The power supply 408 is used to power various components in the terminal 400. The power supply 408 can be an alternating current, a direct current, a disposable battery, or a rechargeable battery. When the power supply 408 includes a rechargeable battery, the rechargeable battery can support wired charging or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0156] In some embodiments, the terminal 400 further includes one or more sensors 409 . The one or more sensors 409 include, but are not limited to: an acceleration sensor 410 , a gyroscope sensor 411 , a pressure sensor 412 , an optical sensor 413 , and a proximity sensor 414 .

[0157] The acceleration sensor 410 can detect the magnitude of acceleration on the three coordinate axes of the coordinate system established by the terminal 400. For example, the acceleration sensor 410 can be used to detect the components of gravity acceleration on the three coordinate axes. The processor 401 can control the display screen 405 to display the user page in a horizontal view or a vertical view according to the gravity acceleration signal collected by the acceleration sensor 410. The acceleration sensor 410 can also be used for collecting game or user motion data.

[0158] The gyro sensor 411 can detect the body direction and rotation angle of the terminal 400, and the gyro sensor 411 can cooperate with the acceleration sensor 410 to collect the user's three-dimensional actions on the terminal 400. The processor 401 can implement the following functions based on the data collected by the gyro sensor 411: motion sensing (such as changing the UI according to the user's tilt operation), image stabilization during shooting, game control, and inertial navigation.

[0159] The pressure sensor 412 can be set on the side frame of the terminal 400 and / or the lower layer of the display screen 405. When the pressure sensor 412 is set on the side frame of the terminal 400, it can detect the user's holding signal of the terminal 400, and the processor 401 performs left and right hand recognition or shortcut operation according to the holding signal collected by the pressure sensor 412. When the pressure sensor 412 is set on the lower layer of the display screen 405, the processor 401 controls the operability controls on the UI page according to the user's pressure operation on the display screen 405. The operability controls include at least one of a button control, a scroll bar control, an icon control, and a menu control.

[0160] The optical sensor 413 is used to collect the ambient light intensity. In one embodiment, the processor 401 can control the display brightness of the display screen 405 according to the ambient light intensity collected by the optical sensor 413. Specifically, when the ambient light intensity is high, the display brightness of the display screen 405 is increased; when the ambient light intensity is low, the display brightness of the display screen 405 is reduced. In another embodiment, the processor 401 can also dynamically adjust the shooting parameters of the camera assembly 406 according to the ambient light intensity collected by the optical sensor 413.

[0161] The proximity sensor 414, also called a distance sensor, is usually arranged on the front panel of the terminal 400. The proximity sensor 414 is used to collect the distance between the user and the front of the terminal 400. In one embodiment, when the proximity sensor 414 detects that the distance between the user and the front of the terminal 400 is gradually decreasing, the processor 401 controls the display screen 405 to switch from the screen-on state to the screen-off state; when the proximity sensor 414 detects that the distance between the user and the front of the terminal 400 is gradually increasing, the processor 401 controls the display screen 405 to switch from the screen-off state to the screen-on state.

[0162] Those skilled in the art will understand that Figure 4 The structure shown in the figure does not constitute a limitation on the terminal 400, and the terminal 400 may include more or less components than those shown in the figure, or combine some components, or adopt a different component arrangement.

[0163] Take the computer device as a server as an example. Figure 5It is a structural diagram of a server provided in an embodiment of the present application. The server 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (Central Processing Units, CPU) 501 and one or more memories 502, wherein the one or more memories 502 store at least one computer program, and the at least one computer program is loaded and executed by the one or more processors 501 to implement the above-mentioned method for generating the image processing model. Of course, the server 500 may also have components such as a wired or wireless network interface, a keyboard, and an input and output interface for input and output. The server 500 may also include other components for realizing device functions, which will not be described in detail here.

[0164] The embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned method for generating an image processing model. Optionally, the computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, etc.

[0165] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0166] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for generating an image processing model, characterized in that: include: Acquire an initial ISP parameter set, the initial ISP parameter set comprising at least one initial ISP parameter, each of the at least one initial ISP parameter being set with an initial value; the initial ISP parameter set indicating an initial scene corresponding to an initial image, and the initial ISP parameter indicating an image effect of the initial image; Screening the at least one initial ISP parameter to obtain at least one target ISP parameter; Based on the sample scene, adjusting the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value to obtain a first sample ISP parameter set, where the first sample ISP parameter set indicates the sample scene corresponding to the sample image; The adjusting, based on the sample scenario, the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value includes: Adjusting an initial value of each target ISP parameter of the at least one target ISP parameter to a corresponding first candidate value to obtain an image effect of a corresponding first candidate image; Based on the sample scene, obtaining a comparison result between an image effect of the first candidate image and an image effect of the initial image; If the comparison result meets the preset condition, the first candidate value of each target ISP parameter in the at least one target ISP parameter is used as the target value of each target ISP parameter in the at least one target ISP parameter; or if the comparison result does not meet the preset condition, the first candidate value of each target ISP parameter in the at least one target ISP parameter is continuously adjusted until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained; Inputting the first sample ISP parameter set into the image processing model to be trained to obtain an output image; The output image is calculated to obtain a first loss function value, and based on the first loss function value, the model parameters of the image processing model to be trained are updated.

2. The method according to claim 1, characterized in that After obtaining the initial ISP parameter set, it also includes: Each of the at least one initial ISP parameter is adjusted at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter.

3. The method according to claim 2, characterized in that The step of adjusting each of the at least one initial ISP parameter at least once to obtain an adjusted ISP parameter corresponding to each of the at least one initial ISP parameter includes: For one adjustment, taking one of the at least one initial ISP parameter as the ISP parameter to be adjusted; Adjusting the initial value of the ISP parameter to be adjusted to a corresponding adjustment value to obtain the adjusted ISP parameter; Based on the adjusted ISP parameters and the remaining initial ISP parameters in the at least one initial ISP parameter, an image effect of the adjusted image is obtained.

4. The method according to claim 3, characterized in that The step of screening at least one of the initial ISP parameters to obtain at least one target ISP parameter includes: The at least one initial ISP parameter is screened based on the image effect of the adjusted image to obtain the at least one target ISP parameter.

5. The method according to claim 1, characterized in that Before inputting the first sample ISP parameter set into the image processing model to be trained to obtain the output image, the method further includes: Acquire first environment data; the first environment data indicates the environment corresponding to the first sample ISP parameter set.

6. The method according to claim 5, characterized in that The step of inputting the first sample ISP parameter set into the image processing model to be trained to obtain an output image includes: The first environmental data and the first sample ISP parameter set are input into the image processing model to be trained to obtain the output image.

7. A device for generating an image processing model, characterized in that: include: A first acquisition module is used to acquire an initial ISP parameter set, wherein the initial ISP parameter set includes at least one initial ISP parameter, and each of the at least one initial ISP parameter is set with an initial value; the initial ISP parameter set indicates an initial scene corresponding to an initial image, and the initial ISP parameter indicates an image effect of the initial image; A screening module, configured to screen the at least one initial ISP parameter to obtain at least one target ISP parameter; A first adjustment module, configured to adjust, based on a sample scene, an initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value, to obtain a first sample ISP parameter set, wherein the first sample ISP parameter set indicates the sample scene corresponding to the sample image; The adjusting, based on the sample scenario, the initial value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set to a corresponding target value includes: Adjusting an initial value of each target ISP parameter of the at least one target ISP parameter to a corresponding first candidate value to obtain an image effect of a corresponding first candidate image; Based on the sample scene, obtaining a comparison result between an image effect of the first candidate image and an image effect of the initial image; If the comparison result meets the preset condition, the first candidate value of each target ISP parameter in the at least one target ISP parameter is used as the target value of each target ISP parameter in the at least one target ISP parameter; or if the comparison result does not meet the preset condition, the first candidate value of each target ISP parameter in the at least one target ISP parameter is continuously adjusted until the target value of each target ISP parameter in the at least one target ISP parameter in the initial ISP parameter set is obtained; A training module, used for inputting the first sample ISP parameter set into the image processing model to be trained to obtain an output image; An updating module is used to calculate the output image to obtain a first loss function value, and based on the first loss function value, update the model parameters of the image processing model to be trained.

8. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory is used to store at least one program, and the at least one program is loaded by the processor and executes the method for generating an image processing model according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one program, and the at least one program is loaded and executed by the processor to implement the method for generating an image processing model according to any one of claims 1 to 6.

Images