Electronic device for correcting and outputting image and control method

Abstract

This electronic device comprises: a memory for storing one or more storage media and instructions; a camera; and at least one processor communicatively connected to the memory and the camera and including processing circuitry, wherein the instructions, when executed individually or collectively by the at least one processor, cause the electronic device to: when a capturing event is identified, acquire a plurality of raw images having different exposure values by using the camera; acquire an output image by using the plurality of raw images; acquire weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image; acquire semantic segmentation information corresponding to the output image; acquire noise information on the basis of the semantic segmentation information and the weight information; and acquire a corrected output image on the basis of the output image and the noise information.

Description

Electronic device and control method for correcting and outputting an image

[0001] The present disclosure relates to an electronic device and a control method for correcting and outputting an image.

[0002] There are several ways to add noise to an image, most of which are used in image processing and computer vision.

[0003] Dithering, which adds noise to an image, can be used in digital image processing to minimize color banding, flatness, or other visual artifacts in low bit depth images or displays by adding small amounts of noise or patterns to the image.

[0004] The above information may be provided as background art to aid in understanding the present disclosure. No claim or determination is made as to whether any of the above-described matters constitute prior art related to the present disclosure.

[0005] Aspects of the present disclosure address at least the problems and / or disadvantages mentioned above, and provide at least the advantages described below. Accordingly, aspects of the present disclosure provide an electronic device for correcting and outputting an image, and a control method thereof.

[0006] Additional aspects will be set forth in part in the description below, and in part will be apparent from the description or may be learned by practicing the embodiments provided.

[0007] According to one embodiment of the present disclosure, an electronic device includes: a memory storing one or more storage media and instructions; a camera; and at least one processor including processing circuitry communicatively connected to the memory and the camera; wherein the instructions, when individually or collectively executed by the at least one processor, cause the electronic device (100) to, when a photographing event is identified, acquire a plurality of raw images having different exposure values ​​through the camera (130), acquire an output image using the plurality of raw images, acquire weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image, acquire semantic segmentation information corresponding to the output image, acquire noise information based on the semantic segmentation information and the weight information, and acquire a corrected output image based on the output image and the noise information.

[0008] A method for controlling an electronic device according to an embodiment of the present disclosure includes, when a shooting event is identified, an operation of acquiring a plurality of raw images having different exposure values ​​through the camera; an operation of acquiring an output image using the plurality of raw images; an operation of acquiring weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image; an operation of acquiring semantic segmentation information corresponding to the output image; an operation of acquiring noise information based on the semantic segmentation information and the weight information; and an operation of acquiring a corrected output image based on the output image and the noise information.

[0009] One or more non-transitory computer-readable storage media storing one or more computer programs including computer-executable instructions that, when individually or collectively executed by one or more processors, cause an electronic device to perform operations, the operations comprising: when a photographing event is identified, acquiring a plurality of raw images having different exposure values ​​through the camera; acquiring an output image using the plurality of raw images; acquiring weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image; acquiring semantic segmentation information corresponding to the output image; acquiring noise information based on the semantic segmentation information and the weight information; and acquiring a corrected output image based on the output image and the noise information.

[0010] Other aspects, advantages and key features of the present disclosure will become apparent to those skilled in the art from the following detailed description taken in conjunction with the accompanying drawings.

[0011] The above and other aspects, features and advantages of specific embodiments of the present disclosure will become more apparent from the following description taken in conjunction with the accompanying drawings.

[0012] FIG. 1 is a drawing for explaining an image processing method of an electronic device (100) according to one embodiment of the present disclosure.

[0013] FIGS. 2a, 2b, 2c, and 2d are drawings for explaining an image processing method of an electronic device (100) according to one embodiment of the present disclosure.

[0014] FIG. 3 illustrates an example of a block diagram of an electronic device (100) according to one embodiment of the present disclosure.

[0015] FIG. 4 is a flowchart for explaining the operation of an electronic device (100) according to one embodiment of the present disclosure.

[0016] FIG. 5a and FIG. 5b are drawings for explaining a method for obtaining weight information according to an embodiment of the present disclosure.

[0017] FIG. 6 is a diagram for explaining a method for obtaining semantic segmentation information according to one embodiment of the present disclosure.

[0018] FIG. 7 is a diagram for explaining a noise information acquisition method according to an embodiment of the present disclosure.

[0019] FIG. 8 is a drawing for explaining an output image processed according to one embodiment of the present disclosure.

[0020] FIG. 9a and FIG. 9b are diagrams for explaining a method for obtaining noise information based on weight information according to an embodiment of the present disclosure.

[0021] FIG. 10A and FIG. 10B are diagrams for explaining a method for obtaining noise information based on semantic segmentation information according to an embodiment of the present disclosure.

[0022] FIG. 11a and FIG. 11b are diagrams for explaining a method for obtaining noise information based on semantic segmentation information according to an embodiment of the present disclosure.

[0023] FIG. 12 is a drawing for explaining an example of an output image correction method according to one embodiment of the present disclosure.

[0024] FIG. 13 is a block diagram of an electronic device within a network environment according to various embodiments.

[0025] It should be noted that the same reference numbers are used throughout the drawings to indicate identical or similar elements, features and structures.

[0026]

[0027] The following description, with reference to the attached drawings, is provided to assist in a comprehensive understanding of various embodiments of the present disclosure, as defined by the claims and their equivalents. While this description includes various specific details to aid in this understanding, these should be considered merely exemplary. Accordingly, those skilled in the art will recognize that various changes and modifications can be made to the various embodiments described herein without departing from the spirit and scope of the present disclosure. Furthermore, descriptions of well-known functions and configurations may be omitted for clarity and conciseness.

[0028] The terms and words used in the following description and claims are not limited to their dictionary meanings, but rather have been used by the inventors to clearly and consistently describe the present disclosure. Therefore, those skilled in the art will readily recognize that the following description of various embodiments of the present disclosure is for illustrative purposes only and is not intended to limit the present disclosure, which is defined by the appended claims and their equivalents.

[0029] Unless the context clearly indicates otherwise, the singular forms "a," "an," and "the" should be understood to include plural references. Thus, for example, a reference to "a component surface" should be interpreted to include one or more such surfaces.

[0030] In this specification, expressions such as “has,” “can have,” “includes,” or “may include” indicate the presence of a feature (e.g., a component such as a number, function, operation, or part), and do not exclude the presence of additional features.

[0031] The expression "at least one of A and / or B" should be understood to mean either "A" or "B" or "A and B".

[0032] As used herein, expressions such as “first,” “second,” “first,” or “second,” can describe various components, regardless of order and / or importance, and are only used to distinguish one component from another, but do not limit said components.

[0033] When it is said that a component (e.g., a first component) is “(operatively or communicatively) coupled with / to” or “connected to” another component (e.g., a second component), it should be understood that the component may be directly coupled to the other component, or may be coupled via another component (e.g., a third component).

[0034] Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, terms such as "comprise" or "consist of" are intended to indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the specification, but should be understood not to preclude the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0035] In the embodiments of the present disclosure, a "module" or "part" performs at least one function or operation and may be implemented in hardware or software, or a combination of hardware and software. Furthermore, multiple "modules" or multiple "parts" may be integrated into at least one module and implemented as at least one processor, excluding any "modules" or "parts" that require specific hardware implementation.

[0036] In this disclosure, the term user may refer to a person using an electronic device or a device (e.g., an artificial intelligence electronic device) using an electronic device.

[0037] It should be noted that the blocks and combinations of flowcharts in each flowchart can be executed by one or more computer programs containing computer-executable instructions. One or more computer programs may be stored entirely in a single memory device, or may be divided into multiple memory devices, each with its own portion stored therein.

[0038] Any function or operation described in this disclosure may be processed by one processor or a combination of multiple processors. Such a processor or combination of processors may include a circuit that performs a processing function, such as an application processor (AP, e.g., a central processing unit (CPU)), a communication processor (CP, e.g., a modem), a graphical processing unit (GPU), a neural processing unit (NPU, e.g., an artificial intelligence (AI) chip), a Wi-Fi chip, a Bluetooth™ chip, a global positioning system (GPS) chip, a near field communication (NFC) chip, connectivity chips, a sensor controller, a touch controller, a fingerprint sensor controller, a display drive integrated circuit (IC), an audio CODEC chip, a universal serial bus (USB) controller, a camera controller, an image processing IC, a microprocessor unit (MPU), a system on a chip (SoC), an integrated circuit (IC), or similar circuits.

[0039] The various elements and areas in the drawings are schematically drawn. Therefore, the technical concepts of the present disclosure are not limited by the relative sizes or spacings drawn in the attached drawings.

[0040] Hereinafter, embodiments of the present disclosure will be described in more detail with reference to the attached drawings.

[0041] FIG. 1, FIG. 2a, FIG. 2b, FIG. 2c, and FIG. 2d are drawings for explaining an image processing method of an electronic device (100) according to one embodiment of the present disclosure.

[0042] According to one embodiment, referring to FIGS. 1, 2a, 2b, 2c and 2d, the electronic device (100) can acquire raw data by performing a photographing operation through the camera (130) according to a photographing event.

[0043] Raw data can be an image file containing raw data captured by a camera sensor, stored in its unprocessed state. For example, raw data can contain pure image sensor output data without any image processing applied.

[0044] For example, raw data may include a plurality of raw images (10) as illustrated in FIG. 1. The plurality of raw images (or raw frames) may also be referred to as a multi-image raw file (Multi-Frame RAW file, MFR file).

[0045] Multiple raw images may be generated through multiple captures. For example, the camera (130) may capture multiple images in succession at very short intervals, and all of these images may be stored in raw data format. Each image may have subtle differences during capture. For example, each image may have a different exposure value.

[0046] According to one embodiment of the present disclosure, an electronic device (100) can synthesize and image-process a plurality of raw images (10) using an artificial intelligence model (20) (or a neural network model). For example, the electronic device (100) can input a plurality of raw images into an artificial intelligence model (20) to obtain an output image (30) in which a plurality of raw images are synthesized.

[0047] According to one embodiment of the present disclosure, the artificial intelligence model (20) may be implemented as an Image-to-Image AI model that outputs an image-processed image when an image is input. The Image-to-Image AI model may be a deep learning model that performs a task of generating an output image through image processing when at least one image is input. For example, the Image-to-Image AI model may be implemented as an AI model that uses a deep learning algorithm such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), or Diffusion Models. The image processing may be digital image processing including at least one of image enhancement, image restoration, image transformation, image analysis, image understanding, image compression, image decoding, or scaling.

[0048] Typically, the output image (30) from the Image-to-Image AI model appears blurry due to the regression to the mean phenomenon. Since the user's eyes perceive sharpness when pixel value changes significantly, if there is a regression to the mean phenomenon, the range of this change is reduced, making the image appear relatively blurry.

[0049] As an example, the upper drawing of FIG. 2A is an example of a first image (210) before being input to an artificial intelligence model (20), and the lower drawing is an example of a second image (220) output from the artificial intelligence model (20). As an example, when comparing a band area (211) of a hat portion in the first image (210) and a band area (221) of the same hat portion in the second image (220), a phenomenon occurs in which the image appears blurred due to a regression to the mean phenomenon in the band area (221) of the hat portion in the second image (220).

[0050] For example, FIGS. 2b, 2c and 2d are charts showing R (Red) pixel values, G (Green) pixel values ​​and B (Blue) pixel values ​​of an 8x8 grid area within a band area (221) of a hat portion in the first image (210) and the second image (220).

[0051] Referring to FIG. 2b, the line R-210 represents the R pixel value of the first image (210), and R-220 represents the R pixel value of the second image (220).

[0052] Referring to FIG. 2c, the line G-210 represents the G pixel value of the first image (210), and G-220 represents the G pixel value of the second image (220).

[0053] Referring to FIG. 2d, the line B-210 represents the B pixel value of the first image (210), and B-220 represents the B pixel value of the second image (220).

[0054] Referring to FIG. 2b, FIG. 2c and FIG. 2d, it can be confirmed that the pixel value change range of the second image (220) output from the artificial intelligence model (20) is smaller than the pixel value change range of the first image (210) before being input to the artificial intelligence model (20).

[0055] In this case, adding appropriate noise to the second image (220) can alleviate the regression to the mean phenomenon, improve the texture, and make the image look more natural.

[0056] Below, various embodiments that can improve the quality of a second image (220) by generating effective noise using the weight information of an artificial intelligence model (20) rather than simply generating uniform noise will be described.

[0057] FIG. 3 illustrates an example of a block diagram of an electronic device (100) according to one embodiment of the present disclosure.

[0058] Referring to FIG. 3, according to various embodiments, the electronic device (100) may be at least partially similar to the electronic device (1301) of FIG. 13, or may include other embodiments of the electronic device.

[0059] In one embodiment of the present disclosure, in terms of being owned by a user, the electronic device (100) may be referred to as a terminal (or user terminal). The terminal may include, for example, a personal computer (PC) such as a laptop or desktop. The terminal may include, for example, a smartphone, a smartpad, and / or a tablet PC. The terminal may include a smart accessory such as a smartwatch and / or a head-mounted device (HMD). According to one embodiment, the electronic device (100) may include a deformable housing. Based on the deformability, the housing of the electronic device (100) may be divided into a plurality of parts.

[0060] According to one embodiment of the present disclosure, an electronic device (100) may include at least one of a processor (110), a memory (120), a camera (130), a display (140), or a communication circuit (150). The processor (110), the memory (120), the camera (130), the display (140), and the communication circuit (150) may be electronically and / or operably coupled with each other by an electronic component such as a communication bus.

[0061] In one embodiment of the present disclosure, the hardwares of the electronic device (100) being operatively coupled may mean that a direct connection or an indirect connection is established between the hardwares, either wired or wireless, so that the second hardware is controlled by the first hardware among the hardwares. Although illustrated based on different blocks, the embodiment is not limited thereto, and some of the hardwares of FIG. 3 (e.g., at least a portion of the processor (110), the memory (120), and the communication circuit (150)) may be included in a single integrated circuit such as a system on a chip (SoC). The type and / or number of hardware included in the electronic device (100) is not limited to that illustrated in FIG. 3. For example, the electronic device (100) may include only some of the hardware components illustrated in FIG. 2.

[0062] According to one embodiment of the present disclosure, a processor (110) of an electronic device (100) may include hardware for processing data based on one or more instructions. The hardware for processing data may include, for example, an arithmetic and logic unit (ALU), a floating point unit (FPU), a field programmable gate array (FPGA), a central processing unit (CPU), a graphics processing unit (GPU), a neural processing unit (NPU), and / or an application processor (AP). The number of processors (110) may be one or more. For example, the processor (110) may have a multi-core processor structure such as a dual core, a quad core, or a hexa core.

[0063] A CPU is a general-purpose processor capable of performing not only general calculations but also artificial intelligence calculations. Its multi-layered cache structure allows for the efficient execution of complex programs. CPUs are advantageous for serial processing, enabling organic linking of previous and subsequent calculation results through sequential calculations. A general-purpose processor is not limited to the examples described above, except where specifically designated as a CPU.

[0064] A GPU is a processor designed for large-scale computations, such as floating-point operations used in graphics processing. It integrates a large number of cores to perform large-scale computations in parallel. In particular, GPUs may be advantageous over CPUs in parallel processing methods, such as convolution operations. Furthermore, GPUs can be used as coprocessors to supplement the functions of CPUs. Processors for large-scale computations are not limited to the examples described above, except in cases where they are specifically referred to as GPUs.

[0065] An NPU is a processor specialized in artificial intelligence operations using artificial neural networks, and each layer of the artificial neural network can be implemented in hardware (e.g., silicon). Since NPUs are designed specifically according to the company's specifications, they have less freedom than CPUs or GPUs, but can efficiently process the artificial intelligence operations requested by the company. Meanwhile, as a processor specialized in artificial intelligence operations, an NPU can be implemented in various forms, such as a Tensor Processing Unit (TPU), an Intelligence Processing Unit (IPU), or a Vision Processing Unit (VPU). Except as specifically designated as an NPU, the artificial intelligence processor is not limited to the examples described above.

[0066] According to one embodiment of the present disclosure, the memory (120) of the electronic device (100) may include a hardware component for storing data and / or instructions input and / or output to the processor (110). The memory (120) may include, for example, a volatile memory such as a random-access memory (RAM) and / or a non-volatile memory such as a read-only memory (ROM). The volatile memory may include, for example, at least one of a dynamic RAM (DRAM), a static RAM (SRAM), a cache RAM, and a pseudo SRAM (PSRAM). The non-volatile memory may include, for example, at least one of a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), a flash memory, a hard disk, a compact disc, a solid state drive (SSD), and an embedded multimedia card (eMMC).

[0067] According to an embodiment of the present disclosure, one or more instructions (or commands) representing operations and / or actions to be performed on data by a processor (110) may be stored in a memory (120) of an electronic device (100). A set of one or more instructions may be referred to as firmware, an operating system, a process, a routine, a sub-routine, and / or an application. For example, the electronic device (100) and / or the processor (110) may perform various operations when a set of a plurality of instructions distributed in the form of an operating system, firmware, a driver, and / or an application is executed. Hereinafter, the fact that an application is installed in an electronic device (100) may mean that one or more instructions provided in the form of an application are stored in the memory (120) of the electronic device (100), and that the one or more applications are stored in a format executable by the processor (110) of the electronic device (100) (e.g., a file having an extension specified by the operating system of the electronic device (100).

[0068] One or more processors (110) control input data to be processed according to predefined operating rules or artificial intelligence (AI) models stored in memory (120). The predefined operating rules or AI models are characterized by being created through learning. Being created through learning means that predefined operating rules or AI models with desired characteristics are created by applying a learning algorithm to a plurality of learning data. This learning may be performed on the device itself on which the artificial intelligence according to the present disclosure is performed, or may be performed through a separate server / system.

[0069] An AI model may be composed of multiple neural network layers. At least one layer has at least one weight value and performs its operation through the operation result of the previous layer and at least one defined operation. Examples of neural networks include a convolutional neural network (CNN), a recurrent neural network (RNN), a deep neural network (DNN), a restricted boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), deep Q-networks, and a transformer. The neural networks in the present disclosure are not limited to the above-described examples unless otherwise specified.

[0070] A learning algorithm is a method for training a target device (e.g., a robot) using a large amount of learning data, enabling the target device to make decisions or predictions on its own. Examples of learning algorithms include supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. Unless otherwise specified, the learning algorithms in this disclosure are not limited to the aforementioned examples.

[0071] According to one embodiment, the camera (130) of the electronic device (100) may be turned on and perform shooting according to a preset event. The camera (130) may convert a captured (or photographed) image into an electrical signal and generate image data based on the converted signal. For example, a subject may be converted into an electrical image signal through a semiconductor optical element (CCD; Charge Coupled Device), and the converted image signal may be amplified and converted into a digital signal and then signal processed. For example, the camera (130) may include at least one of a general (or basic) camera, an ultra-wide-angle camera, and a depth camera.

[0072] According to one embodiment, the display (140) of the electronic device (100) can output visualized information to the user. For example, the display (140) can be controlled by a controller such as a GPU (graphic processing unit) to output visualized information to the user. The display (140) can include an OLED (Organic Light Emitting Diodes) display, an LED (Light Emitting Diodes), a micro LED, a Mini LED, a PDP (Plasma Display Panel), a QD (Quantum Dot) display, a QLED (Quantum Dot Light-Emitting Diodes), and / or an e-ink display or / and an e-paper display. According to one example, the display (140) can be implemented as a flat display, a curved display, a foldable or / and rollable flexible display, etc.

[0073] The communication circuit (150) of the electronic device (100) according to one embodiment may include hardware for supporting transmission and / or reception of electrical signals between the electronic device (100) and an external device (e.g., a server). The communication circuit (150) may include, for example, at least one of a modem, an antenna, and an optical / electronic (O / E) converter. The communication circuit (150) may support various wireless communication technologies, such as Ethernet, a local area network (LAN), a wide area network (WAN), wireless fidelity (WiFi), near field communication (NFC), and Bluetooth. TM , BLE (bluetooth low energy), ZigBee, LTE (long term evolution), 5G (fifth generation) NR (new radio) and / or 6G (sixth generation) can support transmission and / or reception of electrical signals based on various types of protocols.

[0074] In one example, the electronic device (100) may be connected to the server based on a wired network and / or a wireless network. The wired network may include a network such as the Internet, a local area network (LAN), a wide area network (WAN), Ethernet, or a combination thereof. The wireless network may include a network such as long term evolution (LTE), 5G new radio (NR), wireless fidelity (WiFi), Zigbee, near field communication (NFC), Bluetooth, Bluetooth low-energy (BLE), or a combination thereof. In one example, the electronic device (100) and the server may be indirectly connected via an intermediate node within the network.

[0075] FIG. 4 is a flowchart for explaining the operation of an electronic device (100) according to one embodiment of the present disclosure.

[0076] In the following examples, the operations may be performed sequentially, but are not necessarily performed sequentially. For example, the order of the operations may be changed, and at least two operations may be performed in parallel.

[0077] According to one embodiment of the present disclosure, referring to FIG. 4, operations 410 to 470 may be understood to be performed in the processor (110) of the electronic device (100).

[0078] In operation 410, the electronic device (100) may identify a shooting event. The shooting event may be an event that initiates shooting via the camera (130). For example, the electronic device (100) may identify that a shooting event has occurred based on at least one of receiving a shooting command (e.g., button input, voice reception, gesture recognition, etc.), the passage of a preset time, and satisfaction of a preset condition.

[0079] When a shooting event is identified (operation 410:Y), the electronic device (100) may acquire multiple raw images through the camera (130) in operation 420. In one example, the camera (130) may generate multiple raw images (or multi-image raw files) through multiple exposure shooting. Multiple exposure shooting may be a method of shooting the same scene multiple times with different exposure settings. For example, a low exposure image for details in dark areas, a high exposure image for details in bright areas, and an image with medium brightness may be shot. Accordingly, the multiple raw images may have different exposure values.

[0080] For example, when a shooting event is identified, the camera (130) can set auto-focus (AF) and auto-exposure (AE) before the shutter opens. During this process, the camera (130) can measure the distance to the subject and determine appropriate exposure settings (e.g., ISO, shutter speed, aperture) for the environment. After this, the shutter of the camera (130) opens, allowing light to enter the image sensor (e.g., CMOS or CCD) to capture an image. During this process, the image sensor of the camera (130) can convert the light into an electrical signal.

[0081] For example, the camera (130) can take multiple pictures in succession at very short time intervals while changing the exposure settings for multiple exposure shooting and store the obtained raw data in the memory (120).

[0082] In operation 430, the electronic device (100) may obtain an output image using a plurality of raw images. In one example, the electronic device (100) may obtain an output image synthesized from a plurality of raw images using an image conversion module (or image processing software). In one example, the image conversion module may include a first artificial intelligence model.

[0083] For example, the first AI model may be implemented as an Image-to-Image AI model for improving image quality. The Image-to-Image AI model may be a deep learning model that processes at least one input image to generate an output image. For example, the Image-to-Image AI model may be implemented as an AI model utilizing deep learning algorithms such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), or Diffusion Models.

[0084] For example, when multiple raw images are input, the first artificial intelligence model can synthesize the multiple raw images to output a single output image. For example, the first artificial intelligence model can denoise the multiple raw images to generate a single output image. However, the present invention is not limited thereto, and the first artificial intelligence model can synthesize the multiple raw images through various image processing such as multi-exposure synthesis, quality improvement, color correction, white balance, and HDR (High Dynamic Range) adjustment to generate an output image.

[0085] For convenience of explanation, the following description assumes that the first artificial intelligence model generates an output image (or output images) by performing multi-exposure fusion on multiple raw images. Multi-exposure fusion may be an image processing method that generates an output image by synthesizing multiple raw images captured under different exposure conditions.

[0086] Exposure refers to the degree to which the image sensor is exposed to light, and can be determined by shutter speed, aperture, and / or ISO settings. For example, a higher exposure (overexposure) can result in a brighter image, while a lower exposure (underexposure) can result in a darker image.

[0087] Fusion can combine multiple raw images, selecting only the optimally exposed portions of each raw image to create a single output image. During this process, the first AI model can analyze the brightness, contrast, and color of each raw image and combine the appropriate portions. For example, the first AI model can select a low-exposure image for bright areas and a high-exposure image for dark areas to create the output image. For example, the first AI model can select a bright sky area from the low-exposure image and a dark shadow area from the high-exposure image to create the output image. In the low-exposure image, the bright sky area is not overexposed but the dark areas are barely visible, whereas in the high-exposure image, the dark areas are clearly visible and the sky area may be overly bright, losing detail. During this process, the first AI model can blend the edges to create a natural transition and hide the seam.

[0088] In operation 440, the electronic device (100) can obtain weight information for each area related to the degree of synthesis of multiple raw images.

[0089] According to an embodiment of the present disclosure, the electronic device (100) may obtain weight information from the first artificial intelligence model. The weight information may include information on the degree of synthesis of multiple raw images for each area included in the output image. The degree of synthesis may be information indicating how much a specific area of ​​each raw image will contribute to the final composite image when the first artificial intelligence model combines raw images captured under different exposure conditions. For example, the weight information may be determined based on the brightness, contrast, noise level, etc. of each raw image. For example, if a raw image with a low EV value (e.g., -2 EV) contributes to the synthesis, noise is likely to remain in the output image, and therefore, the weight information reflects the noise characteristics of the multiple raw images.

[0090] For example, weight information used in image synthesis of the first artificial intelligence model may be stored in the memory (120). Accordingly, the electronic device (100) may obtain the weight information stored in the memory (120).

[0091] For example, the weight information may be in the form of a weight map (hereinafter, referred to as a first map) that includes weights for each region of the image. For example, the weight map may include a weight corresponding to at least one of a plurality of raw images for each region. In the present disclosure, a "region" is a term referring to a portion of an image and means at least one pixel block or a set of pixel blocks. In addition, a "pixel block" means a set of adjacent pixels that include at least one pixel.

[0092] In operation 450, the electronic device (100) can obtain semantic segmentation information corresponding to the output image.

[0093] For example, the electronic device (100) can obtain semantic segmentation information corresponding to an output image using a second artificial intelligence model.

[0094] Semantic segmentation information may be information that classifies each pixel in an image into a specific class (e.g., person, sky, tree, car, etc.). For example, the semantic segmentation information may be in the form of a map (hereinafter, referred to as a second map) containing semantic information for each region of the image. The second map may be in the form of a label (e.g., color, indicator, flag, text, etc.) that indicates which object or background each pixel in the image belongs to. For example, semantic segmentation may not distinguish individual objects, but may classify all objects belonging to the same class equally. For example, in an image with multiple people, all people may be classified into the same "person" class. For example, the second map may have the same resolution as the input image and may provide accurate classification on a pixel-by-pixel basis.

[0095] For example, the second AI model can assign each pixel in an image to a specific class. For example, in a photographic image of a road, each pixel can be classified into a class such as "road," "car," "pedestrian," or "sky." For example, in the second map, each pixel can include a predefined class label. A class label can be information that specifies which class a pixel belongs to within the image. For example, if the label is expressed as a color, a road may be gray, a car red, and a pedestrian green.

[0096] For example, the second AI model can be implemented as a deep learning model utilizing a convolutional neural network (CNN). For example, the second AI model can be implemented as a fully convolutional network (FCN), U-Net, SegNet, DeepLab, etc. The second AI model can learn the characteristics of an image and predict the most appropriate class label for each pixel.

[0097] In operation 460, the electronic device (100) may acquire noise information based on semantic segmentation information and weight information. For example, the noise information may be noise information to be added to an output image through dithering.

[0098] According to one embodiment of the present disclosure, the electronic device (100) can obtain noise information based on a first map including weight information, a second map including semantic segmentation information, and an output image.

[0099] For example, the exposure value can be a combination of aperture (f / value), shutter speed, and ISO, which determine the brightness of an image. A higher exposure value (e.g., EV +1, +2) makes the image brighter, and a lower exposure value (e.g., EV -1, -2) makes the image darker. Noise is unwanted noise in an image, which can primarily occur in the image sensor of a camera (130). For example, increasing the ISO increases the sensitivity of the image sensor, but may also increase noise. Accordingly, as the EV value changes, the amount and characteristics of noise appearing in the image may also change.

[0100] For example, images with a low EV (e.g., captured darker than the default 0 EV) are more likely to have more noise in the dark areas. This is because the image sensor tries to be more sensitive to light when the camera (130) tries to compensate for the dark areas, which increases noise. On the other hand, images with a high EV (e.g., captured brighter than the default 0 EV) are less likely to have noise.

[0101] Accordingly, the weighting information, i.e., the first map, can be a feature that includes noise characteristics for multiple raw images. Since the first map includes noise characteristics for each frame, the level of noise in the synthesized result can be predicted, and based on this, the level of noise can be adjusted to add effective noise. Accordingly, when noise information is generated based on the first map including weighting information, noise suitable for the output image can be generated.

[0102] According to one embodiment of the present disclosure, an electronic device (100) may acquire noise information by inputting a first map including weight information, a second map including semantic segmentation information, and an output image into a second artificial intelligence model. In one example, the noise information may include digital grain information. Digital grain may be a small particle-like pattern appearing in a digital image.

[0103] In operation 470, the electronic device (100) may obtain a corrected output image (or final output image) based on the output image and noise information. For example, the noise information may be in the form of a noise image (or noise map) including noise values ​​for each pixel. In one example, the electronic device (100) may obtain the output image through dithering.

[0104] According to an example, the electronic device (100) may obtain a corrected output image by blending a noise image with an output image. For example, the electronic device (100) may blend pixel values ​​of the output image and pixel values ​​included in a noise map through alpha blending. For example, the electronic device (100) may blend pixel values ​​of the output image and pixel values ​​included in a noise map based on the equation “blended image = α × output image + (1-α) × noise image”. The α value may be a weighted sum ratio of the output image and the noise image. For example, the closer α is to 1, the more the output image is reflected, and the closer α is to 0, the more the noise image is reflected. The α value may be a preset value, but may be set / changed according to user input.

[0105] FIG. 5a and FIG. 5b are drawings for explaining a method for obtaining weight information according to an embodiment of the present disclosure.

[0106] According to one embodiment, the electronic device (100) can obtain weight information using the first artificial intelligence model (20). For example, the electronic device (100) can input a plurality of raw images obtained through shooting as described in FIG. 1 into the first artificial intelligence model (20) to obtain an output image synthesized from the plurality of raw images.

[0107] Fig. 5a illustrates an example of an output image (510). For example, the output image (510) may be a somewhat blurred image due to the regression to the mean phenomenon described in Fig. 1.

[0108] FIG. 5b illustrates an example of weight information obtained from the first artificial intelligence model (20), for example, the first map (520).

[0109] For example, the first map (520) may include weighting information for synthesis information for each region of raw frames having different EVs. For example, in the first map (520), a higher proportion of the "A" color may indicate that raw images with low EV values ​​are relatively synthesized, while a higher proportion of the "B" color may indicate that many raw images with high EV values ​​are synthesized.

[0110] For example, an image with a low EV is more likely to have more noise in its dark areas, while an image with a high EV is less likely to have noise. Accordingly, the weighting information, i.e., the first map, can be a feature that includes noise characteristics for multiple raw images. Since the first map includes noise characteristics for each frame, the level of noise in the synthesized result can be predicted, and based on this, the level of noise can be adjusted to add effective noise. Accordingly, when noise information is generated based on the first map including the weighting information, noise suitable for the output image can be generated.

[0111] FIG. 6 is a diagram for explaining a method for obtaining semantic segmentation information according to one embodiment of the present disclosure.

[0112] According to one embodiment, referring to FIG. 6, the electronic device (100) may input an output image (510) into a second artificial intelligence model (610) as illustrated in FIG. 6 to obtain semantic segmentation information (620). According to one example, the semantic segmentation information may be in the form of a map (second map) that includes semantic information for each region of the image.

[0113] For example, the second artificial intelligence model (610) may assign each pixel included in the output image (510) to a specific class. Accordingly, the semantic segmentation information (620) may include class information corresponding to each pixel. For example, the semantic segmentation information (620) may be implemented as a second map in which each pixel includes a corresponding class label. For example, if the class labels are expressed as colors, the semantic segmentation information (620) may be implemented as a second map in which each pixel included in the second map includes a color corresponding to a class such as sky, cloud, building, tree, etc. If the class labels are colors, the electronic device (100) may pre-store class information corresponding to each color. For example, if the class labels are expressed as text, the semantic segmentation information (620) may be implemented as a second map in which each pixel included in the second map includes a text corresponding to a class such as sky, cloud, building, tree, etc.

[0114] FIG. 7 is a diagram for explaining a noise information acquisition method according to an embodiment of the present disclosure.

[0115] According to an embodiment, referring to FIG. 7, the electronic device (100) may input an output image (510), weight information, and semantic segmentation information (second map) (620) into a third artificial intelligence model (710) to obtain noise information (720), as illustrated in FIG. 7. According to an example, the noise information may include digital grain information, as illustrated in FIG. 7. Digital grain may be a small particle pattern appearing in a digital image.

[0116] According to one embodiment of the present disclosure, the electronic device (100) can acquire noise information by adjusting the weight for each region included in the weight information of the output image based on semantic segmentation information for each region included in the output image.

[0117] According to an example, the electronic device (100) may adjust noise information based on semantic segmentation information of the first region and the second region when the first region and the second region included in the output image include pixel values ​​of raw images having relatively low exposure values ​​among a plurality of raw images. For example, when the electronic device (100) identifies a first object included in the first region as an object requiring detail enhancement processing compared to a second object included in the second region based on the semantic segmentation information of the first region and the second region, the electronic device (100) may acquire noise information such that noise in the first region is greater than noise in the second region.

[0118] As an example, the electronic device (100) can obtain noise information by inputting weight information and semantic segmentation information together into the learned third artificial intelligence model (710) as illustrated in FIG. 7.

[0119] Hereinafter, a method for obtaining noise information according to various embodiments will be described with reference to FIGS. 8, 9a, 9b, 10a, 10, 11a, and 11b.

[0120] FIG. 8 is a drawing for explaining an output image processed according to one embodiment of the present disclosure.

[0121] Fig. 8 illustrates an output image (810) according to an example. For example, the output image (810) may be an output image obtained through the first artificial intelligence model (20). For convenience of explanation, an area including the sky and trees in the output image (810) is set as a region of interest (811), and a method for obtaining noise information for the region of interest (811) is described.

[0122] FIGS. 9A and 9B are diagrams illustrating a method for obtaining noise information based on weight information according to an embodiment of the present disclosure. Referring to FIGS. 9A and 9B, FIGS. 9A and 9B are examples of weight information (first maps) corresponding to the output image (810) illustrated in FIG. 8, which may be weight maps of different epochs. Epoch may be the number of training cycles completed for all training data when training a neural network. For example, epoch = 10 may mean that training was performed using the entire data 10 times.

[0123] For example, the weight map (910) illustrated in FIG. 9a may be a map corresponding to epoch = 4980, and the weight map (920) illustrated in FIG. 9b may be a map corresponding to epoch = 12020.

[0124] In the weight maps (910, 920) illustrated in FIGS. 9a and 9b, weight information (911, 912) corresponding to the area of ​​interest (811) is compared.

[0125] For example, in the branch area, the weight map (910) illustrated in FIG. 9A may have a large weight corresponding to a low EV frame, while the weight map (920) illustrated in FIG. 9B may have a large weight corresponding to a high EV frame. In this case, the detail of the branch area may be relatively good in the weight map (920) with a large weight corresponding to a high EV frame.

[0126] The sky region may have a weight map (910) illustrated in FIG. 9a having a large weight corresponding to a high EV frame, while the weight map (920) illustrated in FIG. 9b may have a large weight corresponding to a low EV frame. In this case, noise in the sky region may be greater in the weight map (920) having a large weight corresponding to a low EV frame.

[0127] According to FIGS. 9a and 9b, since the characteristics of noise are reflected in the weight map, different noise information can be obtained based on the weight map.

[0128] FIGS. 10A and 10B, 11A and 11B are diagrams for explaining a method for obtaining noise information based on semantic segmentation information according to an embodiment of the present disclosure.

[0129] Fig. 10a illustrates a region of interest image (1010) of the output image corresponding to epoch = 4980. Referring to Fig. 10a, the weight map (911) of the tree branch region requiring detail may have a larger weight corresponding to a low EV frame. Accordingly, by referring to the semantic segmentation information, noise can be further added to the tree branch region to enhance the detail of the tree branch, as shown in the corrected output image (1020) illustrated in Fig. 10b.

[0130] Figures 11a and 11b illustrate an image (1110) (image 1120 in Figure 11b) of a region of interest (921) of an output image corresponding to epoch = 12020. Assuming that both the tree branch region and the sky region have large weights corresponding to low EV frames, as shown in Figure 11a, if the same noise is added to both the tree branch region and the sky region, the sky region is likely to appear relatively noisy. In this case, by referring to the semantic segmentation information, relatively less noise can be added to the sky region, thereby preventing the sky region from becoming excessively noisy.

[0131] FIG. 12 is a drawing for explaining an example of an output image correction method according to one embodiment of the present disclosure.

[0132] According to an embodiment, referring to FIG. 12, each operation module (1210, 1220, 1230, 1240) illustrated in FIG. 12 may be implemented with at least one software, at least one hardware, and / or a combination thereof. For example, each operation module (1210, 1220, 1230, 1240) may be implemented to utilize a predefined algorithm, a predefined formula, and / or an artificial intelligence model. Each operation module (1210, 1220, 1230, 1240) may be included within the electronic device (100), but may be distributed to at least one external device according to an example.

[0133] According to FIG. 12, the electronic device (100) can input a Multi-Frame RAW file (1201) acquired through shooting into an Image-to-Image AI (artificial intelligence) model (1210) to acquire an output frame (1202).

[0134] A multi-frame RAW file may be a format in which multiple raw frames are saved as a single file. For example, a camera (130) may continuously capture multiple images at very short time intervals according to multi-capture and store them in the memory (120) as a single file. The image-to-image AI model (1210) may be implemented as an image signal processor (ISP) or a software image signal processor (SWISP). The ISP (or SWISP) may be hardware that reduces digital noise generated during photo shooting or performs image processing such as image correction or HDR (High Dynamic Range) after shooting.

[0135] For example, the Image-to-Image AI model (1210) may be an implementation example of the first artificial intelligence model described in operation 430 of FIG. 4. The Image-to-Image AI model (1210) may be a deep learning model that, when a Multi-Frame RAW file is input, performs a task of generating an output image by synthesizing multiple low frames included in the Multi-Frame RAW file. For example, the Image-to-Image AI model (1210) may generate one output frame by selecting only a part with optimal exposure from each of the multiple low frames. For example, the Image-to-Image AI model may be implemented as an AI model that utilizes a deep learning algorithm such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), or Diffusion Models.

[0136] In one example, the electronic device (100) may input the output frame into a semantic segmentation model (1220) to obtain a semantic segmentation map (1203). The semantic segmentation model (1220) may be an implementation example of the second artificial intelligence model described in operation 440 of FIG. 4. In one example, the semantic segmentation model (1220) may classify each pixel included in an image into a specific class (e.g., person, sky, tree, car). In one example, the semantic segmentation map may be in the form of a map that includes semantic information about the specific class (e.g., person, sky, tree, car) assigned to each pixel in the image. The semantic information may be in the form of a label (e.g., color, indicator, flag, text, etc.) that indicates which object or background each pixel belongs to.

[0137] In one example, the electronic device (100) may obtain a weight map (1204) from the image-to-image AI model (1210). For example, the weight map (1204) may be information indicating how much a specific area (e.g., each pixel) of each raw image will contribute to the final composite image when the image-to-image AI model (1210) combines raw images captured under different exposure conditions. For example, the weight map (1204) may be an example of the weight information described in operation 450 of FIG. 4.

[0138] In one example, the electronic device (100) may input the output image (1202), the semantic segmentation map (1203), and the weight map (1204) into the noise generator (1230) to obtain a noise image (1206). For example, the noise image may be in the form of a map including noise values ​​for each pixel. In one example, the noise value may be in the form of digital grain. Digital grain may be a small particle-like pattern appearing in a digital image. For example, the noise image (1206) may be an example of the noise information described in operation 450 of FIG. 4.

[0139] According to an example, the electronic device (100) can obtain a noise image (1206) by additionally inputting at least one of the gain information of the camera (130) and the denoise information of the AP (Application Processor) (1205) as well as the output image (1202), the semantic segmentation map (1203) and the weight map (1204) to the noise generator (1230).

[0140] For example, the gain information of the camera (130) may be a digital camera setting that controls the signal amplification of the camera sensor. The camera (130) may provide an automatic gain (or autogain, AGC) function, and the AGC function may be manually turned on / off. The gain may be performed before and / or after the analog-to-digital converter (ADC). The gain may amplify all signals, including related background noise. Accordingly, when the gain value increases, noise may also be amplified as the signal is amplified. In this case, it may be appropriate to relatively reduce noise when the gain value is high, and conversely, to relatively amplify noise when the gain value is low. For example, if gain information is additionally input to the noise generator (1230), the noise generator (1230) can identify the degree of amplification of previously generated noise (e.g., noise of a weight map or a semantic segmentation map) based on the gain information, so that when the gain value is high, the noise can be relatively reduced, and conversely, when the gain value is low, the noise can be relatively amplified to generate a noise image (1206).

[0141] For example, the denoise information of the AP (application processor) may include the denoise information of the Image-to-Image AI model (1210). The denoise information of the Image-to-Image AI model (1210) may be information corresponding to before / after denoising processing when denoising processing is performed to improve image quality degradation caused by low-light shooting in the Image-to-Image AI model (1210). For example, based on the difference between images before / after processing by the Image-to-Image AI model (1210), 'removed noise' can be calculated according to the denoising processing, and the 'removed noise' calculated in this way can be additionally input to the Noise generator (1230) to obtain a noise image (1206). For example, if denoising information of an AP is additionally input to a noise generator (1230), the noise generator (1230) can identify the degree of denoising based on the denoising information of the AP, so that the generated noise can be strongly added to an area where denoising is well done, and the noise can be weakly added to an area where denoising is not well done, thereby generating a noise image (1206). Meanwhile, if denoising processing is performed by an AP (application processor) other than the Image-to-Image AI model (1210), the corresponding denoising information may be additionally input to the Noise generator (1230).

[0142] According to an example, the electronic device (100) can obtain a Dithered Noise Frame (1207) by blending the output image (1202) and the noise image (1206) using the Blending module (1240) (or the synthesis module). Since the noise addition process is called dithering, the image obtained by blending the output image (1202) and the noise image (1206) is called a Dithered Noise Frame (1207). For example, the Dithered Noise Frame (1207) may be an example of the corrected output image described in operation 460 of FIG. 4. For example, dithering is a technique used in digital image processing to reduce color banding or other visual artifacts in low bit depth images or displays, and since it is a technical problem of the present invention, the final corrected image is called a Dithered Noise Frame (1207). For example, the electronic device (100) may obtain a Dithered Noise Frame (1207) by blending a noise image with an output image. For example, the electronic device (100) may blend pixel values ​​of the output image and pixel values ​​included in a noise map through alpha blending. For example, the Blending module (1240) may blend pixel values ​​of the output image and pixel values ​​included in a noise map based on the equation “Blended image = α × output image + (1-α) × noise image”. The α value may be a weighted sum ratio of the output image and the noise image. For example, the closer α is to 1, the more the output image is reflected, and the closer α is to 0, the more the noise image is reflected. The α value may be a preset value, but may be set / changed according to user input.

[0143] The blending module (1240) can perform blending of the output image (1202) and the noise image (1206) in at least one of the YUV domain and the RGB domain.

[0144] For example, the Blending module (1240) may perform blending by adding the R / G / B values ​​of each pixel included in the noise image (1206) of the RGB domain to the R / G / B values ​​of each pixel included in the output image (1202) of the RGB domain.

[0145] For example, the electronic device (100) may perform noise blending by converting an output image (1202) of an RGB domain and a noise image (1206) of an RGB domain into an output image and a noise image of a YUV domain, respectively. In the YUV domain, Y may include brightness (luminance), and U and V may include chrominance information. For example, the blending module (1240) may perform noise blending by applying different weights to the Y channel, the U channel, and the V channel. For example, the blending module (1240) may add relatively strong noise to the Y channel including brightness information, and add relatively fine noise to the U channel and the V channel. This is because the human eye is sensitive to changes in brightness and less sensitive to changes in color.

[0146] According to one embodiment of the present disclosure, through reverse engineering corresponding to an artificial intelligence model among the binaries of other electronic devices, the shape and data type of the input / output of each model layer, the layer name, structure, and the meaning of the input / output can be confirmed. In addition, by comparing the results of the input / output of the artificial intelligence model, noise added to the output can be identified, and based on this, infringement of the present disclosure can be detected.

[0147] FIG. 13 is a block diagram of an electronic device (1301) within a network environment (1300) according to various embodiments of the present disclosure. The electronic device (1301) may be implemented as the electronic device (100) illustrated in FIG. 3, for example.

[0148] Referring to FIG. 13, in a network environment (1300), an electronic device (1301) may communicate with an external electronic device (1302) via a first network (1398) (e.g., a short-range wireless communication network), or may communicate with at least one of an external electronic device (1304) or a server (1308) via a second network (1399) (e.g., a long-range wireless communication network). According to one embodiment of the present disclosure, the electronic device (1301) may communicate with the external electronic device (1304) via the server (1308). According to one embodiment of the present disclosure, the electronic device (1301) may include a processor (1320), a memory (1330), an input module (1350), an audio output module (1355), a display module (1360), an audio module (1370), a sensor module (1376), an interface (1377), a connection terminal (1378), a haptic module (1379), a camera module (1380), a power management module (1388), a battery (1389), a communication module (1390), a subscriber identification module (1396), or an antenna module (1397). In some embodiments of the present disclosure, at least one of these components (e.g., the connection terminal (1378)) may be omitted, or one or more other components may be added to the electronic device (1301). In some embodiments of the present disclosure, some of these components (e.g., sensor module (1376), camera module (1380), or antenna module (1397)) may be integrated into one component (e.g., display module (1360)).

[0149] The processor (1320) may control at least one other component (e.g., hardware or software component) of the electronic device (1301) connected to the processor (1320) by executing, for example, software (e.g., program (1340)), and may perform various data processing or operations. According to one embodiment of the present disclosure, as at least a part of the data processing or operations, the processor (1320) may store commands or data received from other components (e.g., sensor module (1376) or communication module (1390)) in a volatile memory (1332), process the commands or data stored in the volatile memory (1332), and store result data in a non-volatile memory (1334). According to one embodiment of the present disclosure, the processor (1320) may include a main processor (1321) (e.g., a central processing unit or an application processor) or an auxiliary processor (1323) (e.g., a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor) that can operate independently or together with the main processor (1321). For example, when the electronic device (1301) includes the main processor (1321) and the auxiliary processor (1323), the auxiliary processor (1323) may be configured to use lower power than the main processor (1321) or to be specialized for a given function. The auxiliary processor (1323) may be implemented separately from the main processor (1321) or as a part thereof.

[0150] The auxiliary processor (1323) may control at least a part of functions or states associated with at least one component (e.g., a display module (1360), a sensor module (1376), or a communication module (1390)) of the electronic device (1301), for example, on behalf of the main processor (1321) while the main processor (1321) is in an inactive (e.g., sleep) state, or together with the main processor (1321) while the main processor (1321) is in an active (e.g., application execution) state. According to one embodiment of the present disclosure, the auxiliary processor (1323) (e.g., an image signal processor or a communication processor) may be implemented as a part of another functionally related component (e.g., a camera module (1380) or a communication module (1390)). According to one embodiment of the present disclosure, the auxiliary processor (1323) (e.g., a neural network processing device) may include a hardware structure specialized for processing artificial intelligence models. An AI model can be generated through machine learning. This learning can be performed, for example, on the electronic device (1301) where the AI ​​model is executed, or through a separate server (e.g., server (1308)). The learning algorithm may include, for example, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, but is not limited to the examples described above. The AI ​​model may include multiple artificial neural network layers.The artificial neural network may be one of a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a deep Q-network, or a combination of two or more of the above, but is not limited to the examples described above. In addition to, or alternatively to, a hardware structure, an artificial intelligence model may include a software structure.

[0151] The memory (1330) can store various data used by at least one component (e.g., the processor (1320) or the sensor module (1376)) of the electronic device (1301). The data can include, for example, software (e.g., the program (1340)) and input data or output data for commands related thereto. The memory (1330) can include volatile memory (1332) or non-volatile memory (1334).

[0152] The program (1340) may be stored as software in the memory (1330) and may include, for example, an operating system (1442), middleware (1444), or an application (1446). The non-volatile memory 1334 may include built-in memory 1336 or external memory 1338.

[0153] The input module (1350) can receive commands or data to be used in a component of the electronic device (1301) (e.g., a processor (1320)) from an external source (e.g., a user) of the electronic device (1301). The input module (1350) can include, for example, a microphone, a mouse, a keyboard, a key (e.g., a button), or a digital pen (e.g., a stylus pen).

[0154] The audio output module (1355) can output audio signals to the outside of the electronic device (1301). The audio output module (1355) can include, for example, a speaker or a receiver. The speaker can be used for general purposes, such as multimedia playback or recording playback. The receiver can be used to receive incoming calls. According to one embodiment of the present disclosure, the receiver can be implemented separately from the speaker or as part of the speaker.

[0155] The display module (1360) can visually provide information to an external party (e.g., a user) of the electronic device (1301). The display module (1360) may include, for example, a display, a holographic device, or a projector, and a control circuit for controlling the device. According to one embodiment, the display module (1360) may include a touch sensor configured to detect a touch, or a pressure sensor configured to measure the intensity of a force generated by the touch.

[0156] The audio module (1370) can convert sound into an electrical signal, or vice versa, convert an electrical signal into sound. According to one embodiment of the present disclosure, the audio module (1370) can acquire sound through the input module (1350), output sound through the sound output module (1355), or an external electronic device (e.g., electronic device (1302)) (e.g., speaker or headphone) directly or wirelessly connected to the electronic device (1301).

[0157] The sensor module (1376) can detect the operating status (e.g., power or temperature) of the electronic device (1301) or the external environmental status (e.g., user status) and generate an electrical signal or data value corresponding to the detected status. According to one embodiment of the present disclosure, the sensor module (1376) can include, for example, a gesture sensor, a gyro sensor, a barometric pressure sensor, a magnetic sensor, an acceleration sensor, a grip sensor, a proximity sensor, a color sensor, an IR (infrared) sensor, a biometric sensor, a temperature sensor, a humidity sensor, or an illuminance sensor.

[0158] The interface (1377) may support one or more designated protocols that may be used to directly or wirelessly connect the electronic device (1301) with an external electronic device (e.g., the electronic device (1302)). In one embodiment, the interface (1377) may include, for example, a high definition multimedia interface (HDMI), a universal serial bus (USB) interface, an SD card interface, or an audio interface.

[0159] The connection terminal (1378) may include a connector through which the electronic device (1301) may be physically connected to an external electronic device (e.g., an external electronic device (1302)). According to one embodiment of the present disclosure, the connection terminal (1378) may include, for example, an HDMI connector, a USB connector, an SD card connector, or an audio connector (e.g., a headphone connector).

[0160] The haptic module (1379) can convert electrical signals into mechanical stimuli (e.g., vibration or movement) or electrical stimuli that a user can perceive through tactile or kinesthetic sensations. According to one embodiment of the present disclosure, the haptic module (1379) can include, for example, a motor, a piezoelectric element, or an electrical stimulation device.

[0161] The camera module (1380) can capture still images and videos. According to one embodiment, the camera module (1380) may include one or more lenses, image sensors, image signal processors, or flashes.

[0162] The power management module (1388) can manage power supplied to the electronic device (1301). According to one embodiment, the power management module (1388) can be implemented, for example, as at least a part of a power management integrated circuit (PMIC).

[0163] A battery (1389) may power at least one component of the electronic device (1301). In one embodiment, the battery (1389) may include, for example, a non-rechargeable primary battery, a rechargeable secondary battery, or a fuel cell.

[0164] The communication module (1390) may support the establishment of a direct (e.g., wired) communication channel or a wireless communication channel between the electronic device (1301) and an external electronic device (e.g., external electronic device (1302), external electronic device (1304), or server (1308)), and the performance of communication through the established communication channel. The communication module (1390) may operate independently from the processor (1320) (e.g., application processor) and may include one or more communication processors that support direct (e.g., wired) communication or wireless communication. According to one embodiment of the present disclosure, the communication module (1390) may include a wireless communication module (1392) (e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module) or a wired communication module (1494) (e.g., a local area network (LAN) communication module, or a power line communication module). Among these communication modules, the above-mentioned communication module can communicate with an external electronic device (1304) via a first network (1398) (e.g., a short-range communication network such as Bluetooth, wireless fidelity (WiFi) direct, or infrared data association (IrDA)) or a second network (1399) (e.g., a long-range communication network such as a legacy cellular network, a 5G network, a next-generation communication network, the Internet, or a computer network (e.g., a LAN or WAN)). These various types of communication modules can be integrated into a single component (e.g., a single chip) or implemented as multiple separate components (e.g., multiple chips). The wireless communication module (1392) can verify or authenticate the electronic device (1301) within a communication network such as the first network (1398) or the second network (1399) by using subscriber information (e.g., an international mobile subscriber identity (IMSI)) stored in the subscriber identification module (1396).

[0165] The wireless communication module (1392) can support the fourth generation (5G network after the 4G network) and next-generation communication technology, for example, NR access technology (new radio access technology). The NR access technology can support high-speed transmission of high-capacity data (eMBB (enhanced mobile broadband)), minimization of terminal power and connection of multiple terminals (mMTC (massive machine type communications)), or high reliability and low latency (URLLC (ultra-reliable and low-latency communications)). The wireless communication module (1392) can support, for example, a high-frequency band (e.g., millimeter wave (mmWave) band) to achieve a high data transmission rate. The wireless communication module (1392) can support various technologies for securing performance in a high-frequency band, for example, beamforming, massive multiple-input and multiple-output (MIMO), full-dimensional multiple-input and multiple-output (FD-MIMO), and array antenna. The wireless communication module (1392) may support technologies such as antenna, analog beam-forming, or large scale antenna. The wireless communication module (1392) may support various requirements specified in the electronic device (1301), an external electronic device (e.g., the external electronic device (1304)), or a network system (e.g., the second network (1399)). According to one embodiment of the present disclosure, the wireless communication module (1392) may support a peak data rate (e.g., 20 Gbps or more) for eMBB realization, a loss coverage (e.g., 164 dB or less) for mMTC realization, or a U-plane latency (e.g., 0 for downlink (DL) and uplink (UL) respectively) for URLLC realization.It can support latency of 5ms or less, or round trip of 1ms or less.

[0166] The antenna module (1397) can transmit or receive signals or power to or from an external device (e.g., an external electronic device). According to one embodiment of the present disclosure, the antenna module (1397) may include an antenna including a radiator formed of a conductor or a conductive pattern formed on a substrate (e.g., a printed circuit board (PCB)). According to one embodiment of the present disclosure, the antenna module (1397) may include a plurality of antennas (e.g., an array antenna). In this case, at least one antenna suitable for a communication method used in a communication network, such as the first network (1398) or the second network (1399), may be selected from the plurality of antennas by, for example, the communication module (1390). A signal or power may be transmitted or received between the communication module (1390) and an external electronic device via the selected at least one antenna. According to some embodiments of the present disclosure, in addition to the radiator, other components (e.g., a radio frequency integrated circuit (RFIC)) may be additionally formed as part of the antenna module (1397).

[0167] According to various embodiments of the present disclosure, the antenna module (1397) may form a mmWave antenna module. According to one embodiment of the present disclosure, the mmWave antenna module may include a printed circuit board, an RFIC disposed on or adjacent to a first surface (e.g., a bottom surface) of the printed circuit board and capable of supporting a designated high-frequency band (e.g., a mmWave band), and a plurality of antennas (e.g., an array antenna) disposed on or adjacent to a second surface (e.g., a top surface or a side surface) of the printed circuit board and capable of transmitting or receiving signals of the designated high-frequency band.

[0168] At least some of the above components can be interconnected and exchange signals (e.g., commands or data) with each other via a communication method between peripheral devices (e.g., a bus, GPIO (general purpose input and output), SPI (serial peripheral interface), or MIPI (mobile industry processor interface)).

[0169] According to one embodiment, commands or data may be transmitted or received between the electronic device (1301) and an external electronic device (1304) via a server (1308) connected to a second network (1399). Each of the external electronic devices (1302, 1304, or server 1308) may be the same or a different type of device as the electronic device (1301). According to one embodiment of the present disclosure, all or part of operations executed in the electronic device (1301) may be executed in one or more of the external electronic devices (1302, 1304, or 1308). For example, when the electronic device (1301) is to perform a certain function or service automatically or in response to a request from a user or another device, the electronic device (1301) may, instead of or in addition to executing the function or service by itself, request one or more external electronic devices to perform the function or at least a part of the service. One or more external electronic devices that receive the request may execute at least a portion of the requested function or service, or an additional function or service related to the request, and transmit the result of the execution to the electronic device (1301). The electronic device (1301) may process the result as is or additionally and provide it as at least a portion of a response to the request. For this purpose, cloud computing, distributed computing, mobile edge computing (MEC), or client-server computing technology may be used, for example. The electronic device (1301) may provide an ultra-low latency service using, for example, distributed computing or mobile edge computing. In another embodiment of the present disclosure, the external electronic device (1304) may include an Internet of Things (IoT) device. The server (1308) may be an intelligent server utilizing machine learning and / or a neural network.According to one embodiment of the present disclosure, an external electronic device (1304) or server (1308) may be included within a second network (1399). The electronic device (1301) may be applied to intelligent services (e.g., smart homes, smart cities, smart cars, or healthcare) based on 5G communication technology and IoT-related technology.

[0170] According to one embodiment of the present disclosure, an electronic device (100) includes: a memory (120) for storing instructions; a camera (130); And at least one processor (110) including processing circuitry; wherein the instructions, when individually or collectively executed by the at least one processor (110), cause the electronic device (100) to, when a photographing event is identified, acquire a plurality of raw images having different exposure values ​​through the camera (130), acquire an output image synthesized using the plurality of raw images, acquire weight information for each area related to a degree of synthesis of the plurality of raw images used to acquire the output image, acquire semantic segmentation information corresponding to the output image, acquire noise information based on the semantic segmentation and the weight information, and acquire a corrected output image based on the output image and the noise information.

[0171] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause the electronic device to synthesize the plurality of raw images based on exposure values ​​of the plurality of raw images using an image conversion module to obtain the output image.

[0172] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, cause the electronic device to input the plurality of raw images into a first artificial intelligence model included in the image conversion module to obtain the de-noised output image, wherein the weight information may be information including noise characteristics of the plurality of raw images.

[0173] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause the electronic device to obtain a first map including the weight information from the image conversion module, obtain a second map including the semantic segmentation information using a second artificial intelligence model, obtain a third map including the noise information based on the output image, the first map, and the second map, and obtain the corrected output image by blending the output image and the third map.

[0174] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, cause the electronic device to acquire the noise information by adjusting a weight for each region of the output image based on semantic segmentation information for each region included in the output image, wherein the semantic segmentation information may include class information of an object corresponding to a region of the image.

[0175] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause the electronic device to adjust noise of the first region and the second region based on semantic segmentation information of the first region and the second region when the first region and the second region included in the output image include pixel values ​​of a raw image having a relatively low exposure value among the plurality of raw images.

[0176] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause noise information to be acquired such that noise in the first area is greater than noise in the second area, when a first object included in the first area is an object that requires detail enhancement processing compared to a second object included in the second area.

[0177] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause the electronic device to input the output image, the semantic segmentation information, and the weight information into a third artificial intelligence model to obtain the noise information.

[0178] According to one embodiment of the present disclosure, the instructions, when individually or collectively executed by the at least one processor, may cause the electronic device to additionally input at least one of gain information of the camera and denoise information of the image conversion module into the third artificial intelligence model to obtain the noise information.

[0179] According to one embodiment of the present disclosure, the noise information may include digital grain information.

[0180] A method for controlling an electronic device according to an embodiment of the present disclosure includes, when a shooting event is identified, an operation of acquiring a plurality of raw images having different exposure values ​​through the camera; an operation of acquiring an output image using the plurality of raw images; an operation of acquiring weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image; an operation of acquiring semantic segmentation information corresponding to the output image; an operation of acquiring noise information based on the semantic segmentation information and the weight information; and an operation of acquiring a corrected output image based on the output image and the noise information.

[0181] According to one embodiment of the present disclosure, the operation of obtaining the output image may include an operation of obtaining the output image by synthesizing the plurality of raw images based on exposure values ​​of the plurality of raw images using an image conversion module.

[0182] According to one embodiment of the present disclosure, the operation of obtaining the output image includes an operation of inputting the plurality of raw images into a first artificial intelligence model included in an image conversion module to obtain the de-noised output image; and the weight information may be information including noise characteristics of the plurality of raw images.

[0183] According to one embodiment of the present disclosure, the operation of obtaining the corrected output image may include: obtaining a first map including the weight information from the image transformation module; obtaining a second map including the semantic segmentation information using a second artificial intelligence model; obtaining a third map including the noise information based on the output image, the first map, and the second map; and obtaining the corrected output image by blending the output image and the third map.

[0184] According to one embodiment of the present disclosure, the operation of obtaining the noise information includes an operation of obtaining the noise information by adjusting a weight for each region of the output image based on semantic segmentation information for each region included in the output image; wherein the semantic segmentation information may include class information of an object corresponding to the region of the image.

[0185] According to one embodiment of the present disclosure, the operation of obtaining the noise information may include an operation of adjusting noise of the first region and the second region based on semantic segmentation information of the first region and the second region when the first region and the second region included in the output image include pixel values ​​of a raw image having a relatively low exposure value among the plurality of raw images.

[0186] According to one embodiment of the present disclosure, the operation of obtaining the noise information may include an operation of obtaining the noise information such that noise in the first area is greater than noise in the second area, when the first object included in the first area is an object that requires detail enhancement processing compared to the second object included in the second area.

[0187] According to one embodiment of the present disclosure, the operation of obtaining the noise information may include an operation of obtaining the noise information by inputting the output image, the semantic segmentation information, and the weight information into a third artificial intelligence model.

[0188] According to one embodiment of the present disclosure, the operation of obtaining the noise information may include an operation of additionally inputting at least one of gain information of the camera and denoise information of the image conversion module into the third artificial intelligence model to obtain the noise information.

[0189] In accordance with one embodiment of the present disclosure, a non-transitory computer-readable medium storing computer instructions that, when executed by a processor of an electronic device, cause the electronic device to perform an operation, the operation includes: when a photographing event is identified, acquiring a plurality of raw images having different exposure values ​​through the camera; acquiring an output image using the plurality of raw images; acquiring weight information for each region related to a degree of synthesis of the plurality of raw images used to acquire the output image; acquiring semantic segmentation information corresponding to the output image; acquiring noise information based on the semantic segmentation information and the weight information; and acquiring a corrected output image based on the output image and the noise information.

[0190] According to the various embodiments described above, the quality of an image can be improved by performing dithering by generating adaptive noise according to image characteristics.

[0191] Although the various embodiments described above have been described as utilizing multiple individual artificial intelligence models, the operations of at least two artificial intelligence models among the multiple neural network models may be implemented in a single artificial intelligence model.

[0192] Each operation according to the various embodiments described above may be performed by the processor (110), but if necessary, a module for each operation may be utilized. For example, each module may be implemented using at least one software, at least one hardware, and / or a combination thereof. Each module may be implemented to utilize a predefined algorithm, a predefined formula, and / or a learned artificial intelligence model to perform the operation. However, at least some modules may be distributed to an external device.

[0193] The methods according to the various embodiments of the present disclosure described above may be implemented in the form of applications installable on existing electronic devices. Alternatively, the methods according to the various embodiments of the present disclosure described above may be performed using a deep learning-based artificial neural network (or deep artificial neural network), i.e., a learning network model.

[0194] The methods according to the various embodiments of the present disclosure described above can be implemented only with a software upgrade or a hardware upgrade for an existing electronic device.

[0195] The various embodiments of the present disclosure described above can also be performed through an embedded server provided in an electronic device, or an external server of the electronic device.

[0196] According to an exemplary embodiment of the present disclosure, the various embodiments described above may be implemented as software including instructions stored in a machine-readable storage medium that can be read by a machine (e.g., a computer). The device may include an electronic device (e.g., electronic device A) according to the disclosed embodiments, which is a device that can call instructions stored in the storage medium and operate according to the called instructions. When an instruction is executed by a processor, the processor may directly or under the control of the processor perform a function corresponding to the instruction using other components. The instruction may include code generated or executed by a compiler or interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' means that the storage medium does not contain a signal and is tangible, but does not distinguish between data being stored semi-permanently or temporarily in the storage medium.

[0197] Furthermore, according to one embodiment of the present disclosure, the method according to the various embodiments described above may be provided as included in a computer program product. The computer program product may be traded as a product between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or online through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily generated in a storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0198] In addition, each of the components (e.g., modules or programs) according to the various embodiments described above may be composed of a single or multiple entities, and some of the corresponding sub-components described above may be omitted, or other sub-components may be further included in various embodiments. Alternatively or additionally, some components (e.g., modules or programs) may be integrated into a single entity, which may perform the same or similar functions as those performed by each of the corresponding components prior to integration. Operations performed by modules, programs or other components according to various embodiments of the present disclosure may be executed sequentially, in parallel, iteratively or heuristically, or at least some operations may be executed in a different order, omitted, or other operations may be added.

[0199] It will be appreciated that the various embodiments according to the claims and description of the present disclosure may be implemented in the form of hardware, software, or a combination of hardware and software.

[0200] Such software may be stored on a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium stores one or more computer programs (software modules), which may include computer-executable instructions that, when executed by one or more processors of an electronic device, cause the electronic device to perform the method of the present invention.

[0201] Such software may be stored in a volatile or non-volatile storage form, for example, a storage device such as read-only memory (ROM), whether or not erasable or rewritable, or a memory form such as random access memory (RAM), a memory chip, device, integrated circuit, or an optically or magnetically readable medium such as a compact disc (CD), a digital video disc (DVD), a magnetic disk, or a magnetic tape. The storage device and the storage medium may be embodiments of various non-transitory machine-readable storage media suitable for storing a computer program or instructions for executing the same. Accordingly, various embodiments may provide a program comprising code for implementing an apparatus or method as recited in any one of the claims of the present specification, and a non-transitory machine-readable storage medium storing such a program.

[0202] While the invention described in this disclosure has been described and illustrated with reference to various embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the present disclosure as defined by the appended claims and their equivalents.

Claims

1. In an electronic device (100), One or more storage media, memory (120) for storing instructions; Camera (130); and At least one processor (110) connected to the memory and the camera so as to be communicatively connected, and including processing circuitry; The above instructions, when individually or collectively executed by the at least one processor (110), cause the electronic device (100), When a shooting event is identified, multiple raw images with different exposure values ​​are acquired through the camera (130), Obtaining an output image using the above multiple raw images, Obtain weight information for each area related to the degree of synthesis of the plurality of raw images used to obtain the output image, Obtain semantic segmentation information corresponding to the above output image, Obtaining noise information based on the above semantic segmentation information and the above weight information, An electronic device that obtains a corrected output image based on the output image and the noise information.

2. In paragraph 1, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: By using an image conversion module, the plurality of raw images are synthesized based on the exposure values ​​of the plurality of raw images to obtain the output image, Electronic devices.

3. In paragraph 2, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: The above plurality of raw images are input into the first artificial intelligence model included in the image conversion module to obtain the de-noised output image, The above weight information is, An electronic device, which is information including noise characteristics of the plurality of raw images.

4. In paragraph 2, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: Obtain a first map including the weight information from the image conversion module, Obtaining a second map including the semantic segmentation information using a second artificial intelligence model, Obtaining a third map including the noise information based on the output image, the first map, and the second map, An electronic device that obtains the corrected output image by blending the output image and the third map.

5. In paragraph 1, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: The noise information is obtained by adjusting the weight for each region of the output image based on the semantic segmentation information for each region included in the output image. The above semantic segmentation information is, An electronic device including class information of an object corresponding to an area of ​​the image.

6. In paragraph 5, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: An electronic device that adjusts noise of the first region and the second region based on semantic segmentation information of the first region and the second region when the first region and the second region included in the output image include pixel values ​​of a raw image having a relatively low exposure value among the plurality of raw images.

7. In paragraph 6, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: An electronic device that obtains noise information in which the noise in the first area is greater than the noise in the second area, when the first object included in the first area is an object that requires detail enhancement processing compared to the second object included in the second area.

8. In paragraph 2, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: An electronic device that inputs the output image, the semantic segmentation information, and the weight information into a third artificial intelligence model to obtain the noise information.

9. In paragraph 8, The above instructions, when individually or collectively executed by the at least one processor, cause the electronic device to: An electronic device that additionally inputs at least one of the gain information of the camera and the denoise information of the image conversion module into the third artificial intelligence model to obtain the noise information.

10. In paragraph 1, The above noise information is, An electronic device containing digital grain information.

11. In a method for controlling an electronic device, When a shooting event is identified, an operation of acquiring multiple raw images with different exposure values ​​through the camera; An operation of obtaining an output image using the above plurality of raw images; An operation of obtaining weight information for each area related to the degree of synthesis of the plurality of raw images used to obtain the output image; An operation of obtaining semantic segmentation information corresponding to the above output image; An operation of obtaining noise information based on the semantic segmentation information and the weight information; and A control method comprising: an operation of obtaining a corrected output image based on the output image and the noise information.

12. In paragraph 11, The operation of obtaining the above output image is as follows: An operation of obtaining the output image by synthesizing the plurality of raw images based on the exposure values ​​of the plurality of raw images using an image conversion module; Control method.

13. In paragraph 12, The operation of obtaining the above output image is as follows: An operation of inputting the plurality of raw images into a first artificial intelligence model included in the image conversion module to obtain the de-noised output image; The above weight information is, A control method, which is information including noise characteristics of the plurality of raw images.

14. In paragraph 12, The operation of obtaining the above-mentioned corrected output image is as follows: An operation of obtaining a first map including the weight information from the image conversion module; An operation of obtaining a second map including the semantic segmentation information using a second artificial intelligence model; An operation of obtaining a third map including the noise information based on the output image, the first map, and the second map; and A control method comprising: an operation of blending the output image and the third map to obtain the corrected output image.

15. One or more non-transitory computer-readable storage media storing one or more computer programs comprising computer-executable instructions that, when executed individually or collectively by one or more processors, cause an electronic device to perform an operation, The above action is, When a shooting event is identified, the action of acquiring multiple raw images with different exposure values ​​through the camera; An operation of obtaining an output image synthesized from the plurality of raw images using an image conversion module; An operation of obtaining weight information reflecting the noise characteristics of multiple raw images used in the above synthesis; An operation of obtaining noise information based on semantic segmentation information and weight information corresponding to the output image; and A non-transitory computer-readable medium comprising: an operation of obtaining a corrected output image based on the output image and the noise information.

Images