Training method, system, electronic device and mobile device for chromatic aberration correction model

Abstract

The present application discloses a training method, system, electronic device and mobile device for a color difference correction model. The training method and system specifically collect a first color card image taken by a standard eye diagnosis instrument and a second color card image taken by a mobile device; use a color block positioning deep learning network model to process the first color card image and the second color card image to obtain all first image blocks in the first color card image and all second image blocks in the second color card image; extract the first color value of each first image block and the second color value of each second image block; perform iterative regression solution processing on all first color values ​​and all second color values ​​to obtain a color difference correction model. When the color difference correction model is configured on a corresponding mobile device, since the mobile device can obtain an eye image with the same parameters as the standard eye diagnosis instrument based on the color difference correction model, the mobile device can obtain reliable and accurate diagnosis results based on the corresponding eye diagnosis system.

Description

Technical Field

[0001] The present application relates to the field of medical equipment technology, and more specifically, to a training method, system, electronic device and mobile device for a chromatic aberration correction model. Background Art

[0002] Eye diagnosis was first recorded in the classic Chinese medicine book "Huangdi Neijing" and is called "five-color micro-diagnosis". Chinese medicine believes that the five colors in the eyes, namely red, yellow, white, black and green, correspond to the five internal organs of the human body. The theory of Chinese medicine eye diagnosis believes that the color of the sclera and blood vessels on the eyeball reflect the health status of the corresponding area of ​​the human body. Therefore, the key point of developing computer intelligent eye diagnosis auxiliary diagnosis and treatment equipment is that the collected eye image data can highly restore its true color so that it can conform to the observation results of Chinese medicine.

[0003] The existing standard eye diagnosis equipment has achieved breakthrough results in health management by collecting high-quality eye images and analyzing them, but it is difficult to popularize it among the grassroots population due to its large size, high cost and high price. With the popularization of mobile devices such as smartphones, relying on the high-quality imaging cameras on mobile devices, it has become possible to deploy eye diagnosis systems on mobile devices.

[0004] The accuracy of the color of the eye image is crucial for the subsequent diagnosis and analysis. However, different models of mobile devices have different camera hardware and ISP (Image Signal Processing) modules, which will lead to large differences in the color of the eye image collected by different mobile phones under the same light source environment for the same subject. This will inevitably have a very large impact on the diagnosis and it is impossible to obtain reliable and accurate diagnosis results. Therefore, solving the imaging color differences of different models of mobile devices is the key to ensuring that the eye diagnosis system on different models of mobile devices can obtain reliable and accurate diagnosis results. Summary of the invention

[0005] In view of this, the present application provides a training method, system, electronic device and mobile device for a chromatic aberration correction model, which are used to obtain a tool that can correct the chromatic aberration of an eye image, so that a mobile device equipped with such a chromatic aberration correction model can obtain reliable and accurate diagnosis results based on a corresponding eye diagnosis system.

[0006] In order to achieve the above objectives, the proposed solution is as follows:

[0007] A training method for a chromatic aberration correction model is applied to electronic equipment, and the training method comprises the following steps:

[0008] Acquire a first color card image and a second color card image, wherein the first color card image is obtained by photographing a dedicated color card with a standard eye diagnosis instrument, and the second color card image is obtained by photographing the dedicated color card with a mobile device to be calibrated, and the first color card image includes a plurality of first image blocks;

[0009] Processing the second color card image using a color block positioning deep learning network model to obtain all second image blocks in the second color card image;

[0010] Extracting a first color value of each of the first image blocks and a second color value of each of the second image blocks;

[0011] All the first color values ​​and all the second color values ​​are subjected to iterative regression solution processing to obtain the color difference correction model.

[0012] Optionally, the exclusive color card includes a plurality of first color blocks, a plurality of second color blocks, and a plurality of third color blocks arranged on a background plate of 18 degrees of gray, wherein:

[0013] The plurality of first color blocks include some or all of the color blocks in the 24-color ColorCheck color card;

[0014] The plurality of second color blocks include a plurality of color blocks matching the parameters of the mobile device;

[0015] The third color block includes a plurality of color blocks of the same color as those usually appearing in a conventional eye image diagram.

[0016] Optionally, the second color card image is captured by the mobile device under illumination by multiple illumination lamps at different angles, and each second color value is an average color value of the second image block at different illumination angles.

[0017] Optionally, the following steps are also included:

[0018] The correction effect of the chromatic aberration correction model is evaluated.

[0019] A training system for a chromatic aberration correction model, applied to electronic equipment, the training system comprising:

[0020] An image transmission acquisition module is configured to acquire a first color card image and a second color card image, wherein the first color card image is obtained by photographing a dedicated color card by a standard eye diagnosis instrument, and the second color card image is obtained by photographing the dedicated color card by a mobile device to be calibrated, and the first color card image includes a plurality of first image blocks;

[0021] a color block positioning module, configured to process the second color card image using a color block positioning deep learning network model to obtain all second image blocks in the second color card image;

[0022] A color value extraction module, configured to extract a first color value of each of the first image blocks and a second color value of each of the second image blocks;

[0023] The model training module is configured to perform iterative regression solution processing on all the first color values ​​and all the second color values ​​to obtain the color difference correction model.

[0024] Optionally, the exclusive color card includes a plurality of first color blocks, a plurality of second color blocks, and a plurality of third color blocks arranged on a background plate of 18 degrees of gray, wherein:

[0025] The plurality of first color blocks include some or all of the color blocks in the 24-color ColorCheck color card;

[0026] The plurality of second color blocks include a plurality of color blocks matching the parameters of the mobile device;

[0027] The third color block includes a plurality of color blocks of the same color as those usually appearing in a conventional eye image diagram.

[0028] Optionally, the second color card image is captured by the mobile device under illumination by multiple illumination lamps at different angles, and each second color value is an average color value of the second image block at different illumination angles.

[0029] Optionally, also include:

[0030] The effect evaluation module is configured to evaluate the correction effect of the chromatic aberration correction model.

[0031] An electronic device, optionally comprising at least one processor and a memory connected to the processor, wherein:

[0032] The memory is used to store computer programs or instructions;

[0033] The processor is used to execute the computer program or instruction so that the electronic device implements the training method of the chromatic aberration correction model as described above.

[0034] A mobile device is configured with the chromatic aberration correction model as described above.

[0035] It can be seen from the above technical scheme that the present application discloses a training method, system, electronic device and mobile device for a color difference correction model. The training method and system specifically collect a first color card image taken by a standard eye diagnosis instrument and a second color card image taken by a mobile device; use a color block positioning deep learning network model to process the first color card image and the second color card image to obtain all first image blocks in the first color card image and all second image blocks in the second color card image; extract the first color value of each first image block and the second color value of each second image block; perform iterative regression solution processing on all first color values ​​and all second color values ​​to obtain a color difference correction model. When the color difference correction model is configured on a corresponding mobile device, since the mobile device can obtain an eye image with the same parameters as the standard eye diagnosis instrument based on the color difference correction model, the mobile device can obtain reliable and accurate diagnosis results based on the corresponding eye diagnosis system.

[0036] In the color block positioning and color value extraction stage of the technical solution of the present application, a color block positioning model based on yolov4 is trained, the color categories of the color blocks in the color card are not considered, only the color blocks are positioned, the learning of the color block positions is strengthened during the training stage, and then the color values ​​of the color blocks are obtained based on the color block position relationship. This method is just suitable for color correction of cameras of different mobile devices, when the color block color categories are different, the number of color blocks is different, and even the color block layout is irregular and not fixed. When locating the color blocks using the trained positioning model, except for adjusting the size of the color card image of the input model, no preliminary processing of the color card image is required, and the color block positioning is fast and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 A flowchart of a method for training a chromatic aberration correction model according to an embodiment of the present application;

[0039] Figure 2 A flowchart of another method for training a chromatic aberration correction model according to an embodiment of the present application;

[0040] Figure 3 A block diagram of a training system for a chromatic aberration correction model according to an embodiment of the present application;

[0041] Figure 4 A block diagram of a training system for a chromatic aberration correction model according to an embodiment of the present application;

[0042] Figure 5 A block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0044] Embodiment 1

[0045] Figure 1 The present invention is a flowchart of a method for training a chromatic aberration correction model according to an embodiment of the present application.

[0046] The chromatic aberration correction model of the present embodiment is used to perform chromatic aberration correction on an eye image chart or other target image required for visual diagnosis captured by a mobile device equipped with a visual diagnosis system, so that the color of the eye image chart or other target image is the same as the color of the image captured by a standard visual diagnosis instrument, so that the mobile device can obtain the same diagnostic result as the standard visual diagnosis instrument based on the visual diagnosis system configured thereon, that is, correction based on the chromatic aberration correction model can enable the mobile device to obtain reliable and accurate diagnostic results.

[0047] like Figure 1 As shown, the training method provided in this embodiment is used to train the above-mentioned chromatic aberration correction model. The training method is implemented based on an electronic device, and the electronic device can be understood as a computer or server with information processing and data calculation. The training method specifically includes the following steps:

[0048] S1. Collect a first color card image and a second color card image.

[0049] Specifically, the first color card image is acquired from the standard eye examination instrument, and the mobile device to be calibrated acquires the second color card image, where the second color card image is the image that needs to be calibrated. The first color card image is obtained by the standard eye examination instrument by photographing the exclusive color card provided in this embodiment; the second color card image is obtained by the mobile device by photographing the above exclusive color card.

[0050] The exclusive color card in this embodiment includes a plurality of first color blocks, a second color block, and a third color block set on an 18-degree gray background plate. The first color block here refers to part or all of the color blocks in the 24-color ColorCheck color card; the second color block refers to a plurality of color blocks that match the parameters of the mobile device to be calibrated; the third color block includes a plurality of color blocks that are the same as several colors that often appear in the conventional eye image diagram. In order to facilitate subsequent operations, we uniformly number all color blocks, and the size of each color block is selected as a 3mm*3mm square.

[0051] Based on the theory of traditional Chinese medicine, five colors commonly appear in the eyes: red, yellow, white, black, and green. Therefore, the colors commonly appearing in the conventional eye image diagram referred to in this application include the above five colors: red, yellow, white, black, and green, and may also include other colors discovered in practice.

[0052] In addition, since the camera window of the standard eye diagnosis instrument is relatively small, each color block needs to be photographed separately in sequence. The light source system of the eye diagnosis instrument includes point light sources in four directions: up, down, left, and right. Therefore, the standard eye diagnosis instrument photographs the corresponding color card when the four direction lights illuminate the cavity respectively, and the obtained first color card image includes multiple first image blocks with the same number of color blocks. When shooting the exclusive color card, the mobile device to be calibrated is illuminated by light sources in four directions and shoots at the same time, and obtains four images as the above-mentioned second color card image, that is, the second color card image is actually a group of images.

[0053] S2. Locate a second image block in a second color card image.

[0054] Here, the number of second image blocks is the same as the number of color blocks of the exclusive color card. Since the second color card image is a group of images, each image is processed using the color block positioning deep network model to obtain the second image block corresponding to each color block. Each second image block obtained here also includes a group of color block images.

[0055] S3. Extract the first color value and the second color value.

[0056] That is, the color value of each first image block in the first color card image and the color value of each second image block are extracted. For easy distinction, the color value of each first image block is described as a first color value, and the color value of each second image block is described as a second color value.

[0057] When extracting the color value of each image block, first extract the first pixel value of the RGB channel, and then convert the first pixel value from the RGB color space to the CIE Lab color space to obtain the second pixel value of the Lab channel. Since each of the above second image blocks is a group of color block images, it naturally includes a group of second pixel values. For the corresponding mobile device, the multiple second pixel values ​​of the same image block are averaged to obtain the color value of the corresponding second image block, that is,

[0058]

[0059] Here, F(x) represents the conversion from RGB color space to CIE Lab color space.

[0060] S4. Perform model training using the first color value and the second color value.

[0061] After obtaining the above-mentioned multiple first color values ​​and multiple second color values, the first color value is used as the color reference value of the corresponding color block, and the corresponding second color value is used as the actual color observation value of the mobile device to be corrected. The color difference correction model F(x) based on CCM (Color Correction Matrix) is solved by iterative regression, and the obtained model parameters are saved for color correction of color card images and eye image diagrams.

[0062] Here, the number of training iterations is determined by the set maximum number of iterations or the iteration error. When the number of iterations reaches the maximum or the iteration error is less than the set threshold, the model training stops. The relationship model structure between the actual color observation value and the reference value, that is, the color difference correction model is as follows:

[0063]

[0064] R in , G in , B in are the values ​​of the image to be corrected on the R, G, and B channels respectively, R out , G out ,,B out is the value of the RGB three channels of the image after color correction. The model needs to solve c 00 ~c 22 , C1~C3, g1~g3, a total of 15 parameters.

[0065] In the solution process, the R, G, and B values ​​of the actual observed color block are used as the R in , G in , B inThe mean square error (MSE) is used as the objective function, and the model parameters are obtained through iterative training, that is, the MSE is continuously optimized to minimize it. The process is as follows:

[0066] 1) Initialize 15 parameters of the model;

[0067] 2)iter=0; / / iter indicates the number of iterations

[0068] 3) When iter < total number of iterations, execute:

[0069] according to calculate

[0070] calculate / / R std , G std , B std Indicates the color reference value of the color block

[0071] iter++;

[0072] If MSE < threshold:

[0073] Stop iteration.

[0074] It can be seen from the above technical solution that this embodiment provides a method for training a color difference correction model, which is applied to electronic devices, specifically collecting a first color card image taken by a standard eye diagnosis instrument and a second color card image taken by a mobile device; using a color block positioning deep learning network model to process the first color card image and the second color card image to obtain all first image blocks in the first color card image and all second image blocks in the second color card image; extracting the first color value of each first image block and the second color value of each second image block; performing iterative regression solution processing on all first color values ​​and all second color values ​​to obtain a color difference correction model. When the color difference correction model is configured on a corresponding mobile device, since the mobile device can obtain an eye image with the same parameters as the standard eye diagnosis instrument based on the color difference correction model, the mobile device can obtain reliable and accurate diagnosis results based on the corresponding eye diagnosis system.

[0075] In a specific implementation of the present application, the following steps are also included: Figure 2 As shown:

[0076] S5. Evaluate the correction effect of the chromatic aberration correction model.

[0077] The color difference correction model obtained above is used to correct the color chart images and eye image images taken by different mobile devices to be corrected, and the correction effect of the color difference correction model is evaluated by both quantitative and qualitative methods.

[0078] Quantitative evaluation includes calculating the color difference of color blocks and comparing the number of symptoms obtained from the calibrated eye image chart with the symptoms obtained from the standard eye diagnosis instrument; qualitative evaluation includes observing the consistency of color between the eye image chart taken by the calibrated mobile eye diagnosis instrument and the eye image chart taken by the standard eye diagnosis instrument.

[0079] Here, the corrected color block color value obtained in this solution is used, and the color difference δLab of the color block after color correction is calculated, including the average value, standard deviation and maximum value of the color difference of all color blocks. The calculation formula of the color difference δLab is as follows:

[0080]

[0081] Here, i represents the color block number. The color difference mean, standard deviation, and maximum value are

[0082]

[0083]

[0084]

[0085] N represents the total number of color blocks on the color card, std(x) represents the standard deviation function, and Max(x) represents the maximum value function.

[0086] If after color correction, the average color difference value of the color block is less than or equal to 4, the standard deviation is less than or equal to 2, and the maximum value does not exceed 15, the number of TCM syndromes obtained by the eye image analysis of the same subject and the TCM syndromes obtained by the standard eye diagnosis instrument increases after color correction, and at least 2 are the same as those of the standard eye diagnosis instrument, and the color of the eye image diagram observed by the human eye is close to or the same as that of the standard eye diagnosis instrument, then the color correction result can be accepted, and the obtained color difference correction model is encapsulated in the corresponding mobile device to serve the eye diagnosis system in the mobile device. Otherwise, reselect the color block combination through the above scheme to formulate a dedicated color card, and then re-execute the subsequent training plan.

[0087] In the color block positioning and color value extraction stage of the technical solution of the present application, a color block positioning model based on yolov4 is trained, and the color categories of the color blocks in the color card are not considered. The color blocks are only positioned, and the learning of the color block positions is strengthened during the training stage, and then the color values ​​of the color blocks are obtained based on the color block position relationship. This method is suitable for color correction of cameras of different mobile devices, when the color block color categories are different, the number of color blocks is different, and even the color block layout is irregular and not fixed. When locating the color blocks using the trained positioning model, except for adjusting the size of the color card image of the input model, no preliminary processing of the color card image is required, and the color block positioning is fast and accurate.

[0088] Embodiment 2

[0089] Figure 3 A block diagram of a training system for a chromatic aberration correction model according to an embodiment of the present application.

[0090] like Figure 3 As shown, the training system provided in this embodiment is used to train the above-mentioned chromatic aberration correction model. The training system can be understood as a functional module of an electronic device or the electronic device itself. The electronic device can be understood as a computer or server with information processing and data calculation. The training system specifically includes an image acquisition module 10, a color block positioning module 20, a color value extraction module 30 and a model training module 40.

[0091] The image acquisition module is used to acquire the first color card image and the second color card image.

[0092] Specifically, the first color card image is collected from the standard eye examination instrument, and the second color card image is collected from the mobile device to be calibrated, where the second color card image is the image that needs to be calibrated. The first color card image is obtained by the standard eye examination instrument by photographing the exclusive color card provided in this embodiment; the second color card image is obtained by the mobile device by photographing the above exclusive color card.

[0093] The exclusive color card in this embodiment includes a plurality of first color blocks, a second color block, and a third color block set on an 18-degree gray background plate. The first color block here refers to part or all of the color blocks in the 24-color ColorCheck color card; the second color block refers to a plurality of color blocks that match the parameters of the mobile device to be calibrated; the third color block includes a plurality of color blocks that are the same as several colors that often appear in the conventional eye image diagram. In order to facilitate subsequent operations, we uniformly number all color blocks, and the size of each color block is selected as a 3mm*3mm square.

[0094] Based on the theory of traditional Chinese medicine, five colors commonly appear in the eyes: red, yellow, white, black, and green. Therefore, the colors commonly appearing in the conventional eye image diagram referred to in this application include the above five colors: red, yellow, white, black, and green, and may also include other colors discovered in practice.

[0095] In addition, since the camera window of the standard eye diagnosis instrument is relatively small, each color block needs to be photographed separately in sequence. The light source system of the eye diagnosis instrument includes point light sources in four directions: up, down, left, and right. Therefore, the standard eye diagnosis instrument photographs the corresponding color card when the four direction lights illuminate the cavity respectively, and the obtained first color card image includes multiple first image blocks with the same number of color blocks. When shooting the exclusive color card, the mobile device to be calibrated is illuminated by light sources in four directions and shoots at the same time, and obtains four images as the above-mentioned second color card image, that is, the second color card image is actually a group of images.

[0096] The color block positioning module is used to locate the second image block in the image at the second moment.

[0097] Here, the number of second image blocks is the same as the number of color blocks of the exclusive color card. Since the second color card image is a group of images, each image is processed using the color block positioning deep network model to obtain the second image block corresponding to each color block. Each second image block obtained here also includes a group of color block images.

[0098] The color value extraction module is used to extract a first color value and a second color value.

[0099] That is, the color value of each first image block in the first color card image and the color value of each second image block are extracted. For easy distinction, the color value of each first image block is described as a first color value, and the color value of each second image block is described as a second color value.

[0100] When extracting the color value of each image block, first extract the first pixel value of the RGB channel, and then convert the first pixel value from the RGB color space to the CIE Lab color space to obtain the second pixel value of the Lab channel. Since each of the above second image blocks is a group of color block images, it naturally includes a group of second pixel values. For the corresponding mobile device, the multiple second pixel values ​​of the same image block are averaged to obtain the color value of the corresponding second image block, that is,

[0101]

[0102] Here, F(x) represents the conversion from RGB color space to CIE Lab color space.

[0103] The model training module is used to perform model training using the first color value and the second color value.

[0104] After obtaining the above-mentioned multiple first color values ​​and multiple second color values, the first color value is used as the color reference value of the corresponding color block, and the corresponding second color value is used as the actual color observation value of the mobile device to be corrected. The color difference correction model F(x) based on CCM (Color Correction Matrix) is solved by iterative regression, and the obtained model parameters are saved for color correction of color card images and eye image diagrams.

[0105] Here, the number of training iterations is determined by the set maximum number of iterations or the iteration error. When the number of iterations reaches the maximum or the iteration error is less than the set threshold, the model training stops. The relationship model structure between the actual color observation value and the reference value, that is, the color difference correction model is as follows:

[0106]

[0107] R in , G in , B in are the values ​​of the image to be corrected on the R, G, and B channels respectively, R out , G out ,,B out is the value of the RGB three channels of the image after color correction. The model needs to solve c 00 ~c 22 , C1~C3, g1~g3, a total of 15 parameters.

[0108] In the solution process, the R, G, and B values ​​of the actual observed color block are used as the R in , G in , B in The mean square error (MSE) is used as the objective function, and the model parameters are obtained through iterative training, that is, the MSE is continuously optimized to minimize it. The process is as follows:

[0109] 1) Initialize 15 parameters of the model;

[0110] 2)iter=0; / / iter indicates the number of iterations

[0111] 3) When iter < total number of iterations, execute:

[0112] according to calculate

[0113] calculate / / R std , G std , B std Indicates the color reference value of the color block

[0114] iter++;

[0115] If MSE < threshold:

[0116] Stop iteration.

[0117] It can be seen from the above technical solution that this embodiment provides a training system for a color difference correction model, which is applied to electronic devices, specifically to collect a first color card image taken by a standard eye diagnosis instrument and a second color card image taken by a mobile device; use a color block positioning deep learning network model to process the first color card image and the second color card image to obtain all first image blocks in the first color card image and all second image blocks in the second color card image; extract the first color value of each first image block and the second color value of each second image block; perform iterative regression solution processing on all first color values ​​and all second color values ​​to obtain a color difference correction model. When the color difference correction model is configured on a corresponding mobile device, since the mobile device can obtain an eye image with the same parameters as the standard eye diagnosis instrument based on the color difference correction model, the mobile device can obtain reliable and accurate diagnosis results based on the corresponding eye diagnosis system.

[0118] In a specific embodiment of the present application, an effect evaluation module 50 is also included, such as Figure 4 As shown:

[0119] The effect evaluation module is used to evaluate the correction effect of the chromatic aberration correction model.

[0120] The color difference correction model obtained above is used to correct the color chart images and eye image images taken by different mobile devices to be corrected, and the correction effect of the color difference correction model is evaluated by both quantitative and qualitative methods.

[0121] Quantitative evaluation includes calculating the color difference of color blocks and comparing the number of symptoms obtained from the calibrated eye image chart with the symptoms obtained from the standard eye diagnosis instrument; qualitative evaluation includes observing the consistency of color between the eye image chart taken by the calibrated mobile eye diagnosis instrument and the eye image chart taken by the standard eye diagnosis instrument.

[0122] Here, the corrected color block color value obtained in this solution is used, and the color difference δLab of the color block after color correction is calculated, including the average value, standard deviation and maximum value of the color difference of all color blocks. The calculation formula of the color difference δLab is as follows:

[0123]

[0124] Here, i represents the color block number. The color difference mean, standard deviation, and maximum value are

[0125]

[0126]

[0127]

[0128] N represents the total number of color blocks on the color card, std(x) represents the standard deviation function, and Max(x) represents the maximum value function.

[0129] If after color correction, the average color difference value of the color block is less than or equal to 4, the standard deviation is less than or equal to 2, and the maximum value does not exceed 15, the number of TCM syndromes obtained by the eye image analysis of the same subject and the TCM syndromes obtained by the standard eye diagnosis instrument increases after color correction, and at least 2 are the same as those of the standard eye diagnosis instrument, and the color of the eye image diagram observed by the human eye is close to or the same as that of the standard eye diagnosis instrument, then the color correction result can be accepted, and the obtained color difference correction model is encapsulated in the corresponding mobile device to serve the eye diagnosis system in the mobile device. Otherwise, reselect the color block combination through the above scheme to formulate a dedicated color card, and then re-execute the subsequent training plan.

[0130] Embodiment 3

[0131] Figure 5 A block diagram of an electronic device according to an embodiment of the present application.

[0132] like Figure 5 As shown, the electronic device provided in this embodiment can be understood as a computer or server with information processing and data computing capabilities, and the electronic device includes at least one processor 101 and a memory 102, which are connected via a data bus 103. The memory is used to store computer programs or instructions, and the processor is used to execute corresponding computer programs or instructions, so that the electronic device implements the training method of the chromatic aberration correction model provided in the first embodiment.

[0133] The training method specifically includes collecting a first color card image taken by a standard eye diagnosis instrument and a second color card image taken by a mobile device; processing the first color card image and the second color card image using a color block positioning deep learning network model to obtain all first image blocks in the first color card image and all second image blocks in the second color card image; extracting the first color value of each first image block and the second color value of each second image block; performing iterative regression solution processing on all first color values ​​and all second color values ​​to obtain a color difference correction model. When the color difference correction model is configured on a corresponding mobile device, since the mobile device can obtain an eye image with the same parameters as the standard eye diagnosis instrument based on the color difference correction model, the mobile device can obtain reliable and accurate diagnosis results based on the corresponding eye diagnosis system.

[0134] Embodiment 4

[0135] This embodiment provides a mobile device, which can be understood as a smart device such as a mobile phone or a tablet computer. The mobile device is provided with a camera for collecting color card images and eye image diagrams, and is also configured with the color difference correction model and visual diagnosis system provided in the above embodiment. The visual diagnosis system can correct the color of the collected eye image diagram based on the color difference correction model, so that it can obtain reliable and accurate diagnosis results based on the visual diagnosis system.

[0136] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0137] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0138] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0139] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0141] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0142] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0143] The technical solution provided by the present invention is introduced in detail above. Specific examples are used in this article to illustrate the principle and implementation mode of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation mode and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for training a color difference correction model, applied to electronic equipment, characterized in that: The training method comprises the steps of: A first color card image is collected by a standard eye examination instrument and a second color card image is collected by a mobile device to be calibrated, wherein the first color card image is obtained by photographing a dedicated color card by the standard eye examination instrument, and the second color card image is obtained by photographing the dedicated color card by the mobile device to be calibrated, wherein the first color card image includes a plurality of first image blocks; the dedicated color card includes a plurality of first color blocks, a plurality of second color blocks, and a plurality of third color blocks arranged on a background plate of 18 degrees of gray, wherein: the plurality of first color blocks include some or all color blocks in a 24-color ColorCheck color card; the plurality of second color blocks include a plurality of color blocks that match the parameters of the mobile device; the third color blocks include a plurality of color blocks that have the same color as those that often appear in a conventional eye image chart; Processing the second color card image using a color block positioning deep learning network model to obtain all second image blocks in the second color card image; Extracting a first color value of each of the first image blocks and a second color value of each of the second image blocks; All the first color values ​​and all the second color values ​​are subjected to iterative regression solution processing to obtain the color difference correction model.

2. The training method according to claim 1, characterized in that: The second color card image is obtained by photographing the mobile device under illumination by multiple illumination lamps at different angles, and each second color value is an average color value of the second image block at different illumination angles.

3. The training method according to claim 1 or 2, characterized in that: Also includes the steps: The correction effect of the chromatic aberration correction model is evaluated.

4. A training system for a color difference correction model, applied to electronic equipment, characterized in that: The training system comprises: The image acquisition module is configured to acquire a first color card image through a standard eye diagnosis instrument and a second color card image through a mobile device to be calibrated, wherein the first color card image is obtained by photographing a dedicated color card by the standard eye diagnosis instrument, and the second color card image is obtained by photographing the dedicated color card by the mobile device to be calibrated, wherein the first color card image includes a plurality of first image blocks; the dedicated color card includes a plurality of first color blocks, a plurality of second color blocks, and a plurality of third color blocks arranged on an 18-degree gray background plate, wherein: the plurality of first color blocks include part or all of the color blocks in a 24-color ColorCheck color card; the plurality of second color blocks include a plurality of color blocks matching the parameters of the mobile device; the third color blocks include a plurality of color blocks of the same color as those often appearing in a conventional eye image chart; a color block positioning module, configured to process the second color card image using a color block positioning deep learning network model to obtain all second image blocks in the second color card image; A color value extraction module, configured to extract a first color value of each of the first image blocks and a second color value of each of the second image blocks; The model training module is configured to perform iterative regression solution processing on all the first color values ​​and all the second color values ​​to obtain the color difference correction model.

5. The training system according to claim 4, characterized in that The second color card image is obtained by photographing the mobile device under illumination by multiple illumination lamps at different angles, and each second color value is an average color value of the second image block at different illumination angles.

6. The training system according to claim 4 or 5, characterized in that: Also includes: The effect evaluation module is configured to evaluate the correction effect of the chromatic aberration correction model.

7. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs or instructions; The processor is used to execute the computer program or instruction so that the electronic device implements the training method of the chromatic aberration correction model as described in any one of claims 1 to 3.

Images