Image detail adjustment method and electronic device

Abstract

Embodiments of the present application disclose an image detail adjustment method and an electronic device, and relate to the technical field of image processing. The adjustment method comprises: in response to a first operation, performing image processing on a raw image to obtain a noise image, a denoised image and a to-be-output image; determining a target detail intensity distribution; obtaining a first detail difference image according to the noise image and the denoised image; and adjusting the first detail difference image by using the target detail intensity distribution to obtain a second detail difference image, the second detail difference image comprising an adjusted detail difference between the noise image and the denoised image. The image details in the to-be-output image can be enhanced by using the second detail difference image to obtain an image with adjusted details. The image details in the image with adjusted details are enhanced, and the image with adjusted details is a denoised image. By using the present solution, the image can be denoised while the definition of the image details in the image with adjusted details is improved.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method for adjusting image details and an electronic device. Background Technology

[0002] To improve the image quality captured by a mobile phone camera, the camera performs image processing operations such as correction, noise removal (or denoising), and compression on the raw image (or RAW image) to generate the final output image. The output image after image processing is of higher quality and can be displayed by the mobile phone camera.

[0003] However, because denoising in image processing requires smoothing or blurring image details, while this can eliminate noise in the raw image, it also leads to the loss of some image details. In other words, some image details are lost in the output image generated after denoising and other image processing operations on the raw image. This results in a blurry and less realistic output image. Summary of the Invention

[0004] This application provides an image detail adjustment method and electronic device, which can improve the clarity of image details in the adjusted image while achieving image denoising, thereby improving the realism of the adjusted image.

[0005] To achieve the above objectives, the embodiments of this application adopt the following technical solutions:

[0006] In a first aspect, an image detail adjustment method is provided, comprising: first receiving and responding to a first operation, acquiring a raw image, and performing image processing on the raw image to obtain a noisy image, a denoised image, and an image to be output; then determining a target detail intensity distribution map; then obtaining a first detail difference map based on the noisy image and the denoised image; adjusting the first detail difference map using the target detail intensity distribution map to determine a second detail difference map; and finally obtaining a detail-adjusted image based on the denoised image and the second detail difference map.

[0007] The output image is the denoised image. The target detail intensity distribution map is used to indicate the adjustment intensity of image details. The first detail difference map includes the detail differences between the noisy image and the denoised image.

[0008] Understandably, the electronic device performs image processing on the raw image to obtain a noisy image, a denoised image, and an output image. Then, the electronic device determines a target detail intensity distribution map; based on the adjustment intensity of image details indicated by the target detail intensity distribution map, it adjusts a first detail difference map between the noisy image and the denoised image to obtain a second detail difference map. The second detail difference map includes the adjusted detail difference map between the noisy image and the denoised image, and can be used to adjust image details in the output image, for example, to enhance image details in the output image. The electronic device uses the second detail difference map to enhance image details in the output image, obtaining a detail-adjusted image where image details are enhanced. Furthermore, since the output image is a denoised image, and the detail-adjusted image is obtained by adjusting the image details of the output image, it can also be a denoised image, just like the output image. This scheme improves the clarity of image details in the detail-adjusted image while achieving image denoising.

[0009] In conjunction with the first aspect, in one possible implementation, obtaining the second detail difference map based on the noisy image and the denoised image includes: performing a subtraction operation on the noisy image and the denoised image to obtain a third detail difference map; and extracting Y-channel information from the third detail difference map to obtain a first detail difference map of the Y channel.

[0010] The third detail difference map and the denoised image can have the same image format. The denoised image is a three-channel image, and the third detail difference map is also a three-channel image.

[0011] Understandably, the electronic device uses the noisy image and the denoised image to obtain the first detail difference map of the Y channel, thus avoiding the meshing problem present in the raw image. The electronic device then adjusts the first detail difference map based on the target detail intensity distribution map to obtain the second detail difference map. The second detail difference map also does not have the meshing problem associated with the raw image. The second detail difference map is used to adjust the image details in the output image to obtain the detail-adjusted image. The second detail difference map also does not introduce meshing problems into the detail-adjusted image; that is, the detail-adjusted image also does not have the meshing problem present in the raw image.

[0012] Secondly, since the Y-channel information contains brightness information in the image, and brightness information affects the clarity of image details, the phone can extract a first detail difference map containing Y-channel information from the third detail difference map of the three channels. This first detail difference map can then be adjusted to obtain a second detail difference map. This second detail difference map includes the adjusted Y-channel information and, when applied to the image to be output, can adjust the image details within the output image.

[0013] In conjunction with the first aspect, another possible implementation involves having the same dimensions for the first detail difference map, the target detail intensity distribution map, and the denoised image; the target detail intensity distribution map includes the target intensity value corresponding to each pixel. Specifically, the process of adjusting the first detail difference map using the target detail intensity distribution map to obtain the second detail difference map includes multiplying the target intensity value in the target detail intensity distribution map with the Y-channel information in the first detail difference map to obtain the second detail difference map.

[0014] Understandably, the first detail difference map, the target detail intensity distribution map, and the denoised image are all the same size. The target detail intensity distribution map can be used to adjust the detail differences represented by each pixel in the first detail difference map to obtain the second detail difference map. This second detail difference map is the same size as the image to be output and can be used to adjust each pixel in the image to be output, thus achieving the adjustment of image details in the image to be output.

[0015] In conjunction with the first aspect, another possible implementation involves obtaining the detail-adjusted image based on the image to be output and the second detail difference map, including: extracting Y-channel and UV-channel information from the image to be output to obtain a Y-channel image and a UV-channel image; performing an addition operation on the second detail difference map and the Y-channel image to obtain an adjusted Y-channel image; and fusing and format-converting the UV-channel image and the adjusted Y-channel image to obtain the detail-adjusted image.

[0016] Understandably, the second detail difference map includes adjusted Y-channel information. Therefore, the electronic device applies this second detail difference map to the Y-channel image of the image to be output, obtaining an adjusted Y-channel image. Then, the adjusted Y-channel image and the UV-channel images of the image to be output are re-fused to obtain the image with adjusted details. Since Y-channel information represents the brightness information in an image, and the brightness information affects the clarity of image details, the adjusted Y-channel image can adjust image details in the image with adjusted details, for example, enhancing image details in the image with adjusted details.

[0017] In conjunction with the first aspect, in another possible implementation, the noisy image is obtained by performing a first image processing on the raw image, the denoised image is obtained by performing a second image processing on the raw image, and the output image is obtained by performing a third image processing on the denoised image. The first image processing includes depigmentation; the second image processing includes both depigmentation and denoising; and the third image processing includes at least one of brightness adjustment, color adjustment, and sharpening.

[0018] It is understandable that the noisy image, the denoised image, and the output image are all obtained by image processing of the raw image. Therefore, neither the noisy image nor the denoised image has the meshing problem of the raw image. Secondly, the output image is obtained by performing a third image processing on the denoised image; therefore, the output image is also a denoised image.

[0019] In conjunction with the first aspect, another possible implementation of the above adjustment method further includes: saving the image with detailed adjustments in a first storage location; the first storage location includes the storage location corresponding to the gallery app. The above adjustment method may also include: saving the noisy image, the denoised image, and the image to be output in a second storage location different from the first storage location.

[0020] Understandably, since the image with detailed adjustments is the output image that the user can view, the electronic device can store this adjusted image in the corresponding storage location of the gallery app so that the user can view it. Secondly, since the noisy image, the denoised image, and the output image are all unprocessed images and do not require user viewing, the phone can use different storage locations to save the denoised image and the output image.

[0021] In conjunction with the first aspect, another possible implementation of the adjustment method further includes: displaying a first interface, the first interface including the image to be output; and receiving a second operation in the first interface. The above-described method of determining the target detail intensity distribution map based on the denoised image includes: in response to the second operation, determining the target detail intensity distribution map based on the denoised image. The adjustment method further includes: displaying the detail-adjusted image on the first interface.

[0022] Understandably, electronic devices can determine the target detail intensity distribution map only when the user needs to adjust the image details in the output image. The specific process may include: the electronic device first displays a first interface including the image to be output; on the first interface, it receives and responds to a second operation input by the user to determine the target detail intensity distribution map for adjusting the image details in the output image. By displaying the image to be output on the first interface first, followed by the image with adjusted details, the user can visually see that the image details in the adjusted image are clearer.

[0023] In conjunction with the first aspect, another possible implementation method, in which the above-mentioned determination of the target detail intensity distribution map based on the denoised image includes: performing image analysis on the denoised image to obtain the analyzed detail intensity distribution map; and determining the target detail intensity distribution map based on the analyzed detail intensity distribution map.

[0024] Understandably, electronic devices can perform image analysis on denoised images to determine an analyzed detail intensity distribution map adapted to the denoised image. Furthermore, based on the analyzed detail intensity distribution map adapted to the denoised image, the electronic device can determine a target detail intensity distribution map used to indicate the adjustment intensity (or detail intensity) of image details.

[0025] In conjunction with the first aspect, another possible implementation involves performing image analysis on the denoised image to obtain an analyzed detail intensity distribution map, including: analyzing the texture intensity of the denoised image to obtain texture intensity information, and determining a first detail intensity distribution map based on the texture intensity information; performing brightness analysis on the denoised image to obtain brightness information, and determining a second detail intensity distribution map based on the brightness information. The first intensity value in the first detail intensity distribution map corresponds to the texture intensity value in the texture intensity information. The second intensity value in the second detail intensity distribution map corresponds to the brightness value in the brightness information. In this case, the analyzed detail intensity distribution map includes both the first and second detail intensity distribution maps.

[0026] Optionally, the above-described image analysis of the denoised image to obtain the analyzed detail intensity distribution map may further include: performing semantic segmentation on the denoised image to obtain a segmented image labeled with semantic categories, and determining a third detail intensity distribution map based on the segmented image; the third intensity value in the third detail intensity distribution map corresponds to the semantic category in the segmented image. In this case, the analyzed detail intensity distribution map also includes a third detail intensity distribution map.

[0027] It is understandable that the brightness, texture intensity, and semantic category of the denoised image all affect the level of detail intensity that matches the denoised image. Furthermore, since the denoised image and the output image are obtained by processing the same raw image, the level of detail intensity that matches the denoised image also matches the output image. The target detail intensity distribution map determined by this scheme based on the brightness, texture intensity, and semantic category of the denoised image shows a high degree of compatibility with the amount of image detail in both the denoised and output images, thus accurately improving the clarity of image details in the output image.

[0028] In conjunction with the first aspect, in another possible implementation, the first detail intensity distribution map includes a first intensity value that is positively correlated with the texture intensity value; the second detail intensity distribution map includes a second intensity value that is negatively correlated with the brightness value.

[0029] Specifically, determining the target detail intensity distribution map based on the analyzed detail intensity distribution map includes: multiplying the first intensity value in the first detail intensity distribution map, the second intensity value in the second detail intensity distribution map, and the third intensity value in the third detail intensity distribution map to obtain the target detail intensity distribution map.

[0030] Optionally, if the analyzed detail intensity distribution map also includes a third detail intensity distribution map, the above-mentioned determination of the target detail intensity distribution map based on the analyzed detail intensity distribution map may include: multiplying the first intensity value in the first detail intensity distribution map, the second intensity value in the second detail intensity distribution map, and the third intensity value in the third detail intensity distribution map to obtain the target detail intensity distribution map.

[0031] It is understandable that regions with higher texture intensity in a denoised image contain more image details. Applying a higher detail intensity to these regions can amplify the image details, making them richer and clearer. Therefore, the first detail intensity distribution map, which includes first intensity values ​​positively correlated with texture intensity, can amplify the image details in regions of the denoised image that contain more image details. Furthermore, since the output image and the denoised image are obtained by image processing the same raw image, the first detail intensity distribution map, which can amplify the image details in regions of the denoised image that contain more image details, can also amplify the image details in regions of the output image that contain more image details.

[0032] Secondly, in a denoised image, the lower the brightness of the region, the less obvious the image details. Applying a higher detail intensity to these lower-brightness regions can amplify the image details in that region. Therefore, the second detail intensity distribution map, which includes a second intensity value negatively correlated with the brightness value, can amplify the image details in regions of the denoised image where details are not obvious. Furthermore, since the output image and the denoised image are obtained by image processing the same raw image, the first detail intensity distribution map, which can amplify the image details in regions of the denoised image where details are not obvious, can also amplify the image details in regions of the output image where details are not obvious.

[0033] Furthermore, the dimensions of the first detail intensity distribution map, the second detail intensity distribution map, the third detail intensity distribution map, and the image to be output can all be the same. The first, second, and third detail intensity distribution maps can all be used to adjust each pixel in the image to be output. Specifically, the electronic device can multiply the first, second, and third intensity values ​​of the same pixel in the first, second, and third detail intensity distribution maps to obtain the target intensity value for each pixel. The target intensity value for each pixel in the target detail intensity distribution map is determined based on the brightness, texture intensity, and semantic category of the denoised image, resulting in a higher degree of adaptation to the amount of image detail in the denoised image and the image to be output, thus accurately improving the clarity of image details in the image to be output.

[0034] In conjunction with the first aspect, in another possible implementation, the above response to the second operation, determining the target detail intensity distribution map based on the denoised image, includes: determining the specified intensity value selected by the second operation; performing image analysis on the denoised image to obtain the analyzed detail intensity distribution map; and determining the target detail intensity distribution map based on the specified intensity value and the analyzed detail intensity distribution map.

[0035] The determination of the target detail intensity distribution map based on the specified intensity value and the analyzed detail intensity distribution map may include: if the analyzed detail intensity distribution map includes a first detail intensity distribution map and a second detail intensity distribution map, multiplying the first intensity value in the first detail intensity distribution map and the second intensity value in the second detail intensity distribution map to obtain a fourth detail intensity distribution map; or, if the analyzed detail intensity distribution map includes a first detail intensity distribution map, a second detail intensity distribution map, and a third detail intensity distribution map, multiplying the first intensity value in the first detail intensity distribution map, the second intensity value in the second detail intensity distribution map, and the third intensity value in the third detail intensity distribution map to obtain a fourth detail intensity distribution map; and then multiplying the specified intensity value and the fourth intensity value in the fourth detail intensity distribution map to obtain the target detail intensity distribution map.

[0036] Understandably, based on the output image on the first interface, the user determines whether to adjust image details. If so, they can input a second operation to manually select a specified intensity value to adjust the image details in the output image. Secondly, the electronic device can perform image analysis on the denoised image to determine an analyzed detail intensity distribution map that fits the denoised image. This analyzed detail intensity distribution map also fits the output image. Furthermore, based on the user-specified intensity value and the analyzed detail intensity distribution map that fits the output image, the electronic device can determine a target detail intensity distribution map to indicate the adjustment intensity (i.e., detail intensity) for the image details.

[0037] In a second aspect, an electronic device is provided, comprising: a processor, a memory, and a communication interface. The memory and the communication interface are coupled to the processor. The memory stores computer program code, including computer instructions. When the processor executes the computer instructions, it causes the electronic device to perform the image detail adjustment method as described in any one of the first aspects above.

[0038] Thirdly, a computer-readable storage medium is provided that stores computer instructions. When the computer instructions are executed on an electronic device, the electronic device causes the electronic device to perform the image detail adjustment method as described in any one of the first aspects above.

[0039] Fourthly, a computer program product containing computer instructions is provided, which, when executed on an electronic device, causes the electronic device to perform an image detail adjustment method as described in any one of the first aspects above.

[0040] Fifthly, an apparatus (e.g., a chip system) is provided, comprising a processor for supporting an electronic device in implementing the image detail adjustment method described in the first aspect. In one possible design, the apparatus further comprises a memory for storing necessary program instructions and data of the electronic device. When the apparatus is a chip system, it may be composed of chips or may include chips and other discrete components.

[0041] The technical effects of any of the design methods in aspects two through five can be found in the technical effects of different design methods in aspect one, and will not be repeated here. Attached Figure Description

[0042] Figure 1 A schematic diagram of an image to be output and an image after detail adjustment, provided for an embodiment of this application;

[0043] Figure 2This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application;

[0044] Figure 3 A system architecture diagram of the electronic device provided in the embodiments of this application;

[0045] Figure 4 A flowchart illustrating an image detail adjustment method provided in this application embodiment. Figure 1 ;

[0046] Figure 5 Images (a) to (c) in this application are schematic diagrams of an interface for capturing images according to an embodiment of this application.

[0047] Figure 6 A flowchart illustrating an image detail adjustment method provided in this application embodiment. Figure 2 ;

[0048] Figure 7 This is a schematic diagram illustrating the training of a first neural network model provided in an embodiment of this application;

[0049] Figure 8 (a) to (d) in this application are schematic diagrams illustrating an image detail adjustment operation provided in an embodiment of this application;

[0050] Figure 9 A schematic diagram illustrating a first mapping relationship provided in an embodiment of this application;

[0051] Figure 10 A schematic diagram illustrating a second mapping relationship provided in an embodiment of this application;

[0052] Figure 11 A schematic diagram of the input and output of a second neural network model provided in an embodiment of this application;

[0053] Figure 12 A flowchart illustrating an image detail adjustment method provided in this application embodiment. Figure 3 ;

[0054] Figure 13 A flowchart illustrating an image detail adjustment method provided in this application embodiment. Figure 4 ;

[0055] Figure 14 This is a schematic diagram of a chip system provided in an embodiment of this application. Detailed Implementation

[0056] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that “ / ” means “or,” for example, A / B can mean A or B; “and / or” in the text is merely a description of the relationship between related objects, indicating that three relationships can exist, for example, A and / or B can mean: A alone, A and B simultaneously, and B alone.

[0057] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0058] The terms "first" and "second" in the following embodiments of this application are for descriptive purposes only and should not be construed as implying relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0059] To facilitate the explanation of the technical solution of this application, some concepts involved in this application will be explained first below.

[0060] (1) Camera: It can be divided into two parts: hardware and software. The hardware part of the camera can include the camera and the image processing chip, etc. The software part of the camera can include image processing algorithms, such as deep learning models (e.g., Convolutional Neural Network (CNN)). The image processing chip can be used to run image processing algorithms.

[0061] (2) Raw image: The image obtained after the light-sensing element (e.g., image sensor) in the camera converts the captured light source signal into an electrical signal can be called the raw image.

[0062] (3) Output Image: This refers to the final output image obtained by processing the raw image output by the camera. The format of the output image can be JPEG (joint photographic experts group) format or RGB (Red Green Blue) format, etc. The mobile phone can save the output image so that the user can view it on the mobile phone. For example, the output image can be a photographed image or any frame from a video.

[0063] For example, an electronic device can receive and respond to an operation to initiate a camera function, acquire a raw image, and process the raw image to obtain a photographed image, which is the image to be output. The electronic device can then send the photographed image to a gallery application (APP) within the electronic device, and the gallery application saves the photographed image.

[0064] For example, an electronic device can receive and respond to an operation to initiate a video recording function, continuously acquire raw images, and process each raw image to obtain a frame of an image in a video; each frame of the video is an image to be output. The electronic device can also send the video to a gallery application (APP) on the electronic device, and the gallery APP saves the video.

[0065] (4) Raw image processing: The process of processing the raw image to obtain the output image in RGB format. Optionally, the raw image processing can be called RAW-to-RGB conversion. In some embodiments, a deep learning model can be used to implement RAW-to-RGB conversion.

[0066] In some embodiments, RAW-to-RGB conversion may include demosaicing, noise reduction, and compression. Since each pixel in the raw image contains only one color information, an interpolation algorithm is needed to estimate the missing color information in the raw image, generating an RGB image where each pixel includes RGB color information. This process is called demosaicing. Compression refers to compressing the image using JPEG or other image formats for storage and transmission.

[0067] Currently, although mobile phone cameras have improved image quality through software algorithms and hardware improvements, the images captured by mobile phone cameras still have the following problems due to limitations in the hardware performance of mobile phone cameras.

[0068] Firstly, the image processing chip in a mobile phone camera has a limited area. To accommodate this limitation, deep learning models used for denoising are often designed to be small, resulting in limited processing power. These models may not be able to accurately distinguish between real image details and noise in the raw image. Therefore, during denoising, they may treat some real image details as noise and remove them (e.g., smoothing or blurring), causing certain areas in the denoised image to become blurry—the so-called "smearing problem." Furthermore, the smoothed or blurred image details cannot be recovered during denoising. When the mobile phone camera generates an output image based on the denoised image, the image details in that output image are also lost, resulting in a blurry output image. For example, as... Figure 1 As shown, some image details in the output image generated by the mobile phone camera are blurry.

[0069] Secondly, in low-to-medium brightness environments, the image sensor in a mobile phone camera captures relatively less signal (or useful light information), while the noise generated by the physical characteristics of the image sensor itself is relatively high. This results in a lower signal-to-noise ratio (SNR) for the raw image captured by the mobile phone camera. SNR is the ratio of signal strength to noise strength. A higher SNR means a clearer image with richer details; a lower SNR indicates more random noise in the image, which may lead to blurriness or graininess. Therefore, it can be seen that the low SNR of the raw image captured by the mobile phone camera in low-to-medium brightness environments further contributes to the blurriness of the output image generated by the mobile phone camera.

[0070] In summary, existing technologies for mobile phone cameras produce images with blurred details and poor realism.

[0071] To address the aforementioned issues, this application provides a method for adjusting image details. An electronic device obtains a first detail difference map of the Y channel using a noisy image and a denoised image, avoiding the meshing problem present in the raw image. Furthermore, the brightness, texture intensity, and semantic category of the denoised image all affect the level of detail intensity adapted to the denoised image. Since the output image and the denoised image are obtained by image processing the same raw image, the brightness, texture intensity, and semantic category of the denoised image also affect the level of detail intensity adapted to the output image. Therefore, the target intensity value determined based on the brightness, texture intensity, and semantic category of the denoised image in this solution has a higher degree of compatibility with the amount of image detail in the output image, and can accurately improve the clarity of image details in the output image.

[0072] For example, such as Figure 1 As shown, the image detail adjustment method provided in this application produces a more detailed and clearer image. Compared to the original image, the detail-adjusted image is significantly sharper.

[0073] This application provides a method for adjusting image details, which can be applied to electronic devices, such as mobile phones, tablets, personal computers (PCs), laptops, in-vehicle devices, or wearable devices (e.g., smart bracelets). This application does not limit the specific type of electronic device.

[0074] For example, taking a mobile phone as an example, the hardware structure of an electronic device will be introduced. Figure 2 As shown, a mobile phone may include a processor 110, an external memory interface 120, an internal memory 121, a universal serial bus (USB) interface 130, a charging management module 140, a power management module 141, a battery 142, antenna 1, antenna 2, a mobile communication module 150, a wireless communication module 160, an audio module 170, a speaker 170A, a receiver 170B, a microphone 170C, a headphone jack 170D, a sensor module 180, buttons 190, a motor 191, an indicator 192, one or more displays 193, a subscriber identification module (SIM) card interface 194, and a camera 195, etc. The sensor module 180 may include a pressure sensor, a touch sensor, a fingerprint sensor, a temperature sensor, an ambient light sensor, etc.

[0075] Processor 110 may include one or more processing units, such as: application processor (AP), modem processor, graphics processing unit (GPU), image signal processor (ISP), controller, memory, video codec, digital signal processor (DSP), baseband processor, and / or neural network processing unit (NPU), etc. Different processing units may be independent devices or integrated into one or more processors.

[0076] The processor 110 may also include a memory for storing instructions and data. In some embodiments, the memory in the processor 110 is a cache memory. This memory can store instructions or data that the processor 110 has just used or that are used repeatedly. If the processor 110 needs to use the instruction or data again, it can retrieve it directly from the memory. This avoids repeated accesses, reduces the waiting time of the processor 110, and thus improves the efficiency of the system.

[0077] The external memory interface 120 can be used to connect to an external non-volatile memory, thereby expanding the phone's storage capacity. The external non-volatile memory communicates with the processor 110 through the external memory interface 120 to perform data storage functions. For example, files such as images to be output can be saved in the external non-volatile memory.

[0078] The internal memory 121 may include one or more random access memory (RAM) and one or more non-volatile memory (NVM). The RAM can be directly read and written by the processor 110 and can be used to store executable programs (e.g., machine instructions) of the operating system or other running programs, as well as user and application data. The NVM can also store executable programs and user and application data, and can be pre-loaded into the RAM for direct read and write by the processor 110. In this embodiment, the internal memory 121 may store noisy images, denoised images, and images to be output captured by the mobile phone camera.

[0079] A touch sensor, also known as a "touch device," can be located on the display screen 193. A touchscreen consisting of the touch sensor and the display screen 193 is also called a "touchscreen." The touch sensor detects touch operations applied to or near it. The touch sensor can then transmit the detected touch operation to the application processor to determine the type of touch event. Visual output related to the touch operation can be provided through the display screen 193. In some embodiments, the touch sensor may also be located on the surface of the phone, in a different position than the display screen 193.

[0080] In this embodiment, the touch sensor can detect user input operations such as shooting, image viewing, and detail adjustment, and transmit the information of these operations to the processor 110 in real time. The processor 110 analyzes the function executed corresponding to the operation, such as controlling the camera to capture an image in response to the shooting operation (including acquiring a raw image and performing image processing on the raw image to obtain the output image), displaying the output image selected in the image viewing operation in response to the image viewing operation, and determining the specified intensity value selected in the detail adjustment operation in response to the detail adjustment operation.

[0081] Mobile phones can achieve display functions through GPU, display screen 193, and AP. Mobile phones can achieve shooting functions through camera 195, ISP, video codec, GPU, display screen 193, and AP.

[0082] In some embodiments, the mobile phone can capture light through camera 195 and transmit the light signal to the image sensor in camera 195. The light signal is converted into a raw image by the image sensor in camera 195. Then, the mobile phone performs image processing on the raw image output by camera 195 through ISP to obtain the image to be output. The mobile phone can display the image to be output on display screen 193 and can also save the image to be output.

[0083] The display screen 193 includes a display panel. The display panel may be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode (FLED), a minimized display, a microLED, a micro-OLED, a quantum dot light-emitting diode (QLED), etc. In some embodiments, the mobile phone may include one or N displays 193, where N is a positive integer greater than 1.

[0084] Of course, it is understandable that the above... Figure 2 The illustration shown is merely an example when the electronic device is in the form of a mobile phone. If the electronic device is in the form of a tablet, PC, or wearable device, the structure of the electronic device may include more... Figure 2 The fewer structures shown can also include more than Figure 2 The structures shown are not limited here.

[0085] It is understandable that the implementation of electronic device functions requires not only hardware support but also the cooperation of software architecture. In some examples, refer to... Figure 3 As shown, the operating system used by the electronic device is Android. TM Taking a system as an example, the software architecture of the electronic device provided in this application may include: an application layer, an application framework layer, and a kernel layer. It should be noted that the embodiments in this application use Android... TM To illustrate, let's take an example from other operating systems (such as HarmonyOS). TM System, iOS TM Any system (or similar system) can achieve the solution of this application as long as the functions implemented by each functional module are similar to those in the embodiments of this application.

[0086] The application layer can include a series of apps. These apps can be divided into system apps pre-installed on the phone and third-party apps downloaded by the user. For example, ... Figure 3 As shown, the system apps may include camera apps, gallery apps, email apps, music apps, and Bluetooth apps, etc.; third-party apps may include chat apps, shooting apps, etc., and this application embodiment does not impose any restrictions on them.

[0087] The application framework layer provides application programming interfaces (APIs) and programming frameworks for applications in the application layer. The application framework layer includes predefined functions such as a camera access interface, a surface flinger, an activity manager, a window manager, a view system, a resource manager, a notification manager, an audio service, and a camera service, etc. This application embodiment does not impose any limitations on these.

[0088] The camera access interface is used to provide application programming interfaces and programming frameworks for apps. For example, a camera app can use the camera access interface to call the camera driver in the kernel layer to control the camera 195.

[0089] A view system includes visual controls, such as controls for displaying text and controls for displaying images. View systems can be used to construct the display interface of an app. The display interface can consist of one or more views. For example, a preview interface including a preview image, a recording interface including any frame of image, and a first interface including a denoised image, etc.

[0090] Surface flinger is an Android app. TMThis is a system-level service responsible for displaying all 2D and 3D graphics elements in the system. It manages all surfaces on the display screen. The surface flinger is responsible for compositing all surfaces and outputting the final image to the display screen 193.

[0091] The kernel layer is the layer between hardware and software. It interacts directly with the hardware and provides a variety of basic system services and functions. The kernel layer shields upper-layer software from hardware details, ensuring system stability and compatibility. The kernel layer contains drivers for various hardware devices, such as display drivers for the screen 193, camera drivers for the camera 195, audio drivers, sensor drivers, etc., allowing upper-layer software to operate without needing to know the specific implementation details of the hardware.

[0092] In some embodiments, any app on the mobile phone (e.g., a camera app) can receive and respond to operations that trigger a shooting function (e.g., triggering an image preview function, or triggering a photo capture function), calling the camera driver in the kernel layer through the camera access interface, and controlling the camera 195 to start through the camera driver. After the camera 195 starts, it captures light and transmits the light signal to the image sensor in the camera 195. The light signal is converted into a raw image by the image sensor. The app can also start the camera service in the framework layer by calling a preset API. After the camera service starts, it can trigger the image processing chip to perform image processing on the raw image output by the camera 195 to obtain the image to be output (e.g., a captured image, or any frame image from a video). The image processing chip sends the image to be output to the app. The app can call the display driver in the kernel layer to control the display to show the image to be output. The app can also save the image to be output.

[0093] Based on the above hardware and software architecture, the following uses a mobile phone as an example to introduce the method for adjusting image details provided in the embodiments of this application.

[0094] Reference Figure 4 The diagram shown is a flowchart illustrating an image detail adjustment method provided in an embodiment of this application. Figure 4 As shown, the specific process may include S401-S413.

[0095] S401: The mobile phone receives and responds to the shooting operation that triggers the shooting function, and controls the camera to capture a raw image.

[0096] The mobile phone receives and responds to the user's input shooting operation, controlling the camera to capture raw images. The shooting function can be an image preview function, a photo capture function, or a video recording function. When the shooting operation triggers the image preview or video recording function, the mobile phone camera continuously captures multiple frames of images. Continuously capturing multiple frames of images can include: the mobile phone camera continuously capturing raw images and performing image processing on each captured raw image to obtain one frame from the multiple images. Secondly, when the shooting operation triggers the photo capture function, the mobile phone camera captures a single frame of image. Capturing a single frame of image can include the mobile phone camera capturing a raw image and performing image processing on that raw image to obtain a single frame of image.

[0097] It should be noted that, in this embodiment of the application, the example is to capture a raw image with a mobile phone camera and perform image processing on the raw image to obtain a frame image.

[0098] Optionally, the shooting operation can be a voice command, a pressing operation of one or more physical buttons, or a touch operation.

[0099] For example, taking the shooting operation as a touch operation as an example, the process of the mobile phone receiving the shooting operation is introduced, such as... Figure 5 As shown in (a), the mobile phone may include a camera app. When the camera app icon 510 is displayed, the phone can receive a user's click operation (or first click operation) on the camera app icon 510. The first click operation is used to trigger the image preview function. Figure 5 As shown in (b), the mobile phone responds to the first click operation and displays a preview interface 520. The preview interface 520 includes preview images of each frame captured in real time.

[0100] Then, as Figure 5 As shown in (b), when the mobile phone displays a preview interface 520, and the preview interface 520 includes a camera option 521, it can receive a user's click operation on the camera option 521. This user's click operation on the camera option 521 can be referred to as a second click operation. The second click operation is a shooting operation, and it is used to trigger the camera function. In response to the second click operation, the mobile phone acquires a raw image and performs image processing on the raw image to obtain the captured image.

[0101] Furthermore, such as Figure 5As shown in (c), when the mobile phone displays the preview interface 520, and the preview interface 520 includes a video recording option 522, it can receive a user's click operation on the video recording option 522. This user's click operation on the video recording option 522 can be referred to as a third click operation. The third click operation is also a shooting operation, and it is used to trigger the video recording function. In response to the third click operation, the mobile phone continuously captures raw images and performs image processing on each captured raw image to obtain a video frame.

[0102] S402. The mobile phone performs image processing on the raw image to obtain a noisy image, a denoised image, and an image to be output, and saves the noisy image, the denoised image, and the image to be output.

[0103] A mobile phone camera can perform image processing on a raw image, sequentially obtaining a noisy image, a denoised image, and the output image. The noisy image, denoised image, and output image can all be images in a preset image format (e.g., RGB format). This image processing procedure includes not only denoising but also depigmentation, brightness adjustment, color adjustment, and sharpening. The noisy image and denoised image can be obtained during the image processing of the raw image.

[0104] For example, such as Figure 6 As shown, the mobile phone responds to the shooting operation and obtains a noisy image, a denoised image, and an image to be output.

[0105] In this embodiment of the application, the image to be output is obtained by image processing of the raw image. Furthermore, the mobile phone also needs to adjust the image details of the image to be output to obtain the image after detail adjustment.

[0106] In one implementation, the mobile phone can use the image to be output as the final image output by the mobile phone camera (e.g., a photographed image or a frame from a video). The mobile phone then responds to the user's conditional operation on the output image, adjusts the image details of the output image to obtain the adjusted image, and then updates the adjusted image as the photographed image.

[0107] For example, taking the example of a mobile phone capturing an image to be output as the final image output by the phone's camera, the mobile phone captures the image to be output in response to a shooting operation that triggers the photo-taking function; the image to be output is the captured image. Alternatively, the mobile phone captures the image to be output in response to a shooting operation that triggers the video recording function; the image to be output is a frame from a captured video.

[0108] In another implementation, the mobile phone can obtain the image to be output, and then adjust the image details of the image to obtain the image with detailed adjustment. Only the image with detailed adjustment is used as the final image output by the mobile phone camera.

[0109] For example, if a mobile phone uses only the image with adjusted details as the final output image from its camera, the phone acquires an image to be output in response to a shooting operation that triggers the photo-taking function; then, it adjusts the image details of this image to obtain an image with adjusted details, which is the captured image. Alternatively, the phone acquires an image to be output in response to a shooting operation that triggers the video-taking function; then, it adjusts the image details of this image to obtain an image with adjusted details, which is a frame from the captured video.

[0110] It should be noted that in the embodiments described in S403-S413 below, the method of adjusting image details is introduced by taking the image to be output as the final image output by the mobile phone camera.

[0111] In this embodiment, the mobile phone camera captures multiple images during the shooting process, including a noisy image, a denoised image, and an image to be output. The noisy image and the denoised image are images that have not undergone image processing and do not require user viewing. The image to be output has undergone image processing, but the image details have not been adjusted; the user may or may not view it. For example, if the mobile phone uses the image to be output as the final output image, then the image to be output is the one the user needs to view; if the mobile phone only uses the image with adjusted details as the final output image, then the image to be output is the one the user does not need to view. The mobile phone can use different storage methods (e.g., different storage locations or different storage formats) to save the images that do not require user viewing and the images that require user viewing.

[0112] For example, taking an image that doesn't need to be viewed by the user, including both a noisy image and a denoised image, and an image that the user needs to view, including the image to be output, the phone can save the image to be output in a first storage location, and save the noisy image and the denoised image in a second storage location different from the first storage location. The first storage location may include the storage location corresponding to the phone's built-in gallery app. Then, the user can view the image to be output in the gallery app.

[0113] It should be noted that if the phone determines that the image to be output is also an image that does not need to be viewed by the user, the phone can save the image to be output in the second storage location.

[0114] In some embodiments, after the mobile phone obtains the image to be output, it can also momentarily display the image on the screen. Momentarily displaying the image on the screen can mean that the mobile phone displays the image to be output after the user inputs a shooting operation, and then immediately stops displaying the image.

[0115] For example, such as Figure 5As shown in (b), in response to the second click operation used to trigger the camera function, the mobile phone instantly displays the captured image on the preview screen 520, and immediately displays a new preview image on the preview screen 520.

[0116] For example, such as Figure 5 As shown in (c), in response to the third click operation used to trigger the recording function, the mobile phone instantly displays a video frame on the preview interface 520, and then immediately displays the next video frame.

[0117] It should be noted that, Figure 5 (b) and Figure 5 The image shown in (c) can be different. Figure 5 The changes in the image are not shown in (b) and (c).

[0118] In some embodiments, S402 may include: the mobile phone performing a first image processing on the raw image to obtain a noisy image; performing a second image processing on the raw image to obtain a denoised image; and then performing a third image processing on the denoised image to obtain a T-shirt to be output. The first image processing includes depigmentation. The second image processing includes depigmentation and denoising. The third image processing includes at least one of brightness adjustment, color adjustment, and sharpening.

[0119] For example, the mobile phone can perform a first image processing on the raw image to obtain a noisy image through the image front end (IFE) module in the ISP, and then perform a second image processing on the raw image to obtain a denoised image. The mobile phone then performs a third image processing on the denoised image through the image processing engine (IPE) module in the ISP to obtain the image to be output.

[0120] Optionally, the mobile phone can input the raw image into the first neural network model for de-mosaicing and denoising to obtain a noisy image and a denoised image output by the first neural network model. For example, the first neural network model can be a model using a Convolutional Neural Network (CNN), which may include U-Net.

[0121] Furthermore, prior to S402, the mobile phone can train an initial neural network model to obtain the first neural network model. The training process may include: firstly acquiring multiple output samples, each output sample including: multiple denoised sample images, and a corresponding sample noise image for each denoised sample image; then performing the inverse process of RAW-to-RGB conversion on each denoised sample image to obtain a sample raw image, which is the input sample corresponding to each denoised sample image; then, using multiple input samples and multiple output samples to train the initial neural network model to obtain the first neural network model.

[0122] The reverse process of RAW-to-RGB conversion can refer to converting an RGB format image into a raw image.

[0123] Optionally, the mobile phone can acquire multiple denoised sample images captured by the mobile phone camera; then Gaussian Poisson noise is added to each denoised sample image to obtain the sample noise image corresponding to the denoised sample image.

[0124] For example, the process of training the initial neural network model on a mobile phone is described using the initial neural network model of the U-Net architecture as an example. Figure 7 As shown, the U-Net network used in the initial neural network model can include two convolutional layers, each with a configuration parameter of 32*3*3*3. 32*3*3*3 means each convolutional layer includes 32 3*3*3 filters. A 3*3*3 filter means the filter's height and width are both 3, and the number of input channels is also 3. Training the initial neural network model using each input sample and its corresponding output sample can include: inputting each input sample (i.e., the raw image) into the initial neural network model, performing image processing to obtain the actual denoised image and the actual noisy image output by the initial neural network model; then, using a preset loss function, calculating the first difference between the actual denoised image and the denoised image corresponding to the input sample, and calculating the second difference between the actual noisy image and the noisy image corresponding to the input sample; and finally, updating the initial neural network model based on the first and second differences. The mobile phone can iteratively execute the above training process multiple times until the stopping condition is met (such as reaching the predetermined number of training rounds or the loss function converging), and the updated initial neural network model is determined as the first neural network model.

[0125] S403. The mobile phone receives and responds to the triggering of an image viewing operation to display the image to be output, and displays a first interface, which includes the image to be output.

[0126] When the phone captures the image to be output as the final output image, and updates the captured image with the image after detail adjustments, after the phone obtains and saves the image to be output, it can accept user input to view the image and display a first interface including the image to be output. In addition to displaying the image to be output, the first interface can also provide editing functions, sharing functions, deletion functions, and detail adjustment functions for the image to be output. Optionally, the detail adjustment function can be one of the editing functions.

[0127] In some embodiments, the mobile phone may receive other operations before receiving the image viewing operation, such as an operation to open the gallery app where the image to be output is saved.

[0128] For example, the mobile phone saves the image to be output in the Gallery app. The mobile phone can first receive an operation to open the Gallery app, displaying the album browsing interface 610 in the Gallery app. Figure 8 As shown in (a), the photo album browsing interface 610 may include a thumbnail of each of a plurality of captured images, the plurality of captured images including the image to be output. Then, as Figure 8 As shown in (b), the mobile phone can receive and respond to a click operation on the thumbnail 611 of the image to be output, displaying an image viewing interface 620. The image viewing interface 620 is a first interface. The click operation on the thumbnail corresponding to the image to be output is an image viewing operation. The image viewing interface 620 includes the image to be output 621, and also includes editing options 622, deletion options 623, and a first detail adjustment option 624.

[0129] It should be noted that, Figure 8 The image viewing interface 620 shown in (a) includes a first detail adjustment option 624, which is merely an example. The first detail adjustment option 624 may also be included in the editing option 622. This application embodiment does not limit this.

[0130] S404 When the phone displays the first screen, it receives and responds to a detail adjustment operation for adjusting the detail intensity, and determines the specified intensity value selected in the detail adjustment operation.

[0131] The phone can display a second detail adjustment option, including multiple selectable intensity values, on a first screen, or a first prompt message to guide the user to adjust the image's detail intensity. The phone can then receive and respond to the user's input detail adjustment operation, determining the specified intensity value. For example, as... Figure 6 As shown, the phone responds to the detail adjustment operation and determines the specified intensity value selected in the detail adjustment operation.

[0132] The second detailed adjustment option provides multiple selectable intensity values ​​within a preset intensity range for the user to choose from. This second detailed adjustment option can be either a stepless adjustment option or a fixed-level adjustment option. A stepless adjustment option allows free selection within the preset intensity range. A fixed-level adjustment option allows selection between multiple fixed intensity values ​​within the preset intensity range.

[0133] Optionally, fine-tuning operations can be performed via voice commands, pressing one or more physical buttons, or touch operations.

[0134] For example, taking a touch operation as an example of fine-tuning, such as... Figure 8 As shown in (b) above, the first interface displayed on the mobile phone can be an image viewing interface 620, which may include a first detail adjustment option 624. The mobile phone can first receive and respond to the user's click operation on the first detail adjustment option 624, and then display a second detail adjustment option 625 on the image viewing interface 620. For example... Figure 8 As shown in (c), the second detail adjustment option 625 can be a fixed-level adjustment option, including multiple fixed intensity values, such as 0, 1, 2, 3, ..., 9 and 10. The mobile phone can then receive and respond to the user's selection of any intensity value in the second detail adjustment option 625, determining that any intensity value as the specified intensity value. This selection of any intensity value in the second detail adjustment option 625 is a detail adjustment operation.

[0135] S405 The mobile phone analyzes the texture intensity of the denoised image to obtain texture intensity information, and determines the first detail intensity distribution map based on the texture intensity information.

[0136] Since different regions in a denoised image can have different texture intensities, and a higher texture intensity in any region indicates more image detail, a larger detail intensity is needed to amplify the image detail in regions with higher texture intensity. Therefore, the mobile phone can determine the texture intensity information of the denoised image and then determine the corresponding first detail intensity distribution map. The first detail intensity distribution map can include: the first intensity value of each first region in the denoised image. The first intensity value of each first region in the denoised image corresponds to the texture intensity value of that region. Each first region can include one or more pixels.

[0137] In this embodiment, besides the user manually selecting a specified intensity value, the brightness of each region in the denoised image, the texture intensity of each region in the denoised image, and the type of subject in the denoised image all affect the level of detail intensity adapted to the denoised image. Specifically, the lower the brightness of a region in the denoised image, the less obvious the image details. Applying a higher detail intensity to a region with lower brightness in the denoised image can amplify the image details in that region. Regions with higher texture intensity in the denoised image contain more image details. Applying a higher detail intensity to a region with higher texture intensity in the denoised image can amplify the image details in that region, making the image details in that region richer and clearer. Some types of subjects in the denoised image (e.g., people, animals, etc.) contain more image details, while other types of subjects (e.g., flat objects) contain fewer image details. Setting different levels of detail intensity for different types of subjects can reflect the image details contained in different types of subjects.

[0138] In summary, after the mobile phone determines that the user has selected a specific intensity value, it can perform image analysis on the denoised image to determine the intensity value that is suitable for the denoised image. Optionally, the image analysis may include at least one of texture intensity analysis, brightness analysis, and semantic segmentation.

[0139] It should be noted that the following embodiments use image analysis including texture intensity analysis, brightness analysis, and semantic segmentation as examples to illustrate the process of image analysis performed by a mobile phone on a denoised image. Image analysis may also include one or two of texture intensity analysis, brightness analysis, and semantic segmentation, and this application embodiment does not limit this.

[0140] For example, such as Figure 6 As shown, if image analysis includes texture intensity analysis, brightness analysis, and semantic segmentation, the mobile phone performs texture intensity analysis, brightness analysis, and semantic segmentation on the denoised image to obtain a first detail intensity distribution map, a second detail intensity distribution map, and a third detail intensity distribution map, which may specifically include S405-S407.

[0141] For example, if image analysis includes texture intensity analysis and brightness analysis, but excludes semantic segmentation, then it is similar to... Figure 6 As shown, the phone only obtains the first and second detail intensity distribution maps, but not the third detail intensity map.

[0142] In some embodiments, S405 may include: the mobile phone using any texture analysis method to calculate the texture intensity of the denoised image to obtain texture intensity information of the denoised image; the texture intensity information may include the texture intensity values ​​of each first region in the denoised image; then, based on the first mapping relationship and the texture intensity values ​​of each first region in the denoised image, the first intensity value of each first region is determined. The first mapping relationship may refer to the mapping relationship between the texture intensity values ​​and the first intensity values.

[0143] The first detail intensity distribution map and the denoised image can have the same size. The first detail intensity distribution map includes the first intensity values ​​of multiple first regions in the denoised image. The first mapping relationship can include: the larger the texture intensity value of any first region, the larger the first intensity value of that first region can be.

[0144] For example, the first mapping relationship between texture intensity values ​​and first intensity values ​​may include: all texture intensity values ​​within the first texture intensity range correspond to the same first value X1; the larger the texture intensity value within the second texture intensity range, the larger the corresponding second value; and all texture intensity values ​​within the third texture intensity range correspond to the same third value X3. Wherein, the larger the texture intensity value within the second texture intensity range, the larger the corresponding second value, indicating a positive correlation between the texture intensity values ​​within the second texture intensity range and the second value. Texture intensity values ​​within the second texture intensity range are greater than those within the first texture intensity range, and texture intensity values ​​within the third texture intensity range are greater than those within the second texture intensity range. The second value is greater than the first value X1, and the third value X3 is greater than the second value. The first value X1, the second value, and the third value X3 all belong to the first intensity value. For example... Figure 9 As shown, the first texture intensity range can be the range from 0 to the first texture intensity value W1; the second texture intensity range can be the range from the first texture intensity value W1 to the second texture intensity value W2, and the second texture intensity range does not include the first texture intensity value W1; the third texture intensity range can be the range greater than the second texture intensity value W2.

[0145] Among them, such as Figure 9 As shown, the texture intensity values ​​within the first texture intensity range and the second value in the first mapping relationship have a linear relationship. Optionally, with Figure 9 As shown, the texture intensity values ​​and second values ​​within the first texture intensity range in the first mapping relationship can also satisfy a Bézier curve mapping relationship. This application embodiment does not limit the first mapping relationship. A Bézier curve is a smooth parametric curve widely used in computer graphics.

[0146] In some embodiments, the mobile phone may provide a first adjustment option for adjusting the first mapping relationship. By operating this first adjustment option, the user can modify the first mapping relationship, for example, modifying the relationship between the texture intensity values ​​within the second texture intensity range and the second value, or modifying the first value X1 and the third value X3. The first adjustment option can be set on any interface of the mobile phone, for example, an interface in a settings app, or an image viewing interface 620.

[0147] Optionally, any texture analysis method may include: statistical texture features, gray-level co-occurrence matrix (GLCM), or local binary mode. Statistical texture features refer to analyzing texture features by calculating image statistics (such as mean, variance, skewness, kurtosis, etc.). The GLCM calculates the statistical relationships between pixel gray levels in an image, generates a co-occurrence matrix, and extracts texture features from it. Local binary mode analyzes texture features by converting local regions of the image into binary mode.

[0148] For example, the mobile phone calculates the texture intensity of the denoised image using local binary mode. Obtaining the texture intensity information of the denoised image may include: the mobile phone dividing the denoised image into multiple first regions of equal size, and calculating the gray value variance for each of these multiple first regions. The gray value variance of each first region is the texture intensity value of each first region.

[0149] Understandably, if the gray values ​​of all pixels in any region (e.g., any first region) are the same or very close, then the gray value variance of this region will be small, and the corresponding first intensity value will also be small. A small gray value variance indicates that the gray distribution in this region is very uniform, and this region contains relatively few image details. In this case, a small first intensity value can clearly reflect the limited image details in this region, and it also avoids amplifying noise in this region.

[0150] Conversely, if the grayscale values ​​of all pixels in a region differ significantly, then the grayscale variance of that region will be large, and the corresponding first intensity value will also be large. A large difference in grayscale values ​​among all pixels in a region indicates that the grayscale distribution in that region is uneven, and this region may contain more image details or textures. In this case, a large first intensity value for that region can amplify the image details in that region, making the image details in that region richer and clearer.

[0151] S406. The mobile phone performs brightness analysis on the denoised image to obtain brightness information, and determines the second detail intensity distribution map based on the brightness information.

[0152] Since different regions in a denoised image can have varying brightness, and image details are less apparent in lower-brightness regions, a higher detail intensity is needed to amplify these details. Therefore, the phone can determine the brightness information of the denoised image and then determine the corresponding second detail intensity distribution map. The second detail intensity distribution map can include the second intensity value of each second region in the denoised image. The second intensity value of each second region in the denoised image corresponds to the brightness value of that second region. Each second region can include one or more pixels.

[0153] In some embodiments, S406 may include: the mobile phone performing brightness analysis on the denoised image to obtain brightness information; the brightness information may include brightness values ​​of each second region in the denoised image; then, based on a second mapping relationship and the brightness values ​​of each second region in the denoised image, determining a second intensity value for each second region. The second mapping relationship may refer to the mapping relationship between the brightness value and the second intensity value.

[0154] The second detail intensity distribution map and the denoised image can have the same size. The second detail intensity distribution map includes the second intensity values ​​of multiple second regions in the denoised image. The second mapping relationship can include: the lower the brightness of any second region, the higher the second intensity value of that region can be.

[0155] For example, the second mapping relationship between brightness values ​​and second intensity values ​​may include: all brightness values ​​within a first brightness range correspond to the same fourth value X4; the larger the brightness value within the second brightness range, the smaller the corresponding fifth value; and all brightness values ​​within a third brightness range correspond to the same sixth value X6. Here, the larger the brightness value within the second brightness range, the smaller the corresponding fifth value, indicating a positive correlation between the brightness values ​​within the second brightness range and the fifth value. Brightness values ​​within the second brightness range are greater than those within the first brightness range, and brightness values ​​within the third brightness range are greater than those within the second brightness range. The fifth value is less than the fourth value X4, and the sixth value X6 is less than the fifth value. The fourth value X4, the fifth value, and the sixth value X6 all belong to the second intensity value. Figure 10 As shown, the first brightness range can be the range from 0 to the first brightness value L1; the second brightness range can be the range from the first brightness value L1 to the second brightness value L2, and the second brightness range does not include the first brightness value L1; the third brightness range can be the range greater than the second brightness value L2.

[0156] Among them, such as Figure 10 As shown, the brightness values ​​within the first brightness range and the fifth value in the second mapping relationship have a linear relationship. Optionally, with Figure 10As shown, the brightness values ​​within the first brightness range and the fifth value in the second mapping relationship can also satisfy the mapping relationship of the Bézier curve. This application embodiment does not limit the second mapping relationship.

[0157] For example, the process of a mobile phone performing brightness analysis on a denoised image to obtain brightness information may include: the mobile phone dividing the denoised image into multiple second regions of the same size, or the mobile phone dividing multiple pixels with similar pixel values ​​in the denoised image into the same second region, thereby obtaining multiple second regions; and then calculating the average pixel value for each of these multiple second regions. The average pixel value of each second region is the brightness value of each second region.

[0158] Understandably, if the average pixel value of any region (e.g., any second region) is small, the corresponding second intensity value will be large. A small average pixel value in this region indicates low brightness, meaning the image details in this region are not clearly visible in the denoised image. In this case, a larger second intensity value for this region can amplify the image details in that region, making them clearer.

[0159] Conversely, if the average pixel value of a region is large, the corresponding second intensity value will be small. A large average pixel value in a region indicates high brightness, meaning the image details in that region are more apparent in the denoised image and do not need to be amplified. In this case, a smaller second intensity value for that region can avoid amplifying the noise in that region.

[0160] In some embodiments, the mobile phone may provide a second adjustment option for adjusting the second mapping relationship. By operating this second adjustment option, the user can modify the second mapping relationship, for example, modifying the relationship between the brightness values ​​within the second brightness range and the fifth value, or modifying the fourth value x4 and the sixth value x6. The second adjustment option can be set on any interface of the mobile phone, for example, an interface in a settings app, or an image viewing interface 620. The second adjustment option and the first adjustment option can be set on the same interface.

[0161] S407. The mobile phone performs semantic segmentation on the denoised image to obtain a segmented image labeled with semantic categories, and determines the third detail intensity distribution map based on the segmented image.

[0162] Because different types of subjects in a denoised image contain varying amounts of image detail, a mobile phone can perform semantic segmentation on the denoised image to determine the semantic category of each pixel, resulting in a segmented image labeled with semantic categories. The segmented image includes multiple third regions, and each of these third regions is labeled with a corresponding semantic category. Each third region includes one or more pixels. Furthermore, the mobile phone can set a corresponding third intensity value for each third region in the segmented image to generate a third detail intensity distribution map. The third detail intensity distribution map includes the third intensity values ​​of each third region in the segmented image, with each third intensity value corresponding to the semantic category of that third region.

[0163] Semantic segmentation refers to assigning each pixel or region in a denoised image to a predefined semantic category, thereby generating an image of the same size as the denoised image, where each pixel is labeled with a semantic category (i.e., a segmented image). In other words, the segmented image and the denoised image are the same size, and all pixels in the segmented image are labeled with their corresponding semantic categories.

[0164] For example, predefined semantic categories may include: sky, road surface, house, vehicle, person, and plant, etc. Since vehicles, people, and plants contain more image details, while the sky, road surface, and house contain less, the third intensity values ​​corresponding to vehicles, people, and plants can be larger, while the third intensity values ​​corresponding to the sky, road surface, and house can be smaller. The third intensity values ​​corresponding to vehicles, people, and plants can be different; for example, the third intensity value corresponding to people is greater than that corresponding to plants, and the third intensity value corresponding to plants is greater than that corresponding to vehicles. The third intensity values ​​corresponding to the sky, road surface, and house can also be different; for example, the third intensity value corresponding to the sky is greater than that corresponding to houses, and the third intensity value corresponding to houses is greater than that corresponding to roads.

[0165] In some embodiments, S407 may include: the mobile phone inputting the denoised image into a second neural network model to perform semantic segmentation and obtain a segmented image labeled with semantic categories.

[0166] For example, such as Figure 11 As shown, the mobile phone inputs a denoised image into a second neural network model for semantic segmentation, resulting in a segmented image. The second neural network model can be a U-Net model. Each color region in the segmented image corresponds to a third region representing a semantic category.

[0167] It should be noted that the output image and the denoised image are obtained by image processing the same raw image. Therefore, the detail intensity (e.g., the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map) that are adapted to the denoised image, based on the brightness of each region in the denoised image, the texture intensity of each region in the denoised image, and the type of the subject in the denoised image, are also adapted to the output image and can be used to adjust the image details in the output image.

[0168] Optionally, the electronic device can directly perform image analysis on the image to be output to determine the intensity value that is compatible with the image to be output. The specific process by which the electronic device performs image analysis on the image to be output to determine the intensity value that is compatible with the image to be output can be referred to in the detailed description of S405-S407 above regarding image analysis of a denoised image to determine the intensity value that is compatible with the denoised image; this embodiment will not be repeated here.

[0169] It should be noted that since the output image and the denoised image are obtained by performing different image processing on the raw image, and the denoised image does not undergo the third image processing operations including brightness adjustment, color adjustment, and sharpening, the texture intensity information and brightness information obtained by the electronic device from the image analysis of the denoised image are closer to the actual texture intensity and brightness information. In other words, the accuracy of the texture intensity and brightness information obtained by the mobile phone from the image analysis of the denoised image is greater than the accuracy of the texture intensity and brightness information obtained from the image analysis of the output image. Therefore, the detail intensity distribution map (e.g., the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map) obtained by the mobile phone based on the denoised image can more accurately improve the clarity of image details in the output image.

[0170] S408: The mobile phone determines the target detail intensity distribution map based on the specified intensity value, the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map.

[0171] The sizes of the first, second, and third detail intensity maps can be the same, all equal to the size of the denoised image. For example... Figure 6 As shown, the mobile phone can multiply the first intensity value, second intensity value, and third intensity value of the same pixel in the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map to obtain a fourth detail intensity distribution map. The fourth detail intensity distribution map includes the fourth intensity value of every pixel in all pixels. Then, the specified intensity value is multiplied by all the fourth intensity values ​​in the fourth detail intensity distribution map to obtain the target detail intensity distribution map. The target detail intensity distribution map can include the target intensity value of every pixel in all pixels.

[0172] S409. The mobile phone determines the initial detail difference map between the noisy image and the denoised image, and extracts the Y channel information from the initial detail difference map to obtain the first detail difference map.

[0173] Both the noisy image and the denoised image are three-channel images; for example, both the noisy image and the denoised image in RGB format are RGB three-channel images. Figure 6 As shown, the mobile phone can first subtract the noisy image from the denoised image to obtain an initial detail difference map of three channels; then extract the Y channel information from the initial detail difference map to obtain the first detail difference map of the Y channel.

[0174] For example, a mobile phone can perform a subtraction operation on a noisy image and a denoised image to obtain an initial three-channel detail difference map. This can include: the mobile phone can subtract the pixel values ​​in the same channel of the noisy image and the denoised image to obtain a single-channel detail difference map, and then obtain three single-channel detail difference maps; then, the three single-channel detail difference maps are fused to obtain the initial three-channel detail difference map. For example, if the noisy image and the denoised image are RGB three-channel images, then the initial detail difference map is also an RGB three-channel image.

[0175] For example, the mobile phone extracts the Y channel information from the initial detail difference map to obtain the first detail difference map of the Y channel. This can include: if the initial detail difference map is in RGB format, the mobile phone can first convert the format of the initial detail difference map to generate a detail difference map in YUV format; then extract the Y component from the YUV format detail difference map to obtain the first detail difference map of the Y channel.

[0176] A single-channel image includes only one type of channel information. A single-channel difference map includes only the difference in one channel information between the noisy and denoised images. For example, the R-channel difference map only includes the difference in the red channel information between the noisy and denoised images, and the first detail difference map of the Y-channel only includes the difference in the Y-channel information between the noisy and denoised images. A three-channel initial detail difference map can include the differences in all three channels of information between the noisy and denoised images. For example, an RGB three-channel initial detail difference map can include the differences in the RGB three-channel information between the noisy and denoised images.

[0177] Understandably, since the Y channel records the brightness information in an image, and the brightness information affects the clarity of image details, the phone can extract a first detail difference map containing Y channel information from the initial detail difference map of the three channels, and then adjust this first detail difference map to adjust image details.

[0178] It should be noted that the initial detail difference drawing is equivalent to the third detail difference drawing in the invention description.

[0179] S410: The mobile phone uses the target detail intensity distribution map to adjust the first detail difference map to obtain the second detail difference map.

[0180] like Figure 6 As shown, the mobile phone can use the target intensity value in the target detail intensity distribution map to multiply the Y channel information in the first detail difference map of the Y channel to obtain the second detail difference map of the Y channel.

[0181] Understandably, the first detail difference map includes the difference in Y-channel information between the noisy image and the denoised image. Adjusting the first detail difference map by the phone involves adjusting the difference in Y-channel information (i.e., brightness information) between the noisy and denoised images; in other words, adjusting the difference in brightness information between the two images. By multiplying the target intensity value in the target detail intensity distribution map with the first detail difference map, the phone can amplify or reduce the difference in brightness information in any region of the first detail difference map. Amplifying the difference in brightness information in any region of the first detail difference map improves the clarity of image details in that region. Conversely, reducing the difference in brightness information in any region of the first detail difference map reduces the clarity of image details in that region.

[0182] Secondly, since the target intensity value is generated based on the user-selected specified intensity value, the brightness of each region in the denoised image, the texture intensity of each region in the denoised image, and the semantic category to which each region belongs, it can be seen that the target intensity value is adapted to the amount of image detail contained in any region of the denoised image, and also to the amount of image detail contained in any region of the output image. Furthermore, by using this target intensity value to amplify or reduce the difference in Y-channel information (i.e., brightness information in the image) in any region of the first detail difference map, the amplified or reduced brightness information difference can accurately improve the clarity of image details in any region of the output image.

[0183] S411. The mobile phone extracts the Y channel information and UV channel information of the image to be output, obtains the Y channel image and UV channel image, and adds the Y channel image and the second detail difference image to obtain the adjusted Y channel image.

[0184] The image to be output can be in a preset image format (e.g., RGB format). For example, ... Figure 6As shown, the mobile phone can first convert the format of the image to be output to generate a YUV format image; then, it can extract the Y component from the YUV format image to obtain a Y channel image, and extract the UV component from the YUV format image to obtain a UV channel image. Finally, the mobile phone can add the Y channel image and the second detail difference map of the Y channel to obtain the adjusted Y channel image.

[0185] It's understandable that the adjusted Y-channel image, obtained by adding a second detail difference map to the Y-channel image, records the adjusted Y-channel information. This adjusted Y-channel information is essentially the brightness information in the adjusted image. The brightness information in the adjusted image can be used to highlight image details in the output image, thus improving the clarity of those details.

[0186] S412. The mobile phone fuses the UV channel image and the adjusted Y channel image to obtain the fused YUV image.

[0187] For example, such as Figure 6 As shown, the mobile phone fuses the UV channel image and the adjusted Y channel image to obtain the fused YUV image.

[0188] Understandably, the adjusted Y-channel image improves the clarity of image details in the output image. Therefore, the clarity of image details in the fused YUV image obtained by fusing the UV channel image and the adjusted Y-channel image is also improved. Secondly, both the UV channel image and the adjusted Y-channel image are acquired from the output image; therefore, the fused YUV image is also a denoised image.

[0189] S413 The mobile phone performs format conversion on the merged YUV image to obtain an image with adjusted details, and displays the image with adjusted details on the first interface.

[0190] like Figure 6 As shown, the phone camera needs to output an image in a preset image format (e.g., RGB format). Therefore, the phone can convert the format of the merged YUV image to obtain an image with adjusted details in the preset image format. After displaying the image to be output on the first screen, the phone then displays the image with adjusted details, allowing the user to visually see that the image details are clearer.

[0191] For example, such as Figure 8 As shown in (c), the mobile phone displays the image 621 to be output on a first interface (e.g., image viewing interface 620). Then, as... Figure 8 As shown in (d), the phone displays an image 630 with adjusted details on the image viewing interface 620. The phone displays... Figure 8Image 621 to be output, shown in (c) in the figure, is displayed. Figure 8 As shown in (d) of the image 630 with the improved details, the user can visually see that the image details are clearer.

[0192] In some embodiments, the mobile phone can replace the image to be output with an image that has undergone detail adjustments and save the image with the adjusted details.

[0193] In some embodiments, if the user is not satisfied with the image after detail adjustments, they can input a new detail adjustment operation into the mobile phone. The mobile phone responds to the new detail adjustment operation by adjusting the original image to be output, generating a new image with adjusted details; or, the mobile phone responds to the new detail adjustment operation by adjusting the image with adjusted details, generating a new image with adjusted details.

[0194] In some embodiments, unlike the descriptions in S403-S413 above, after obtaining the image to be output, the mobile phone may not use the image to be output as the final image output by the mobile phone camera. Instead, it may continue to adjust the image details of the image to be output to obtain an image with adjusted details, and only use the image with adjusted details as the final image output. In this case, after obtaining the noisy image, the denoised image, and the image to be output, the mobile phone can directly perform image analysis on the denoised image to obtain the analyzed detail intensity distribution map (e.g., including a first detail intensity distribution map, a second detail intensity distribution map, and a third detail intensity distribution map).

[0195] For example, such as Figure 12 As shown in the embodiment of this application, another method for adjusting image details is provided. This method may not include S403-S404, and S405-S407, S701, S409-S412 and S702 are executed after S402.

[0196] S701, the mobile phone determines the target detail intensity distribution map based on the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map.

[0197] Without requiring the user to manually select a specific intensity value, the phone can directly determine the target detail intensity distribution map based on the first, second, and third detail intensity distribution maps. Specifically, the phone can multiply the first, second, and third intensity values ​​of the same pixel in the first, second, and third detail intensity distribution maps to obtain the target detail intensity distribution map.

[0198] For example, with Figure 6Unlike the previous method, the mobile phone does not need to receive detail adjustment operations, nor does it obtain the specified intensity value selected by the detail adjustment operation. The mobile phone can directly obtain the target detail intensity distribution map based on the first detail intensity distribution map, the second detail intensity distribution map, and the third detail intensity distribution map.

[0199] S702. The mobile phone performs format conversion on the merged YUV image to obtain an image with detailed adjustments, and saves the image with detailed adjustments to the first storage location; the first storage location includes the storage location corresponding to the Gallery APP.

[0200] Since the phone uses the image with adjusted details as the final output image, it can save the adjusted image to the first storage location that the user can view, such as the storage location of the gallery app.

[0201] Reference Figure 13 The diagram shown is a flowchart illustrating an image detail adjustment method provided in an embodiment of this application. Figure 13 As shown, the display method may also include the following S801-S805.

[0202] S801, the mobile phone receives and responds to the first operation, acquires a raw image, and performs image processing on the raw image to obtain a noisy image, a denoised image, and an image to be output.

[0203] It should be noted that details of the noisy image, the denoised image, and the image to be output can be found in the above description of the noisy image, the denoised image, and the image to be output; these details will not be repeated here in the embodiments of this application.

[0204] For example, the first operation can be the shooting operation described above.

[0205] For example, S801 may include the above-described S401-S402.

[0206] S802, The mobile phone determines the target detail intensity distribution map; the target detail intensity distribution map is used to indicate the adjustment intensity of image details.

[0207] It should be noted that the details of the target detail intensity distribution map can be found in the above-described description of the target detail intensity distribution map, and will not be repeated here in the embodiments of this application.

[0208] For example, S802 may include S403-S408 as described above. Alternatively, S802 may include S405-S407 and S701 as described above.

[0209] S803, the mobile phone obtains a first detail difference map based on the noisy image and the denoised image; the first detail difference map includes the detail differences between the noisy image and the denoised image.

[0210] For example, S803 may include the above-described S409.

[0211] S804. The mobile phone uses the target detail intensity distribution map to adjust the first detail difference map to obtain the second detail difference map; the second detail difference map includes the adjusted detail difference between the noisy image and the denoised image.

[0212] For example, S804 may include the above-described S410.

[0213] S805: The mobile phone obtains the image with adjusted details based on the image to be output and the second detail difference image.

[0214] It should be noted that the details of the image after the detail adjustments can be found in the above description of the image after the detail adjustments, and will not be repeated here in the embodiments of this application.

[0215] For example, S805 may include S411-S413 as described above. Alternatively, S805 may include S411-S412 and S702 as described above.

[0216] It is understood that, in order to achieve the aforementioned functions, the electronic device includes corresponding hardware structures and / or software modules for performing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this application.

[0217] This application embodiment can divide the above-described electronic device into functional modules based on the method example described above. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing module. The integrated modules can be implemented in hardware or as software functional modules. It should be noted that the module division in this embodiment is illustrative and only represents one logical functional division; in actual implementation, there may be other division methods.

[0218] This application also provides an electronic device, which includes: a memory and one or more processors; the memory is coupled to the processors; wherein the memory stores computer program code, the computer program code including computer instructions, and when the computer instructions are executed by the processor, the electronic device performs the image detail adjustment method provided in the foregoing embodiments.

[0219] This application also provides a computer-readable storage medium including computer instructions that, when executed on an electronic device, cause the electronic device to perform the image detail adjustment method provided in the foregoing embodiments.

[0220] This application also provides a computer program product containing executable instructions that, when run on an electronic device, cause the electronic device to perform the image detail adjustment method provided in the foregoing embodiments.

[0221] This application also provides a chip system, such as... Figure 14 As shown, the chip system 900 includes at least one processor 901 and at least one interface circuit 902. The processor 901 and the interface circuit 902 are interconnected via lines. For example, the interface circuit 902 can be used to receive signals from other devices (e.g., the memory of an electronic device). As another example, the interface circuit 902 can be used to send signals to other devices (e.g., the processor 901).

[0222] For example, interface circuit 902 can read instructions stored in memory and send those instructions to processor 901. When the instructions are executed by processor 901, the chip system can perform the steps in the above embodiments. Of course, the chip system may also include other discrete devices, and this application embodiment does not specifically limit this.

[0223] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, 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 can be divided into different functional modules to complete all or part of the functions described above.

[0224] In the several embodiments provided in this application, it should be understood that the disclosed apparatus / device and method can be implemented in other ways. For example, the apparatus / device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0225] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units; that is, it can be located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0226] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0227] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, essentially or in other words, the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0228] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for adjusting image details, characterized in that, The adjustment method includes: Receive and respond to the first operation, acquire a raw image, and perform image processing on the raw image to obtain a noisy image, a denoised image, and an image to be output, wherein the image to be output is the denoised image; A target detail intensity distribution map is determined based on the denoised image, and the target detail intensity distribution map is used to indicate the adjustment intensity of image details; A first detail difference map is obtained based on the noisy image and the denoised image, the first detail difference map including the detail differences between the noisy image and the denoised image; Using the target detail intensity distribution map, the first detail difference map is adjusted to obtain the second detail difference map; Based on the image to be output and the second detail difference map, the image after detail adjustment is obtained.

2. The method according to claim 1, characterized in that, The step of obtaining a first detail difference map based on the noisy image and the denoised image includes: The noisy image and the denoised image are subtracted to obtain a third detail difference image; The Y-channel information is extracted from the third detail difference map to obtain the first detail difference map of the Y channel.

3. The method according to claim 2, characterized in that, The first detail difference map, the target detail intensity distribution map, and the denoised image are all the same size; the target detail intensity distribution map includes the target intensity value corresponding to each pixel. The step of adjusting the first detail difference map using the target detail intensity distribution map to obtain the second detail difference map includes: The target intensity value in the target detail intensity distribution map and the Y channel information in the first detail difference map are multiplied together to obtain the second detail difference map.

4. The method according to claim 2 or 3, characterized in that, The step of obtaining the detail-adjusted image based on the image to be output and the second detail difference map includes: The Y-channel and UV-channel information of the image to be output are extracted to obtain the Y-channel image and the UV-channel image; The second detail difference image and the Y channel image are added together to obtain the adjusted Y channel image. The UV channel image and the adjusted Y channel image are fused and their formats converted to obtain the image with detailed adjustments.

5. The method according to any one of claims 1-4, characterized in that, The noisy image is obtained by performing a first image processing on the raw image, the denoised image is obtained by performing a second image processing on the raw image, and the image to be output is obtained by performing a third image processing on the denoised image; Wherein, the first image processing includes depigmentation; the second image processing includes depigmentation and noise reduction; and the third image processing includes at least one of brightness adjustment, color adjustment, and sharpening.

6. The method according to any one of claims 1-5, characterized in that, The adjustment method further includes: saving the image after the detailed adjustments in a first storage location; the first storage location includes the storage location corresponding to the gallery APP.

7. The method according to any one of claims 1-6, characterized in that, The adjustment method further includes: Display a first interface, which includes the image to be output; Receive the second operation on the first interface. The step of determining the target detail intensity distribution map based on the denoised image includes: in response to the second operation, determining the target detail intensity distribution map based on the denoised image; The adjustment method further includes: displaying the image after the detailed adjustments on the first interface.

8. The method according to any one of claims 1-7, characterized in that, Determining the target detail intensity distribution map based on the denoised image includes: Image analysis is performed on the denoised image to obtain the detailed intensity distribution map after analysis; Based on the analyzed detail intensity distribution map, the target detail intensity distribution map is determined.

9. The method according to claim 8, characterized in that, The step of performing image analysis on the denoised image to obtain the analyzed detail intensity distribution map includes: The denoised image is analyzed for texture intensity to obtain texture intensity information, and a first detail intensity distribution map is determined based on the texture intensity information; the first intensity value in the first detail intensity distribution map corresponds to the texture intensity value in the texture intensity information. The denoised image is subjected to brightness analysis to obtain brightness information, and a second detail intensity distribution map is determined based on the brightness information; the second intensity value in the second detail intensity distribution map corresponds to the brightness value in the brightness information. The analyzed detail intensity distribution map includes the first detail intensity distribution map and the second detail intensity distribution map.

10. The method according to claim 9, characterized in that, The first detail intensity distribution map includes a first intensity value that is positively correlated with the texture intensity value; the second detail intensity distribution map includes a second intensity value that is negatively correlated with the brightness value. The step of determining the target detail intensity distribution map based on the analyzed detail intensity distribution map includes: The target detail intensity distribution map is obtained by multiplying the first intensity value in the first detail intensity distribution map and the second intensity value in the second detail intensity distribution map.

11. The method according to claim 7, characterized in that, The step of determining a target detail intensity distribution map based on the denoised image in response to the second operation includes: Determine the specified intensity value selected in the second operation; Image analysis is performed on the denoised image to obtain the detailed intensity distribution map after analysis; Based on the specified intensity value and the analyzed detail intensity distribution map, the target detail intensity distribution map is determined.

12. An electronic device, characterized in that, The electronic device includes: a processor, a memory, and a communication interface; the memory and the communication interface are coupled to the processor, the memory is used to store computer program code, the computer program code including computer instructions; wherein, when the processor executes the computer instructions, the electronic device performs the method as described in any one of claims 1-11.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions; when the computer instructions are executed on the electronic device, the electronic device causes the electronic device to perform the method as described in any one of claims 1-11.

14. A chip system, characterized in that, The chip system includes a processor and an interface circuit, which are interconnected via a line; wherein the interface circuit is used to receive signals from an electronic device and send the signals to the processor, the signals including computer instructions; when the processor executes the computer instructions, the chip system performs the method as described in any one of claims 1-11.

Images