Electronic device and method for processing image

Abstract

A method of the present disclosure may comprise the steps of: acquiring a plurality of images including an object; acquiring, for respective images, a body score, which is a score quantifying the importance of the pose of the object included in the images; acquiring a base image from among the plurality of images on the basis of the body score; acquiring one image from among the plurality of images as the highlight image for a first body part of the object; and correcting an area of the first body part in the base image on the basis of the highlight image.

Description

Electronic device and method for processing images

[0001] The present invention relates to an electronic device and method for processing images, wherein a method is provided for analyzing the pose of an object within an image to acquire an image of an object performing an important action or to combine images so that an object performs an important action.

[0002] Recently, due to the rapid development of communication technology, the functions of mobile terminals are gradually expanding, and more diverse User Interfaces (UI) are being provided. In order to increase the utility value of mobile terminals and satisfy the diverse needs of users, various applications that can be executed on mobile terminals are being developed.

[0003] In particular, as user interest in photos and videos increases, most mobile devices now offer digital camera functions. Users take photos using these digital camera features, but they also want to capture important moments, such as the moment a person jumps, a baby smiles, or a pet looks at the camera. While users can manually search for the moment a subject performing a significant action was captured after recording a video, this can be somewhat cumbersome.

[0004] A method disclosed as a technical means for achieving a technical task may include the step of acquiring a plurality of images containing an object. The method may include the step of acquiring a body score for each image, which is a numerical score representing the importance of the pose of the object included in the plurality of images. The method may include the step of acquiring a base image among the plurality of images based on the body score. The method may include the step of acquiring one of the plurality of images as a highlight image for a first body part of the object. The method may include the step of correcting the area of ​​the first body part within the base image based on the highlight image.

[0005] An electronic device disclosed as a technical means for achieving a technical task may include an input / output interface for receiving user input requesting image processing and outputting an image processed according to the user input. The electronic device may include a memory in which commands for processing an image are stored. The electronic device may include at least one processor. By executing a program stored in memory or at least one instruction by the at least one processor, the electronic device may acquire a plurality of images containing an object. The electronic device may acquire a body score for each image, which is a numerical score representing the importance of the pose of an object included in the plurality of images. Based on the body score, the electronic device may acquire a base image among the plurality of images. The electronic device may acquire one of the plurality of images as a highlight image for a first body part of the object. Based on the highlight image, the electronic device may correct the area of ​​the first body part within the base image.

[0006] A computer-readable recording medium disclosed as a technical means for achieving a technical task may store a program for executing at least one of the embodiments of the disclosed method on a computer.

[0007] FIG. 1a is a conceptual diagram illustrating a method for analyzing a plurality of images obtained according to one embodiment of the present disclosure.

[0008] FIG. 1b is a conceptual diagram illustrating a method of combining images based on a plurality of images analyzed according to one embodiment of the present disclosure.

[0009] FIG. 2 is a flowchart illustrating the overall flow of a method for generating a best image according to one embodiment of the present disclosure.

[0010] FIG. 3 is a flowchart illustrating a method for generating a best image according to one embodiment of the present disclosure.

[0011] FIG. 4 is a conceptual diagram illustrating a method for extracting feature points of an object in an image according to one embodiment of the present disclosure.

[0012] FIG. 5 is a conceptual diagram illustrating a method for evaluating the pose of an object in an image based on feature points according to one embodiment of the present disclosure.

[0013] FIG. 6 is a conceptual diagram illustrating a method for evaluating the completeness of a shot based on the composition of an image according to one embodiment of the present disclosure.

[0014] FIG. 7a is a drawing illustrating an exemplary composition that can be implemented in an image according to one embodiment of the present disclosure.

[0015] FIG. 7b is a drawing illustrating an exemplary composition that can be implemented in an image according to one embodiment of the present disclosure.

[0016] FIG. 8 is a flowchart illustrating a method for evaluating the pose of an object in an image according to one embodiment of the present disclosure.

[0017] FIG. 9 is a flowchart illustrating a method for determining a base image among a plurality of images according to one embodiment of the present disclosure.

[0018] FIG. 10 is a conceptual diagram illustrating a method for evaluating the visual completeness of body parts of an object in a plurality of images according to one embodiment of the present disclosure.

[0019] FIG. 11 is a flowchart illustrating a method for determining a highlight image of a first body part of an object among a plurality of images according to one embodiment of the present disclosure.

[0020] FIG. 12 is a flowchart illustrating a method for correcting a base image according to one embodiment of the present disclosure.

[0021] FIG. 13 is a flowchart illustrating a method for extracting a piece image to be inpainted onto a base image according to one embodiment of the present disclosure.

[0022] FIG. 14 is a flowchart illustrating a method for generating a best image according to one embodiment of the present disclosure.

[0023] FIG. 15 is a block diagram illustrating the configuration of an electronic device according to one embodiment of the present disclosure.

[0024] In describing the present disclosure, technical details that are well known in the technical field to which the present disclosure belongs and are not directly related to the present disclosure are omitted. This is intended to convey the essence of the present disclosure more clearly without obscuring it by omitting unnecessary explanations. Furthermore, the terms described below are defined considering their functions within the present disclosure, and these definitions may vary depending on the intentions or practices of the user or operator. Therefore, their definitions should be based on the content throughout this specification.

[0025] For the same reason, some components in the attached drawings have been exaggerated, omitted, or schematically depicted. Additionally, the dimensions of each component do not entirely reflect their actual dimensions. Identical or corresponding components in each drawing have been assigned the same reference numbers.

[0026] The advantages and features of the present disclosure, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but may be implemented in various different forms. The disclosed embodiments are provided to ensure that the disclosure of the present disclosure is complete and to fully inform those skilled in the art of the scope of the disclosure. An embodiment of the present disclosure may be defined according to the claims. Throughout the specification, the same reference numerals indicate the same components. Furthermore, in describing an embodiment of the present disclosure, if it is determined that a detailed description of a related function or configuration might unnecessarily obscure the essence of the present disclosure, such detailed description is omitted. Additionally, terms described below are defined considering their functions in the present disclosure, and these may vary depending on the intentions or conventions of the user or operator. Therefore, their definitions should be based on the content throughout the specification.

[0027] In one embodiment, each block of the flowcharts and combinations of the flowcharts may be executed by computer program instructions. Computer program instructions may be loaded onto a processor of a general-purpose computer, a computer for special purposes, or other programmable data processing equipment, and the instructions executed through the processor of the computer or other programmable data processing equipment may create means for performing the functions described in the flowchart block(s). Computer program instructions may also be stored in computer-available or computer-readable memory that can be directed toward the computer or other programmable data processing equipment to implement functions in a specific manner, and instructions stored in computer-available or computer-readable memory may produce a manufactured item containing instruction means for performing the functions described in the flowchart block(s). Computer program instructions may also be loaded onto a computer or other programmable data processing equipment.

[0028] Additionally, each block of the flowchart may represent a module, segment, or part of code containing one or more executable instructions for executing a specified logical function(s). In one embodiment, the functions mentioned in the blocks may occur out of order. For example, two blocks shown in succession may be executed substantially simultaneously or in reverse order depending on the function.

[0029] In one embodiment of the present disclosure, the term “part” used may refer to software or hardware components such as a Field Programmable Gate Array (FPGA) or an Application Specific Integrated Circuit (ASIC), and the “part” may perform a specific role. Meanwhile, the “part” is not limited to software or hardware. The “part” may be configured to reside in an addressable storage medium or may be configured to run one or more processors. In one embodiment, the “part” may include components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. Functions provided through a specific component or a specific “part” may be combined or separated into additional components to reduce their number. Additionally, in one embodiment, the “part” may include one or more processors.

[0030] The meanings of the terms used in the present disclosure are explained below.

[0031] In the present disclosure, the base image may be an image selected as a basis among a plurality of images. The base image may be selected among the plurality of images as an image with a high degree of completeness in the capture or as an image where the importance of the pose taken by an object within the image is high. The base image may be an image requiring minimal correction. For example, a method according to one embodiment of the present disclosure may replace a part of the base image with another image based on the base image. A part of the base image, for example, a part of the body part of the person within the base image, may be replaced so as to emphasize the pose of the person within the base image. The term "base image" is used simply to mean an image used as a basis for image processing, and may be replaced with various terms such as reference image or basic image.

[0032] In the present disclosure, a highlight image may be an image selected from a plurality of images to be inserted into a portion of a base image. The highlight image may be an image selected for a specific body part of an object within the image. For example, a highlight image for the head of an object may be selected from a plurality of images. A method according to one embodiment may correct a base image by replacing the head area of ​​an object within the highlight image with the head area of ​​an object within the base image.

[0033] In the present disclosure, a fragment image may be a part of a highlight image necessary to correct a portion of a base image. For example, assume a case where a highlight image of the head portion of an object in an image is selected to correct the head portion of an object in a base image. An electronic device may extract a fragment image limited to the head portion of the object among the highlight images. The fragment image may be a part of the highlight image.

[0034] In the present disclosure, the best image may be an image finally obtained by correcting a base image based on a highlight image. The best image may be a base image corrected based on a highlight image. For convenience of explanation, the best image and the base image are distinguished, but the best image may be a base image corrected based on a highlight image, or it may be an image newly generated based on a highlight image and a base image.

[0035] In the present disclosure, a keypoint refers to a point within an image that is distinguishable from or easily identifiable from the surrounding background, and corresponds to a key point of the body. For example, a keypoint for a hand may include points corresponding to a plurality of joints included in the hand. A keypoint may be expressed as a three-dimensional positional coordinate value, which is positional information regarding the x-axis, y-axis, and z-axis of a key point of the body.

[0036] Embodiments of the present disclosure are described below with reference to the attached drawings so that those skilled in the art can easily implement them. However, the present disclosure may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present disclosure in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification have been given similar reference numerals. Also, the reference numerals used in each drawing are for the purpose of explaining each drawing, and different reference numerals used in different drawings are not intended to indicate different elements. The present disclosure will be described in detail below with reference to the attached drawings.

[0037] FIG. 1a is a conceptual diagram illustrating a method for analyzing a plurality of images obtained according to one embodiment of the present disclosure.

[0038] Referring to FIG. 1a, an electronic device may acquire a plurality of images (100). The plurality of images (100) may include a first to eighth image (100a, 100b, 100c, 100d, 100e, 100f, 100g, 100h). FIG. 1a illustrates only eight images, and this does not limit the technical scope of the present disclosure.

[0039] In one embodiment, the plurality of images (100) may be images that are continuously captured over a set period. The plurality of images (100) may include an image sequence, which refers to a series of sets of images arranged in order according to the flow of time. For example, the plurality of images (100) may include at least one of a set of images captured over a set period, a motional image composed of images captured over a set period, and a video.

[0040] In one embodiment, the plurality of images (100) may be images containing objects. The objects may include not only people but also living organisms. For example, the plurality of images (100) may be images containing people. As another example, the plurality of images (100) may be images containing dogs.

[0041] In one embodiment, a plurality of images (100), which is an image sequence, may be images containing the same object. Of course, some of the plurality of images (100) may not contain the object, but the image sequence of the plurality of images (100) may be images with the same object as the primary target.

[0042] In one embodiment, the plurality of images (100) may be images capturing the appearance of an object changing over time. Over time, some of the plurality of images (100) may be images containing the object. For example, over time, the object may enter or exit the shooting area. One of the plurality of images may contain the object, and another of the plurality of images may not contain the object. For convenience of explanation, in FIG. 1a, the plurality of images are all depicted as containing the object.

[0043] The user may want to capture a moment when an object assumes an important pose, such as when the object jumps, when a baby smiles, or when a pet looks at the camera. The object included in the multiple images (100) may be photographed in different poses as time passes. Therefore, the image desired by the user may be a part of the multiple images (100) that captures the scene where the object assumes an important pose. Manually selecting the image in which the object assumes an important pose after taking multiple images (100) may be cumbersome for the user.

[0044] In one embodiment, the electronic device can detect an object from a plurality of images (100). The electronic device can detect a person within the plurality of images (100). For example, the electronic device can detect a child jumping off a park bench as an object from the plurality of images (100). Of course, unlike what is shown in FIG. 1a, the object within the plurality of images (100) may be a living creature rather than a person.

[0045] For the sake of convenience of explanation, the electronic device has only described the operation of detecting an object within an image using the first image (100a), but the electronic device can perform the operation of detecting an object for a plurality of images (100).

[0046] The electronic device can detect objects within a plurality of images (100) using an object detection algorithm. The object detection algorithm can be implemented using various deep learning models and does not limit the technical scope of the present disclosure.

[0047] In one embodiment, the electronic device may determine one of the plurality of images (100) as a base image.

[0048] The electronic device can detect objects included in a plurality of images (100) and obtain a body score for each image, which is a numerical score of the importance of the posture of the object in the image. For example, the direction and angle of the arms and legs at the moment of jumping, the facial expression at the moment a baby smiles, and the angle of the head at the moment a pet looks at the camera may be important postures.

[0049] The electronic device may use an artificial intelligence model trained to quantify the importance of a posture based on an exemplary important posture. The electronic device may use the trained artificial intelligence model to obtain a body score for the posture of an object included in a plurality of images (100). Based on the body score, the electronic device may determine one of the plurality of images (100) as a base image.

[0050] For example, among the plurality of images (100), the fourth image (100d) may be an image capturing an object falling while raising its hand. The electronic device may evaluate the pose of the object in the fourth image (100d) as an important pose by a learned artificial intelligence model and assign a high body score to the fourth image (100d). The electronic device may determine the fourth image (100d) among the plurality of images (100) as the base image.

[0051] In one embodiment, the electronic device may acquire a fourth image (100d) containing an important pose by inputting a plurality of images (100) into an artificial intelligence model learned based on an exemplary important pose. In one embodiment, the electronic device may determine the subject of the plurality of images (100) through a model that analyzes the images, determine an important pose corresponding to the determined subject (e.g., a jumping child), and acquire a fourth image (100d) corresponding to the determined important pose.

[0052] In one embodiment, the electronic device may acquire one of a plurality of images (100) as a highlight image for a specific body part of an object. The electronic device may acquire a highlight image containing a body part of an object among the plurality of images (100). For example, the electronic device may acquire a second image (100b) containing a head part of a child among the plurality of images (100) as a highlight image for the head part. As another example, the electronic device may acquire a third image (100c) containing an upper body part of a child among the plurality of images (100) as a highlight image for the upper body part. As another example, the electronic device may acquire a fifth image (100e) containing a right leg part of a child among the plurality of images (100) as a highlight image for the right leg part. As another example, the electronic device may acquire a sixth image (100f) containing a right arm part of a child among the plurality of images (100) as a highlight image for the right arm part.

[0053] Specifically, in one embodiment, the electronic device can detect subdivided objects from a plurality of images (100). The electronic device can detect a person within the plurality of images (100) by body part. For example, the electronic device can acquire a first image (100a). The electronic device can detect the head part (110a) of a child from the first image (100a). The electronic device can detect the upper body part (120a) of a child from the first image (100a). The electronic device can detect the right arm part (130a) of a child from the first image (100a). The electronic device can detect the right leg part (140a) of a child from the first image (100a). For convenience of explanation, the electronic device is shown as detecting four parts of the child from the image, but it may detect more subdivided parts or fewer simplified parts.

[0054] In one embodiment, the electronic device detects an object included in a plurality of images (100), and further subdivided parts of the object, and can obtain a part score, which is a numerical score of the visual fidelity of a part of the body of an object in an image, for each image and for each body part. For example, the electronic device can obtain a part score corresponding to the head part (110a) of a child in a first image (100a). As another example, the electronic device can obtain a part score corresponding to the upper body part (120a) of a child in a first image (100a). As yet another example, the electronic device can obtain a part score corresponding to the head part (110b) of a child in a second image (100b).

[0055] The electronic device may utilize an artificial intelligence model trained to quantify the visual completeness of a body part of an object within an image. By utilizing the trained artificial intelligence model, the electronic device may obtain a part score for a body part of an object included in a plurality of images (100). Based on the part score, the electronic device may determine one of the plurality of images (100) as a highlight image for a specific body part. For example, the electronic device may determine one of the plurality of images (100) that has a high part score for the head of the object as a highlight image for the head. The electronic device may obtain a second image (100b) that has a high part score for the head of the object among the plurality of images (100) as a highlight image for the head.

[0056] In one embodiment, the visual completeness may be the degree to which a part of an object in an image has a direction and angle suitable for the object's important pose, and whether it appears clear considering the brightness, focus, etc. of the image. The visual completeness may be determined based on at least one of the degree to which a part of the body contributes to the importance of the object's pose in the image, the clarity of the part of the body, and whether the part of the body is occluded.

[0057] For example, multiple images (100) may be images containing a series of sequences of an object jumping off a park bench, and the posture of the object falling with its hands raised may be evaluated as a key posture. The right arm portion (130a) of the first image (100a) may be obtained with the arm lowered and may not be an arm movement suitable for the key posture of the object falling with its hands raised. In comparison, the sixth image (100f) captures the appearance of an object falling with its arms raised, and the right arm portion (130f) of the sixth image (100f) may be an arm movement suitable for the key posture as an arm raised posture. The electronic device may use an artificial intelligence model to highly evaluate the visual completeness of the right arm portion (130f) of the sixth image (100f) and determine the sixth image (100f) as a highlight image for the right arm portion.

[0058] As another example, one of the multiple images (100) may not be sharp with respect to the right arm portion of the object. For example, the image may not be sharp because the focus is not accurate, or the sharpness of the right arm portion may decrease as the object shakes its right arm. The electronic device may rate the visual completeness as low for one image in which the sharpness of the object's right arm portion is reduced.

[0059] As another example, a part of the object's body may be obscured in one of the multiple images (100) over time. For example, a part of the object's right arm may be obscured by a flying bird. The obscured right arm in the image may not be suitable for replacing the object's right arm in the base image, and the electronic device may rate the visual completeness of the image containing the obscured right arm as low.

[0060] The electronic device can obtain a part score, which is a numerical score of the visual completeness of a body part of an object within multiple images, by image and by body part using a second artificial intelligence model trained to numerically quantify the visual completeness of a body part. Based on the part score, the electronic device can determine the second image (100b) among the multiple images (100) as a highlight image for the head part. The electronic device can obtain the head part (110b) of the second image (100b) as a fragment image. Based on the part score, the electronic device can determine the third image (100c) among the multiple images (100) as a highlight image for the upper body part. The electronic device can obtain the upper body part (120c) of the third image (100c) as a fragment image. Based on the part score, the electronic device can determine the fifth image (100e) among the multiple images (100) as a highlight image for the right leg part. The electronic device can acquire the right leg portion (140e) of the fifth image (100e) as a fragment image. Based on the part score, the electronic device can determine the sixth image (100f) among the plurality of images (100) as a highlight image for the right arm portion. The electronic device can acquire the right arm portion (130f) of the sixth image (100f) as a fragment image.

[0061] FIG. 1b is a conceptual diagram illustrating a method of combining images based on a plurality of images analyzed according to one embodiment of the present disclosure.

[0062] For reference, the fourth image (100d) shown in FIG. 1b may be the fourth image (100d) determined as the base image in FIG. 1a. The first fragment image (110b) shown in FIG. 1b may be an image of the child's head area detected from the second image (100b) in FIG. 1a. The second fragment image (120c) shown in FIG. 1b may be an image of the child's upper body area detected from the third image (100c) in FIG. 1a. The third fragment image (140e) shown in FIG. 1b may be an image of the child's right leg area detected from the fifth image (100e) in FIG. 1a. The fourth fragment image (130f) shown in FIG. 1b may be an image of the child's right arm area detected from the sixth image (100f) in FIG. 1a.

[0063] Referring to FIGS. 1a and 1b, the electronic device can acquire a fourth image (100d) among a plurality of images as a base image. The electronic device can acquire a first fragment image (110b) of the head portion of an object from a second image (100b) among the plurality of images (100). The second image (100b) may be a highlight image of the head portion. Likewise, the electronic device can acquire a second fragment image (120c) of the upper body portion of an object from a third image (100c) among the plurality of images (100). The electronic device can acquire a third fragment image (140e) of the right leg portion of an object from a fifth image (100e) among the plurality of images (100). The electronic device can acquire a fourth fragment image (130f) of the right arm portion of an object from a sixth image (100f) among the plurality of images (100).

[0064] In one embodiment, the electronic device can correct a base image based on at least one fragment image. The electronic device can correct a fourth image (100d) based on the first to fourth fragment images (110b, 120c, 140e, 130f). The electronic device can correct the areas of body parts of an object within the fourth image (100d) based on the first to fourth fragment images (110b, 120c, 140e, 130f).

[0065] For example, the electronic device can correct the area of ​​the object's head portion in the fourth image (100d) based on the first fragment image (110b) of the object's head portion. The electronic device can replace the area of ​​the object's head in the fourth image (100d) with the first fragment image (110b). The electronic device can insert the first fragment image (110b) into the area of ​​the object's head in the fourth image (100d). The electronic device can change the area of ​​the object's head in the fourth image (100d) by inpainting it based on the first fragment image (110b).

[0066] As another example, the electronic device can correct the area of ​​the object's upper body in the fourth image (100d) based on the second fragment image (120c) of the object's upper body. The electronic device can change the area of ​​the object's upper body in the fourth image (100d) by inpainting it based on the first fragment image (110b). Similarly, the electronic device can correct the area of ​​the object's right leg in the fourth image (100d) based on the third fragment image (140e) of the object's right leg. The electronic device can correct the area of ​​the object's right arm in the fourth image (100d) based on the fourth fragment image (130f) of the object's right arm.

[0067] In one embodiment, the electronic device can obtain a best image (200) by correcting the corresponding body part area within the base image based on a fragment image for each body part. The best image (200) may be an image in which the head part of an object within the base image (the fourth image; 100d) is corrected based on the first fragment image (110b). The best image (200) may be an image in which the upper body part of an object within the base image (the fourth image; 100d) is corrected based on the second fragment image (120c). The best image (200) may be an image in which the right leg part of an object within the base image is corrected based on the third fragment image (140e). The best image (200) may be an image in which the right arm part of an object within the base image (the fourth image; 100d) is corrected based on the fourth fragment image (130f).

[0068] FIG. 2 is a flowchart illustrating the overall flow of a method for generating a best image according to one embodiment of the present disclosure.

[0069] In step S10, the electronic device can acquire an image. The electronic device may include a camera configuration and can capture an image. The electronic device can acquire the captured image through a communication unit. The acquired image may be a plurality of images, or an image sequence, video, etc., that is continuously captured over a set period.

[0070] In step S110, the electronic device can determine a base image among the acquired multiple images. In step S111, the electronic device can analyze the clarity, composition, and pose of the main objects of the acquired multiple images. In step S112, the electronic device can determine one image among the acquired multiple images as the base image. The electronic device can determine the base image based on the analysis result of step S111.

[0071] In one embodiment, the electronic device can analyze the quality of image capture. The quality of image capture may be an evaluation of whether the image was captured well, including image clarity and image composition, but is not limited to clarity and composition.

[0072] In one embodiment, the electronic device can analyze the sharpness of a plurality of acquired images. For example, the electronic device can quantitatively measure sharpness by analyzing the boundary intensity of the image using a Sobel filter. As another example, the electronic device can evaluate the sharpness of the image by applying a Fourier transform to the image to identify the distribution of high-frequency components corresponding to sharp edges.

[0073] In one embodiment, the electronic device can analyze the composition of a plurality of acquired images. For example, if an object is located in the center of an image, the electronic device can evaluate the composition of the image highly. As another example, if an object is located in the periphery of an image, the electronic device can evaluate the composition of the image low.

[0074] In addition, the composition of an image can be implemented in various ways, not just by placing an object in the center, but also through the Rule of Thirds (whether the subject is placed at the point of the third division of the image), Symmetry and Balance (whether the subject or background within the image is symmetrical or balanced), Diagonal Composition (whether the subject is placed along a diagonal line), Framing (a composition that encloses the subject within the frame using natural elements (doors, tree branches, etc.), Leading Lines (a method of guiding the gaze in a specific direction using linear elements such as roads, railways, and fences), Negative Space, Triangular Composition (arranging three main elements within the image in a triangular shape), and Golden Ratio Composition (arranging the subject by applying the Fibonacci sequence).

[0075] A method for analyzing the completeness of an image shot by evaluating the composition of the image will be explained once again using Figs. 6, 7a, and 7b.

[0076] In one embodiment, the electronic device may use an artificial intelligence model trained to quantify the completeness of image capture based on the composition and sharpness of the image. The electronic device may obtain an image score regarding the completeness of image capture of a plurality of images using the trained artificial intelligence model. Based on the image score, the electronic device may determine one of the plurality of images as a base image.

[0077] In one embodiment, the electronic device can analyze the pose of an object within an image. In one embodiment, the electronic device can acquire feature points for an object included in a plurality of images. The electronic device can determine the pose of the object based on the feature points. The electronic device can acquire a body score for each image based on the determined pose of the object. The body score may be a score that quantifies the importance of the pose of an object included in a plurality of images.

[0078] In one embodiment, the importance of a posture may vary depending on the body alignment that is critical to achieving a specific purpose. For example, the direction and angle of the arms and legs at the moment of jumping, the facial expression of a baby smiling, or the angle of a pet's head when looking at a camera may be postures of high importance. Postures of high importance may be primarily selected by the user and secondarily evaluated based on an artificial intelligence model trained to acquire a body score that quantifies the importance of the posture based on exemplary important postures.

[0079] A method for analyzing the pose of an object by extracting feature points from an object within an image is explained once again using Figures 4 and 5.

[0080] In one embodiment, the electronic device may utilize an artificial intelligence model trained to quantify the importance of a posture based on an exemplary important posture. The exemplary important posture may be a body alignment critical to achieving a specific purpose, but may be pre-selected by the user to train the artificial intelligence model. The electronic device may obtain a body score for the posture of an object included in a plurality of images using the trained artificial intelligence model. Based on the body score, the electronic device may determine one of the plurality of images as a base image.

[0081] In one embodiment, the electronic device may utilize a first artificial intelligence model trained to quantify the completeness of image capture and a second artificial intelligence model trained to quantify the importance of pose. Of course, the first artificial intelligence model and the second artificial intelligence model have been distinguished merely for the sake of convenience of explanation; the first artificial intelligence model and the second artificial intelligence model may be a single artificial intelligence model trained to obtain a single score quantifying the completeness of image capture and the importance of pose. The electronic device may obtain image scores and body scores for multiple images using the trained artificial intelligence model. Based on the image scores and body scores, the electronic device may determine one of the multiple images as a base image.

[0082] In step S120, the electronic device can determine, for each body part, a highlight image containing the optimal body part among a plurality of images.

[0083] In step S121, the electronic device can recognize a main object included in a plurality of images. For example, as in FIG. 1a, the electronic device can acquire a plurality of images (100). The plurality of images (100) may be images that include a child as a main subject, and the electronic device can recognize the child as the main object in the plurality of images (100). To recognize the object, the electronic device may use an object detection algorithm, which does not limit the technical scope of the present disclosure.

[0084] In step S122, the electronic device can detect feature points of objects in each image. In step S123, the electronic device can distinguish body parts of objects within the image based on the detected feature points.

[0085] In one embodiment, the electronic device can detect feature points corresponding to objects in an image using a feature point detection algorithm. The electronic device can correspond feature points to major body points of the detected objects in the image and can detect combinations of multiple feature points for each body.

[0086] In one embodiment, the electronic device can distinguish body parts of an object within an image based on feature points. The electronic device can detect body parts of an object within an image based on feature points. For example, the electronic device can acquire feature points corresponding to key points of the right arm. The acquired feature points may include, for example, feature points corresponding to the hand, feature points corresponding to the elbow, and feature points corresponding to the shoulder. The electronic device can detect the right arm of an object within an image based on feature points corresponding to key points of the right arm.

[0087] Of course, the electronic device can directly detect specific body parts of an object within an image using an object detection algorithm. However, the electronic device can perform more accurate body part detection by distinguishing the body parts of an object based on feature points.

[0088] In step S124, the electronic device can select highlight images of different body parts of an object from among the captured images.

[0089] In one embodiment, the electronic device can obtain a visual completeness of a body part of an object within captured images. The visual completeness of the body part may include the degree to which a part of the object within the image has a direction and angle suitable for the object's important pose, and whether it appears clear considering the brightness and focus of the image. The electronic device can evaluate the body part of an object within the images by using an artificial intelligence model trained to obtain a part score that quantifies the visual completeness of the body part of the object within the images.

[0090] In one embodiment, the electronic device may acquire highlight images for each body part among a plurality of images based on the acquired visual completeness. For example, regarding the head part of an object, the electronic device may acquire a second image among a plurality of images as a highlight image for the head part. As another example, regarding the upper body part of an object, the electronic device may acquire a third image among a plurality of images as a highlight image for the upper body part.

[0091] In step S20, the electronic device can obtain whether the base image and the highlight image selected for each body part are the same image. The electronic device can verify whether the base image and the highlight image selected for each body part are the same image.

[0092] For example, the electronic device may acquire a first image among the captured images as a base image. The electronic device may acquire a second image among the captured images as a highlight image of the head portion of the object. In this case, since the base image (first image) and the highlight image of the head portion (second image) are different from each other, step S131 is performed.

[0093] In step S131, the electronic device can correct the base image based on the highlight image. The electronic device can correct the area of ​​the first body part of the object within the base image based on the highlight image of the first body part. The electronic device can replace the area of ​​the first body part of the object within the base image based on the highlight image of the first body part. The electronic device can extract the first body part from the highlight image of the first body part as a first fragment image. The electronic device can insert the first fragment image into the area of ​​the first body part of the object within the base image. The electronic device can inpaint the area of ​​the first body part of the object within the base image based on the highlight image of the first body part.

[0094] In the subsequent step S132, the electronic device can generate a best image. The best image may be a base image corrected based on a highlight image of at least one body part of the object.

[0095] As another example, the electronic device may acquire a first image among the captured images as a base image. The electronic device may acquire the first image among the captured images as a highlight image of the head region of an object. In this case, since the base image and the highlight image of the head region are the same as the first image, step S132 is performed without performing step S131. That is, the electronic device may not perform the operation of replacing, inserting, or inpainting the area of ​​the head region of the highlight image of the head region (first image) with the area of ​​the head region of the object in the base image (first image). The electronic device may not correct the head region of the base image based on the highlight image of the head region, and may acquire the uncorrected base image as the best image. Of course, the area of ​​another part of the object in the base image may be corrected based on the highlight image of another part. The electronic device may acquire the corrected base image as the best image.

[0096] In one embodiment, at step S20, the electronic device can determine whether the base image and the highlight image selected for each body part are similar images. The electronic device can verify whether the base image and the highlight image selected for each body part are similar images. Whether they are similar images can be determined, for example, based on whether the numerically expressible image similarity exceeds a threshold value.

[0097] For example, an electronic device may acquire a first image among the captured images as a base image. The electronic device may acquire a second image among the captured images as a highlight image of the head portion of an object. In this case, the base image (the first image) and the highlight image of the head portion (the second image) may be different from each other. However, the head portion of the object in the first image and the head portion of the object in the second image may be similar.

[0098] In this case, since the head region of the object in the base image and the head region of the object in the highlight image are similar, step S132 may be performed without performing step S131. That is, the electronic device may not perform the operation of replacing, inserting, or inpainting the area of ​​the head region of the highlight image (second image) for the head region into the area of ​​the head region of the object in the base image (first image). The electronic device may not correct the head region of the base image based on the highlight image for the head region, and may obtain the uncorrected base image as the best image. Of course, the area of ​​another part of the object in the base image may be corrected based on the highlight image for another part. The electronic device may also obtain the corrected base image as the best image.

[0099] Techniques for determining whether the head regions of objects within an image are similar may utilize various algorithms, such as the SIFT (Scale-Invariant Feature Transform) algorithm, which extracts and compares feature points within an image, and methods for determining similar regions within an image based on color histograms; however, the present disclosure is not limited to specific algorithms.

[0100] FIG. 3 is a flowchart illustrating a method for generating a best image according to one embodiment of the present disclosure.

[0101] For the sake of convenience of explanation, parts that overlap with those explained using Figure 2 will be simplified.

[0102] In step S310, the electronic device can acquire multiple images containing an object.

[0103] In one embodiment, the electronic device may acquire a plurality of images. The electronic device may include a camera configuration and may capture images. The electronic device may acquire captured images through a communication unit. The acquired images may be image sequences, videos, etc., that are continuously captured over a set period.

[0104] In one embodiment, the acquired plurality of images may include an object. The acquired plurality of images may be images taken with the object as the subject. For example, the plurality of images may be images taken of a person. As another example, the object may include not only a person but also a living organism. The plurality of images may be images taken of a dog.

[0105] In step S320, the electronic device can acquire a body score for each image, which is a numerical score representing the importance of the pose of an object included in a plurality of images. The electronic device can analyze the pose of an object included in a plurality of images.

[0106] In one embodiment, the electronic device can detect objects within a plurality of images. The electronic device can detect objects within an image using an object detection algorithm.

[0107] In one embodiment, the electronic device can correspond feature points to major body points of the detected object. The electronic device can acquire feature points corresponding to major body points of the detected object. For example, the electronic device can correspond feature points to each point of the head, shoulders, elbows, hands, pelvis, knees, and feet, and acquire the corresponding feature points.

[0108] In one embodiment, the electronic device can acquire the attitude of an object based on the relationship between acquired feature points. For example, the electronic device can calculate the angle or movement of a body part based on the 2D coordinates of each feature point. As another example, the electronic device can evaluate the attitude of an object by calculating the relative position or distance between each body part. The method of acquiring the attitude of an object based on the relationship between feature points does not limit the technical scope of the present disclosure.

[0109] In one embodiment, the electronic device may obtain a body score, which is a numerical score representing the importance of the pose of an object included in an image. The electronic device may utilize an artificial intelligence model trained to numerically represent the importance of a pose based on exemplary important poses. The electronic device may obtain a body score regarding the pose of an object included in a plurality of acquired images by utilizing the trained artificial intelligence model. For example, the electronic device may obtain a body score of an object included in a first image by utilizing the trained artificial intelligence model.

[0110] For example, exemplary important poses could include the direction and angle of the arms and legs at the moment of jumping, the facial expression of a baby smiling, or the angle of a pet's head when looking at a camera. These exemplary important poses can be selected as training data for an artificial intelligence model based on the user's intent.

[0111] In step S330, the electronic device can acquire a base image among a plurality of images based on the body score.

[0112] For example, the electronic device can determine the first image having the highest body score among a plurality of images as the base image.

[0113] As another example, the electronic device may acquire multiple base images. The electronic device may acquire multiple images, and objects within the multiple images captured over time may sequentially assume various important poses. The electronic device may determine at least one image among the multiple images that has a body score exceeding a threshold value as the base image. The electronic device may determine at least one image among the multiple images that has a body score of a local maximum value as the base image. The criteria for the body score determining the base image are merely examples and do not limit the technical scope of the present disclosure.

[0114] In step S340, the electronic device can acquire one of the plurality of images as a highlight image of the first body part of the object.

[0115] In one embodiment, the electronic device may acquire a highlight image of a first body part of an object among a plurality of images. The electronic device may compare the first body part of the object with respect to each of the plurality of images. The electronic device may acquire a part score that quantifies the visual fidelity with respect to the first body part within the plurality of images. For example, the electronic device may determine a second image having the highest part score for the first body part among the plurality of images as the highlight image for the first body part.

[0116] Visual completeness may include the degree to which a part of the body contributes to the importance of the object's posture. For example, if an electronic device analyzes the object in the first image as raising its arms as an important posture and determines the first image as the base image, the object's arm portion among multiple images may have high visual completeness when it is in a posture extended toward the sky. As another example, if an electronic device analyzes the object in the first image as kicking as an important posture and determines the first image as the base image, the image may have high visual completeness as the object's leg portion included in multiple images resembles the 'kicking' motion.

[0117] In step S350, the electronic device can correct the area of ​​the first body part within the base image based on the highlight image. The electronic device can inpaint the area of ​​the first body part within the base image based on the highlight image.

[0118] In one embodiment, the electronic device may obtain a fragment image of a first body part of an object from a highlight image. The electronic device may insert the fragment image into an area of ​​the first body part within a base image. The electronic device may correct an area of ​​the first body part within the base image based on the fragment image. By correcting an area of ​​the first body part within the base image based on the fragment image, the electronic device may generate a best image. The best image may be a base image in which the area of ​​the first body part within the base image is corrected based on the fragment image.

[0119] FIG. 4 is a conceptual diagram illustrating a method for extracting feature points of an object in an image according to one embodiment of the present disclosure.

[0120] Referring to Fig. 4, a method for extracting feature points for an object in an image is explained.

[0121] In one embodiment, the electronic device may acquire a plurality of images. The first image (400a) may be one of the plurality of images. The electronic device may detect an object (410a) within the first image (400a). The electronic device may detect an object (410a) within the first image (400a) using an object detection algorithm.

[0122] In one embodiment, the electronic device can correspond feature points to major body points for the detected object (410b). The electronic device can extract feature points corresponding to major body points of the detected object (410b).

[0123] In one embodiment, feature points (420) may be extracted as shown in the second image (400b). An electronic device may acquire feature points (420) corresponding to major body points of the object (410b). Although FIG. 4 is illustrated as acquiring 13 feature points (420) for the body of the object (410b), the number of feature points does not limit the technical concept of the present disclosure. For example, an electronic device may acquire 17 or 33 feature points for the body.

[0124] In one embodiment, the feature points (420) extracted for an object (410b) in an image can be used to determine a base image. For example, the electronic device may evaluate the posture determined through the feature points (420) using a body score, and then determine the image that is determined to include an object taking a significant posture as the base image.

[0125] In one embodiment, the feature points (420) extracted for the object (410b) in the image can be used to determine a highlight image for a specific body part. For example, the electronic device can more accurately distinguish the body part of the object (410b) in the image through the feature points (420). The electronic device can determine the highlight image by evaluating the visual completeness of the accurately distinguished body part. The electronic device can evaluate the visual completeness of the distinguished first body part of the object (410b) in the image through a part score, and then determine the image containing the best-captured first body part as the highlight image for the first body part.

[0126] FIG. 5 is a conceptual diagram illustrating a method for evaluating the pose of an object in an image based on feature points according to one embodiment of the present disclosure.

[0127] For the sake of convenience of explanation, parts that overlap with those explained using FIG. 1a and FIG. 4 are simplified or omitted. For reference, FIG. 5 describes a method for extracting feature points and evaluating the pose of an object for the second image (100b), the fourth image (100d), and the seventh image (100g) among the multiple images (100) of FIG. 1a.

[0128] Referring to FIG. 5, an electronic device may acquire a plurality of images (100). For example, the electronic device may acquire a second image (100b), a fourth image (100d), and a seventh image (100g) among the plurality of images. The second image (100b), the fourth image (100d), and the seventh image (100g) may each include an object. FIG. 5 illustrates images each including a child in a pose.

[0129] In one embodiment, the electronic device can detect an object within the second image (100b). The electronic device can extract feature points for the object detected within the second image (100b). The extracted feature points may consist of 13 points, and the 13 feature points extracted from the second image (100b) constitute a first pose (150b) taken by the corresponding object. The feature points within the first pose (150b) may correspond to major body points of the object within the second image (100b).

[0130] Likewise, the electronic device can detect an object in the fourth image (100d). The electronic device can extract a second pose (150d) for the object detected in the fourth image (100d). Feature points in the second pose (150d) may correspond to major body points of the object in the fourth image (100d). The electronic device can detect an object in the seventh image (100g). The electronic device can extract a third pose (150g) for the object detected in the seventh image (100g). Feature points in the third pose (150g) may correspond to major body points of the object in the seventh image (100g).

[0131] In one embodiment, the electronic device can obtain a body score regarding the posture of an object in an image by analyzing the posture of the object determined based on feature points of the object in the image. The body score may be a numerical score representing the importance of the object's posture. For example, the direction and angle of the arms and legs at the moment of jumping, the facial expression of a baby at the moment of smiling, and the angle of the head of a pet at the moment of looking at a camera may be important postures.

[0132] In one embodiment, the electronic device may use an artificial intelligence model trained to quantify the importance of a posture based on an exemplary important posture. The electronic device may obtain a body score that evaluates the posture of an object in an image using the trained artificial intelligence model. For example, the electronic device may assign a body score of 5 to the second image (100b) through the first posture (150b). The electronic device may assign a body score of 9 to the fourth image (100d) through the second posture (150d). The electronic device may assign a body score of 6 to the seventh image (100g) through the third posture (150g). For example, the artificial intelligence model may be a model trained based on the posture at the moment of jumping as an exemplary important posture, and may evaluate the posture of the object in the fourth image (100d) with high importance.

[0133] In one embodiment, the electronic device may determine one of a plurality of images as a base image based on a body score. For example, the electronic device may determine a fourth image (100d) that records a higher body score among the plurality of images as the base image.

[0134] FIG. 6 is a conceptual diagram illustrating a method for evaluating the completeness of a shot based on the composition of an image according to an embodiment of the present disclosure. For reference, FIG. 6 describes the composition of an image based on an example image (600) and describes a change in the composition of the image when a portion of the example image (600) is selected according to the first box (610), the second box (620), the third box (630), and the fourth box (640). Furthermore, a method for evaluating an image score regarding the completeness of a shot of an image based on the composition of the image is described.

[0135] As an example image (600), an image is shown in which a horse and a person riding a horse are photographed as subjects (650).

[0136] In one embodiment, the electronic device can evaluate the quality of image capture based on the composition of the image. The composition of the image can be implemented in various ways, not only by simply placing an object in the center, but also by the Rule of Thirds composition (whether the subject is placed at the point of thirds of the image), Symmetry & Balance (whether the subject or background within the image is symmetrical or maintains balance), Diagonal Composition (whether the subject is placed along a diagonal line), Framing (a composition that encloses the subject within the screen using natural elements (doors, tree branches, etc.), Leading Lines (a method of guiding the gaze in a specific direction using linear elements such as roads, railways, and fences), Negative Space, Triangular Composition (arranging three main elements within the image in a triangular shape), and Golden Ratio Composition (arranging the subject by applying the Fibonacci sequence). In Fig. 6, the quality of the image capture is explained by focusing on whether the object is placed in the center of the image as an example, but various compositions such as symmetry, balance, framing, and triangular composition can be considered for the image composition.

[0137] In one embodiment, for the image corresponding to the first box (610), the subject (650) may be located at the bottom right of the center of the image. For the image corresponding to the second box (620), the subject (650) may be located at the center of the image. For the image corresponding to the third box (630), the subject (650) may be located at the bottom right of the center of the image, and the legs of the horse, which is the subject (650), may not be visible. The legs of the horse, which is the subject (650), may not be located within the third box (630). For the image corresponding to the fourth box (640), the subject (650) may be located at the bottom left of the center of the image. Accordingly, the electronic device can evaluate the shooting quality highly for the image corresponding to the second box (620) in which the subject (650) is positioned at the center of the image among the first to fourth boxes (610 to 640). The electronic device can obtain a high image score, which is a numerical score of the shooting quality, for the image corresponding to the second box (620) among the first to fourth boxes (610 to 640).

[0138] For example, the electronic device can obtain an image score of 4.1 for the image corresponding to the first box (610). The electronic device can obtain an image score of 4.5 for the image corresponding to the second box (620). The electronic device can obtain an image score of 2.2 for the image corresponding to the third box (630). The electronic device can obtain an image score of 1.7 for the image corresponding to the fourth box (640).

[0139] FIG. 7a is a drawing illustrating an exemplary composition that can be implemented in an image according to one embodiment of the present disclosure. FIG. 7b is a drawing illustrating an exemplary composition that can be implemented in an image according to one embodiment of the present disclosure.

[0140] Using Figs. 7a and 7b, a method for evaluating the completeness of an image shot according to the composition of the image is further explained as an example.

[0141] Referring to FIG. 7a, the electronic device can obtain a first example image (700a). The first example image (700a) may include a subject (750a).

[0142] In one embodiment, the electronic device may use an artificial intelligence model trained to quantify the quality of image capture based on the composition of the image. The electronic device may obtain an image score regarding the quality of image capture by using the trained artificial intelligence model.

[0143] For example, in the first example image (700a), the subject (750a) may be offset to the right from the center of the image. Additionally, the first example image (700a) may include a somewhat wide blank area above the subject (750a). The electronic device may obtain an image score for the first example image (700a) using a learned artificial intelligence model. For example, the fact that the subject (750a) is offset to the right from the center of the image, the fact that there is a somewhat wide blank area above the subject (750a), and the fact that part of the subject's (750a) left foot is cut off may have an adverse effect on the composition of the image.

[0144] Referring to FIG. 7b, the electronic device can obtain a second example image (700b). The second example image (700b) may include a subject (750b).

[0145] In one embodiment, the electronic device may obtain an image score regarding the completeness of the image capture using a learned artificial intelligence model. For example, in the second example image (700b), the subject (750b) may be located at the center of the image. The second example image (700b) may be an image targeting the upper body of the subject (750b). The second example image (700b) may be evaluated positively based on the fact that the subject (750b) is located at the center of the image and that it is an image targeting the upper body of the subject (750b).

[0146] For example, the electronic device may use a trained artificial intelligence model to evaluate the composition of the first example image (700a) as relatively poor and assign a low image score. The electronic device may evaluate the composition of the second example image (700b) as relatively good and assign a high image score. Of course, the first example image (700a) and the second example image (700b) were compared and explained only for the convenience of explanation; the completeness of the image capture can be evaluated by considering the training data of the artificial intelligence model, the degree to which each element influences the image composition, etc.

[0147] FIG. 8 is a flowchart illustrating a method for evaluating the pose of an object in an image according to one embodiment of the present disclosure.

[0148] For the convenience of explanation, parts that overlap with those explained using Figures 2 and 3 are simplified or omitted.

[0149] Step S320 of Fig. 3 may include steps S810, S820, and S830.

[0150] In step S810, the electronic device can acquire feature points for an object included in a plurality of images.

[0151] In one embodiment, the electronic device can acquire an image containing an object. The electronic device can correspond feature points to body parts of the object. The feature points may include three-dimensional position coordinate data corresponding to body parts of the object. The electronic device can detect an object included in a plurality of images. For example, the object may be a person, and the electronic device can acquire feature points corresponding to major body points of the detected person.

[0152] In one embodiment, the feature points may include feature points corresponding to 13 major body points of a person. However, the number of feature points is merely an example and the technical concept of the present disclosure is not limited thereto, and the electronic device may include feature points corresponding to 17 or 33 major body points.

[0153] In one embodiment, the object may be a living organism other than a human. The electronic device may acquire feature points corresponding to key points of the detected living organism. For example, the electronic device may acquire feature points corresponding to key points of a detected dog. The feature points of the dog may be eyes, nose, mouth, ears, neck, forepaws, hind paws, shoulders, hips, the tip of the tail, etc.

[0154] In step S820, the electronic device can determine the pose of an object based on feature points. In step S830, the electronic device can acquire a body score corresponding to the determined pose of the object for each image.

[0155] In one embodiment, the electronic device may use an artificial intelligence model trained to quantify the importance of a pose based on an exemplary important pose. The electronic device may use the trained artificial intelligence model to obtain a body score for the pose of an object included in a plurality of images.

[0156] In one embodiment, the artificial intelligence model may perform a series of operations including extracting feature points of an object within an image, determining the pose of the object based on the feature points, and obtaining a body score based on the determined pose of the object. For example, an electronic device can obtain a body score regarding the pose of an object by inputting an image into the artificial intelligence model. As another example, the electronic device can obtain an image-specific body score regarding the pose of an object by inputting multiple images into the artificial intelligence model.

[0157] FIG. 9 is a flowchart illustrating a method for determining a base image among a plurality of images according to one embodiment of the present disclosure.

[0158] For the sake of convenience of explanation, parts that overlap with those explained using FIGS. 2 and FIGS. 3 are simplified or omitted. For reference, regarding the body score, since it overlaps with the explanation using step S320 of FIG. 3, it is simplified or omitted.

[0159] Step S330 of Fig. 3 may include steps S910 and S920.

[0160] In step S910, the electronic device can further acquire an image score for each image, which is a numerical score representing the completeness of capturing multiple images. The electronic device can analyze the completeness of capturing multiple images.

[0161] In one embodiment, the quality of the image capture may be an evaluation of whether the image was captured well, including image clarity and image composition, but is not limited to clarity and composition.

[0162] In one embodiment, the electronic device can analyze the sharpness of a plurality of acquired images. For example, the electronic device can quantitatively measure the sharpness of an image by analyzing the boundary intensity of the image using a Sobel filter. As another example, the electronic device can evaluate the sharpness of an image by applying a Fourier transform to the image and identifying the distribution of high-frequency components corresponding to sharp edges. The method of analyzing the sharpness of an image is merely an example and does not limit the technical scope of the present invention. The higher the sharpness of the image, the higher the quality of image capture the electronic device can determine.

[0163] In one embodiment, the electronic device can analyze the composition of a plurality of acquired images. For example, if an object is located in the center of an image, the electronic device may evaluate the composition of the image highly. As another example, if an object is located in the periphery of an image, the electronic device may evaluate the composition of the image low. The higher the evaluation of the image composition, the higher the electronic device can determine that the quality of the image capture is high.

[0164] In addition, when evaluating the composition of an image, not only is the object placed in the center considered, but also the Rule of Thirds (whether the subject is placed at the point of thirds of the image), Symmetry and Balance (whether the subject or background within the image is symmetrical or balanced), Diagonal Composition (whether the subject is placed along a diagonal line), Framing (a composition that encloses the subject within the frame using natural elements (doors, tree branches, etc.), Leading Lines (a method of guiding the gaze in a specific direction using linear elements such as roads, railways, and fences), Negative Space, Triangular Composition (arranging three main elements within the image in a triangular shape), and Golden Ratio Composition (arranging the subject by applying the Fibonacci sequence).

[0165] In one embodiment, the electronic device may further acquire an image score, which is a numerical score representing the completeness of image capture. The electronic device may utilize an artificial intelligence model trained to numerically represent the completeness of image capture, including image composition and sharpness. The electronic device may acquire image scores for a plurality of acquired images by utilizing the trained artificial intelligence model. For example, the electronic device may acquire an image score for a first image by utilizing the trained artificial intelligence model.

[0166] In step S920, the electronic device can acquire a base image among a plurality of images based on an image score and a body score.

[0167] For example, an electronic device may determine a first image as the base image among a plurality of images in which the sum of the image score and the body score is the highest. As another example, the electronic device may determine a first image as the base image among a plurality of images in which the image score exceeds a threshold value in which the first image has the highest body score. The method of determining a base image based on the image score and the body score is merely an example, and the technical concept of the present invention is not limited thereto. For example, by assigning a higher weight to either the image score or the body score, whether to use the calculated score value to determine the base image may vary depending on the intention of the user utilizing the method of the present disclosure.

[0168] FIG. 10 is a conceptual diagram illustrating a method for evaluating the visual completeness of body parts of an object in a plurality of images according to one embodiment of the present disclosure.

[0169] For the convenience of explanation, parts that overlap with those explained using FIGS. 1a to 9 are simplified or omitted.

[0170] Referring to FIG. 10, an electronic device may acquire a first image (1100) among a plurality of images. The first image (1100) may be any image among the plurality of images (100 in FIG. 1). For reference, FIG. 10 describes a method for determining a highlight image for each body part among a plurality of images by performing a series of image processing processes on any image. As the image processing processes are applied sequentially, the first images are shown as 1100, 1200, and 1300, and the first objects included within each first image are shown as 1110, 1210, and 1310.

[0171] In one embodiment, the image may include a plurality of objects. For example, the first image (1100) may include a first object (1110) and a second object (1120). An electronic device may detect objects within the image. For example, the electronic device may detect at least one object within the first image (1100). The electronic device may detect the first object (1110) and the second object (1120) included within the first image (1100). The electronic device may detect the first object (1110) and the second object (1120) within the first image (1100) using an object detection algorithm. The object detection algorithm may be implemented using various deep learning models and does not limit the technical scope of the present disclosure.

[0172] In one embodiment, the electronic device may determine one of a plurality of objects in an image as the main object. For example, the electronic device may determine the first object (1210) as the main object among a plurality of objects in the first image (1200).

[0173] For example, the electronic device can determine a primary object among multiple objects based on the size of a bounding box used to detect multiple objects included in a first image (1110). The electronic device can detect the first object (1110) based on a relatively large bounding box. The electronic device can detect the second object (1120) based on a relatively small bounding box. The electronic device can determine the first object (1110) detected using the larger bounding box among the first object (1110) and the second object (1120) in the first image (1100) as the primary object. The electronic device can determine the first object (1210) in the first image (1200) as the primary object.

[0174] As another example, the electronic device may acquire multiple images, which are image sequences acquired over time. The first image (1100) may be one of the multiple images. The electronic device may determine a primary object among the multiple objects based on the duration of a bounding box used to detect multiple objects included in the first image (1100) (i.e., the time it takes to successfully detect the same object through any bounding box within the image). Among the multiple objects, the first object (1110) may be detected across the entirety of the multiple images. Among the multiple objects, the second object (1120) may be detected across a portion of the multiple images. The electronic device may determine the first object (1110) detected across the entirety of the multiple images as the primary object. The electronic device may determine the first object (1210) within the first image (1200) as the primary object.

[0175] Of course, the method of determining one of multiple objects within an image as the primary object within the image is merely an example, and the technical concept of the present disclosure is not limited thereto.

[0176] In one embodiment, the electronic device may acquire feature points corresponding to major points on the body of a major object. The electronic device may acquire feature points (1320) for a first object (1310) determined as a major object. The feature points (1320) may include combinations composed of a total of 13 location data, and the number of feature points (1320) does not limit the technical concept of the present disclosure.

[0177] In one embodiment, the electronic device can extract body parts of an object in an image based on feature points (1320). The electronic device can extract body parts of a first object (1310) in a first image (1300) based on feature points (1320). The electronic device can obtain fragment images (1410 to 1460) corresponding to body parts of the first object (1310). For example, the electronic device can obtain a first fragment image (1410) for the head part of the first object (1310), a second fragment image (1420) for the upper body part, a third fragment image (1430) for the right arm part, a fourth fragment image (1440) for the left arm part, a fifth fragment image (1450) for the left foot part, and a sixth fragment image (1460) for the right foot part.

[0178] For example, the electronic device can define a body part based on the relationship of feature points (1320). The electronic device can define the right arm part as the area connecting the right shoulder feature point, the right elbow feature point, and the right wrist feature point among the feature points (1320) corresponding to the first object (1310) composed of 13 parts. The electronic device can acquire a third piece image (1430) to include the defined right arm part.

[0179] In one embodiment, a segmentation technique may be used to extract body parts of an object within an image. An electronic device may acquire fragment images (1410 to 1460) of body parts of a first object (1310) by using a segmentation model trained to distinguish each part of the body. The electronic device may acquire fragment images (1410 to 1460) of body parts of the first object (1310) by further inputting feature points (1320) of the first object (1310) into the segmentation model. By further inputting feature points, the segmentation model may distinguish body parts of the object within the image more precisely.

[0180] In one embodiment, the electronic device can acquire fragment images (1410 to 1460) of body parts of the first object (1310) without inputting feature points by using a learned segmentation model. However, the electronic device can distinguish the body parts of the first object (1310) more accurately based on feature points and acquire precise fragment images (1410 to 1460).

[0181] In one embodiment, the electronic device can analyze the visual fidelity of a body part of an object within an image. The electronic device can obtain a part score, which is a numerical score of visual fidelity, for each body part of the object within multiple images. Based on the part score, the electronic device can determine a highlight image for each body part among the multiple images.

[0182] For example, the electronic device may acquire a first image (1300) among a plurality of images as a highlight image of the head portion of the first object (1310) based on a part score for the head portion. The electronic device may acquire a first fragment image (1410) of the head portion of the first object (1310) from the first image (1300). The fragment image may be a part of the highlight image. As another example, the electronic device may acquire a second image as a highlight image of the upper body portion of the first object based on a part score for the upper body portion. The electronic device may acquire a fragment image of the upper body portion of the first object from the second image.

[0183] In one embodiment, the visual completeness of the fragment image can be determined based on the quality of the image, such as brightness and illumination, visibility, such as whether the target object is occluded by another object, and the pose of the target object.

[0184] For example, the electronic device can determine the visual completeness of the first fragment image (1410) based on the sharpness of the first fragment image (1410). The electronic device can determine the visual completeness of the first fragment image (1410) to be higher the higher the sharpness of the first fragment image (1410) including the head portion of the first object (1310). For example, the electronic device can quantitatively measure the sharpness by analyzing the boundary intensity of the image using a Sobel filter. As another example, the electronic device can evaluate the sharpness of the image by applying a Fourier transform to the image to check the distribution of high-frequency components corresponding to sharp edges. High sharpness may mean a low degree of blur or that the image is well in focus.

[0185] As another example, the electronic device may determine the visual quality of the first piece image (1410) based on the lighting of the first piece image (1410). The electronic device may determine the lighting condition based on the average brightness of the image or the brightness distribution of the image through histogram analysis (e.g., whether the brightness distribution is uniform). For example, if the brightness is too low or too high, the electronic device may determine the visual quality of the first piece image (1410) to be low. An appropriate brightness can be selected by the user. As another example, an uneven lighting distribution may be a factor in the visual quality being evaluated as low.

[0186] As another example, the electronic device may determine the visual completeness of the first fragment image (1410) based on whether the first fragment image (1410) is obscured. The electronic device may determine the visual completeness based on whether the head portion of the first object (1310) included in the first fragment image (1410) is obscured by another object. For example, if the head portion of the first object (1310) within the first fragment image (1410) is obscured by a passing pedestrian, the electronic device may determine the visual completeness of the first fragment image (1410) to be low.

[0187] As another example, the electronic device may determine the visual completeness of the first fragment image (1410) based on the posture of the body part of the first object (1310) within the first fragment image (1410). For example, the electronic device may acquire an image containing an object that assumes a posture most similar to the first important posture among multiple images as a base image by using an artificial intelligence model trained to quantify the importance of a posture based on an exemplary important posture. The electronic device may use the artificial intelligence model to compare the first body part of the first important posture with the first body parts of the first object (1310) within multiple images. The electronic device may determine the image having the arrangement most similar to the first body part of the first important posture as a highlight image for the first body part. The electronic device may extract a fragment image for the first body part from the determined highlight image.

[0188] For example, if the electronic device determines the base image to include a posture similar to the important posture of 'kicking a ball,' then when determining the highlight image, it may also determine the highlight image to include a right leg shape arranged similarly to the right leg shape of the 'kicking ball' motion.

[0189] FIG. 11 is a flowchart illustrating a method for determining a highlight image of a first body part of an object among a plurality of images according to one embodiment of the present disclosure.

[0190] For the convenience of explanation, parts that overlap with those explained using Figures 2 and 3 are simplified or omitted.

[0191] Referring to FIG. 11, step S340 of FIG. 3 may include step S1110 and step S1120.

[0192] In step S1110, the electronic device can obtain a part score, which is a numerical score of the visual completeness of a part of a body of an object in an image, for each body part of an object in multiple images.

[0193] In one embodiment, the electronic device can detect an object within a plurality of images. The electronic device can detect a body part of an object within a plurality of images. The method of detecting a body part does not limit the technical scope of the present disclosure. For example, a body part may be detected based on feature points. As another example, a body part may be directly detected by an object detection algorithm.

[0194] In one embodiment, the electronic device may assign a part score, which is a numerical score representing the visual completeness of each body part of an object within a plurality of images. For example, the electronic device may obtain a part score for the head part of an object within a first image. The electronic device may obtain a part score for the head part of an object within a second image. The electronic device may obtain a part score for the head part of an object for each of the plurality of images. Likewise, the electronic device may also obtain a part score for the right arm part of an object for each of the plurality of images.

[0195] In step S1120, the electronic device can acquire a highlight image of the first body part among a plurality of images based on a part score for the first body part of the object.

[0196] For example, the electronic device can acquire a part score for the head region of an object for each of a plurality of images. The electronic device can acquire the image among the plurality of images with the highest part score for the head region of the object as a highlight image for the head region. As another example, the electronic device can acquire a part score for the right arm region of an object for each of a plurality of images. The electronic device can acquire the image among the plurality of images with the highest part score for the right arm region of the object as a highlight image for the right arm region.

[0197] FIG. 12 is a flowchart illustrating a method for correcting a base image according to one embodiment of the present disclosure.

[0198] For the convenience of explanation, parts that overlap with those explained using Figures 2 and 3 are simplified or omitted.

[0199] Referring to FIG. 12, step S350 of FIG. 3 may include step S1210 and step S1220.

[0200] In step S1210, the electronic device may obtain a first fragment image of a first body part from a highlight image of a first body part. The first fragment image may include the first body part of an object as part of the highlight image of a first body part.

[0201] In step S1220, the electronic device can correct the base image by inserting a first piece image into the area of ​​the first body part within the base image.

[0202] In one embodiment, the electronic device may insert a first fragment image into an area of ​​a first body part within a base image. The electronic device may correct the area of ​​the first body part within the base image based on the first fragment image. The electronic device may inpaint the area of ​​the first body part within the base image based on the first fragment image.

[0203] In one embodiment, the electronic device can remove a first body part of an object within a base image. The electronic device can adjust the first fragment image to match the environment of the base image. For example, the electronic device can naturally change the first fragment image to the resolution of the base image. As another example, the electronic device can change the orientation and angle of the first body part of the first fragment image to match the connection point with the first body part within the base image. As yet another example, the electronic device can change the first fragment image to match the brightness of the base image. The electronic device can inpaint the first fragment image, on which image editing has been performed, onto the area of ​​the first body part of the base image.

[0204] The above operation can be performed by a trained artificial intelligence model. The series of image processing operations can be performed by inputting a base image and a first fragment image into the artificial intelligence model. By inputting the base image and the first fragment image into the artificial intelligence model, the electronic device can obtain a base image corrected based on the first fragment image.

[0205] FIG. 13 is a flowchart illustrating a method for extracting a piece image to be inpainted onto a base image according to one embodiment of the present disclosure.

[0206] For the sake of convenience of explanation, parts that overlap with those explained using FIG. 3 and FIG. 12 are simplified or omitted.

[0207] Referring to FIG. 13, step S1210 of FIG. 12 may include step S1310, step S1320 and step S1330.

[0208] In step S1310, the electronic device can detect an object included in the highlight image.

[0209] In one embodiment, the electronic device can detect an object included in a highlight image. The electronic device can detect an object included in a plurality of images, and since the operation of detecting an object included in a highlight image is no different, it is simplified.

[0210] The electronic device can detect objects within a highlight image using an object detection algorithm. The object detection algorithm can be implemented using various deep learning models and does not limit the technical scope of the present disclosure.

[0211] In step S1320, the electronic device can acquire feature points for the detected object. Step S1320 is omitted because it overlaps with the explanation using step S810 of FIG. 8.

[0212] In step S1330, the electronic device can acquire a first piece image of a first body part distinguished based on feature points.

[0213] In one embodiment, the electronic device can detect a body part of an object within a highlight image based on feature points. The electronic device can distinguish a body part of an object within a highlight image based on feature points. For example, the electronic device can distinguish a first body part region within a highlight image relating to a first body part based on feature points. The electronic device can acquire a first fragment image including the distinguished first body part region. The first fragment image may be a part of a highlight image relating to a first body part.

[0214] Of course, the electronic device can directly detect specific body parts of an object within an image using an object detection algorithm. However, the electronic device can perform more accurate body part detection by distinguishing the body parts of an object based on feature points.

[0215] FIG. 14 is a flowchart illustrating a method for generating a best image according to one embodiment of the present disclosure.

[0216] For the sake of convenience of explanation, parts that overlap with those explained using FIG. 3 are simplified or omitted. Steps S1410 to S1440 overlap with steps S310 to S340 of FIG. 3, so they are omitted, and the explanation focuses on steps S1450 to S1470.

[0217] Referring to FIG. 14, in step S1450, the electronic device can compare the area of ​​a first body part of an object in a base image with the area of ​​a first body part of a corresponding object in a highlight image. The highlight image may be an image relating to the first body part. The electronic device can compare whether the area of ​​the first body part in the base image and the highlight image is the same or similar.

[0218] An algorithm for comparing whether a first body part region within an image is similar may include various algorithms, such as a Scale-Invariant Feature Transform (SIFT) algorithm that extracts and compares feature points within an image, or a method for determining similar regions within an image based on a color histogram. However, this is merely an example, and the present disclosure is not limited to any specific algorithm.

[0219] In one embodiment, the electronic device may obtain a difference value between the regions of a first body part in a base image and a highlight image. If the obtained difference value exceeds a threshold value, the electronic device may determine that the regions of the first body part in the base image and the highlight image are dissimilar to each other. If the obtained difference value is less than the threshold value, the electronic device may determine that the regions of the first body part in the base image and the highlight image are identical or similar to each other.

[0220] In step S1460, if the electronic device determines that the first body parts are similar or identical based on the comparison result of step S1450, it may not correct the area of ​​the first body part within the base image based on the highlight image.

[0221] In step S1460, if the electronic device determines that the first body parts are dissimilar to each other as a result of the comparison in step S1450, it may perform step S1470. In step S1470, the electronic device may correct the area of ​​the first body part within the base image based on the highlight image.

[0222] In one embodiment, the electronic device can correct the area of ​​the first body part within a base image based on a highlight image of the first body part. The electronic device can substitute the area of ​​the first body part within the highlight image of the first body part within the base image with the area of ​​the first body part within the base image. The electronic device can insert the highlight image of the first body part into the area of ​​the first body part within the base image. The electronic device can generate a best image by inpainting the area of ​​the first body part within the base image based on the highlight image of the first body part. The electronic device can obtain a first fragment image including the first body part from the highlight image of the first body part. The electronic device can correct the area of ​​the first body part within the base image based on the first fragment image.

[0223] Hereinafter, with reference to FIG. 15, the configuration of an electronic device for performing the image processing operations described so far will be described. FIG. 15 is a block diagram illustrating the configuration of an electronic device according to one embodiment of the present disclosure.

[0224] For the convenience of explanation, parts that overlap with those explained using FIGS. 1 to 14 are simplified or omitted.

[0225] Referring to FIG. 15, an electronic device (1500) according to one embodiment may include an input / output interface (1510), memory (1520), and a processor (1530). However, the components of the electronic device (1500) are not limited to the examples described above, and the electronic device (1500) may include more components than the components described above, or fewer components. In one embodiment, some or all of the input / output interface (1510), memory (1520), and processor (1530) may be implemented in the form of a single chip, and the processor (1530) may include one or more processors.

[0226] The input / output interface (1510) may include an input interface (e.g., touch screen, hard button, microphone, etc.) for receiving control commands or information from a user, and an output interface (e.g., display panel, speaker, etc.) for displaying the result of an operation or the status of an electronic device (1500) according to the user's control.

[0227] For example, the electronic device (1500) can acquire multiple images based on a user's image capture command obtained through an input / output interface (1510). The processor (1530) of the electronic device (1500) can perform image processing operations described using FIGS. 1 to 14.

[0228] Memory (1520) is a configuration for storing various programs or data and may be composed of a storage medium or a combination of storage media such as ROM, RAM, hard disk, CD-ROM, and DVD. Memory (1520) may not exist separately but may be configured to be included in the processor (1530). Memory (1520) may be composed of volatile memory, non-volatile memory, or a combination of volatile and non-volatile memory. Programs or instructions for performing operations according to the embodiments described above with reference to FIGS. 1 through 14 may be stored in memory (1520). Memory (1520) may provide stored data to the processor (1530) upon the request of the processor (1530).

[0229] A processor (1530) is configured to control a series of processes to enable an electronic device (1500) to operate according to embodiments described above with reference to FIGS. 1 to 14, and may be composed of one or more processors. One or more processors included in the processor (1530) may be circuitry such as a System on Chip (SoC) or an Integrated Circuit (IC). In this case, one or more processors may be general-purpose processors such as a CPU, AP, or DSP (Digital Signal Processor), graphics-dedicated processors such as a GPU or VPU (Vision Processing Unit), or artificial intelligence-dedicated processors such as an NPU. For example, if one or more processors are artificial intelligence-dedicated processors, the artificial intelligence-dedicated processors may be designed with a hardware structure specialized for processing a specific artificial intelligence model.

[0230] The processor (1530) can write data to memory (1520) or read data stored in memory (1520), and in particular, can process data according to a predefined operation rule or artificial intelligence model by executing a program or at least one instruction stored in memory (1520). Accordingly, the processor (1530) can perform the operations described in the previously described embodiments, and the operations described as being performed by the electronic device (1500) in the previously described embodiments can be seen as being performed by the processor (1530) unless otherwise specified.

[0231] A method according to one embodiment may include the step of acquiring a plurality of images containing an object. The method may include the step of acquiring a body score for each image, which is a numerical score representing the importance of the pose of the object included in the plurality of images. The method may include the step of acquiring a base image among the plurality of images based on the body score. The method may include the step of acquiring one of the plurality of images as a highlight image for a first body part of the object. The method may include the step of correcting the area of ​​the first body part within the base image based on the highlight image.

[0232] In one embodiment, the method may consist of a plurality of images that are continuously captured over a set period.

[0233] In one embodiment, the step of acquiring the body score may include acquiring keypoints for an object included in a plurality of images, determining the pose of the object based on the keypoints, and acquiring a body score corresponding to the determined pose of the object for each image.

[0234] In one embodiment, the step of acquiring the base image may include the step of acquiring an image score, which is a numerical score of the shooting completeness of a plurality of images, and the step of acquiring the base image among a plurality of images based on the image score and the body score.

[0235] In one embodiment, the shooting quality may be determined based on at least one of the sharpness of the image and the composition of the image.

[0236] In one embodiment, the step of acquiring the highlight image may include acquiring a part score, which is a numerical score of the visual fidelity of a part of a body of an object in an image, for each body part of an object in a plurality of images, and acquiring a highlight image of a first body part among the plurality of images based on the part score for a first body part of an object.

[0237] In one embodiment, visual completeness may be determined based on at least one of the degree to which a body part contributes to the importance of the object's posture, the clarity of the body part, and whether the body part is occluded.

[0238] In one embodiment, the step of correcting the area of ​​the first body part within the base image may include the step of obtaining a first fragment image relating to the first body part from a highlight image and the step of correcting the base image by inserting the first fragment image into the area of ​​the first body part within the base image.

[0239] In one embodiment, the step of acquiring the first fragment image may include the step of detecting an object included in the highlight image, the step of acquiring keypoints for the detected object, and the step of acquiring a first fragment image relating to a first body part distinguished based on the keypoints.

[0240] In one embodiment, the step of correcting the area of ​​the first body part in the base image may include the step of comparing the area of ​​the first body part of the object in the base image with the area of ​​the first body part in the highlight image, and, based on the comparison result, if the difference between the area of ​​the first body part in the base image and the area of ​​the first body part in the highlight image exceeds a threshold value, the step of correcting the base image by substituting the area of ​​the first body part in the highlight image with the area of ​​the first body part in the base image.

[0241] An electronic device according to one embodiment may include an input / output interface for receiving user input requesting image processing and outputting an image processed according to the user input. The electronic device may include a memory in which commands for processing an image are stored. The electronic device may include at least one processor. By executing a program stored in memory or at least one instruction by at least one processor, the electronic device may acquire a plurality of images containing an object, acquire a body score for each image which is a numerical score of the importance of the pose of the object included in the plurality of images, acquire a base image among the plurality of images based on the body score, acquire one image among the plurality of images as a highlight image for a first body part of the object, and correct the area of ​​the first body part within the base image based on the highlight image.

[0242] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device can acquire keypoints for an object included in a plurality of images, determine the pose of the object based on the keypoints, and acquire a body score for each image based on the determined pose of the object.

[0243] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device further acquires an image score, which is a numerical score of the completeness of capturing a plurality of images, and can acquire a base image among the plurality of images based on the image score and the body score.

[0244] In one embodiment, the shooting quality may be determined based on at least one of the sharpness of the image and the composition of the image.

[0245] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device can acquire a part score, which is a numerical score of the visual fidelity of a part of a body of an object in an image, for each body part of an object in a plurality of images, and acquire a highlight image of a first body part among the plurality of images based on the part score for a first body part of an object.

[0246] In one embodiment, visual completeness may be determined based on at least one of the degree to which a body part contributes to the importance of the object's posture, the clarity of the body part, and whether the body part is occluded.

[0247] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device can obtain a first fragment image relating to a first body part from a highlight image and correct the base image by inserting the first fragment image into an area of ​​the first body part within the base image.

[0248] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device can detect an object included in a highlight image, acquire keypoints for the detected object, and acquire a first fragment image of a first body part distinguished based on the keypoints.

[0249] In one embodiment, by executing a program stored in memory or at least one instruction, the electronic device compares the area of ​​a first body part of an object in a base image with the area of ​​a first body part in a highlight image, and based on the comparison result, if the difference between the area of ​​the first body part in the base image and the area of ​​the first body part in the highlight image exceeds a threshold value, the base image can be corrected by replacing the area of ​​the first body part in the highlight image with the area of ​​the first body part in the base image.

[0250] In one embodiment, a computer-readable recording medium may be provided on which a program for performing the method of any one claim on a computer is recorded.

[0251] Various embodiments of the present disclosure may be implemented or supported by one or more computer programs, and computer programs may be formed from computer-readable program code and stored on a computer-readable medium. In the present disclosure, “application” and “program” may represent one or more computer programs, software components, instruction sets, procedures, functions, objects, classes, instances, related data, or parts thereof suitable for implementation in computer-readable program code. “Computer-readable program code” may include various types of computer code, including source code, object code, and executable code. “Computer-readable medium” may include various types of media accessible by a computer, such as read-only memory (ROM), random access memory (RAM), hard disk drive (HDD), compact disc (CD), digital video disc (DVD), or various types of memory.

[0252] Additionally, a device-readable storage medium may be provided in the form of a non-transitory storage medium. Here, a 'non-transitory storage medium' is a tangible device and may exclude wired, wireless, optical, or other communication links that transmit transient electrical or other signals. Meanwhile, this 'non-transitory storage medium' does not distinguish between cases where data is stored semi-permanently and cases where it is stored temporarily. For example, a 'non-transitory storage medium' may include a buffer in which data is stored temporarily. A computer-readable medium may be any available medium accessible by a computer and may include both volatile and non-volatile media, as well as removable and non-removable media. A computer-readable medium includes media in which data can be stored permanently and media in which data can be stored and subsequently overwritten, such as rewritable optical discs or erasable memory devices.

[0253] According to one embodiment, the method according to the various embodiments disclosed herein may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or distributed online (e.g., download or upload) through an application store or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product (e.g., a downloadable app) may be temporarily stored or temporarily created on a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0254] The foregoing description of the present disclosure is for illustrative purposes only, and those skilled in the art will understand that modifications can be easily made to other specific forms without altering the technical spirit or essential features of the present disclosure. For example, suitable results may be achieved even if the described techniques are performed in a different order than described, and / or components such as systems, structures, devices, circuits, etc., described are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.

[0255] The scope of the present disclosure is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts thereof should be interpreted as being included within the scope of the present disclosure.

Claims

1. A step of acquiring a plurality of images containing an object; A step of obtaining a body score for each image, which is a numerical score representing the importance of the pose of the object included in the plurality of images; A step of acquiring a base image among the plurality of images based on the body score above; A step of acquiring one of the plurality of images as a highlight image of a first body part of the object; and A method comprising the step of correcting the area of ​​the first body part within the base image based on the above highlight image.

2. In Paragraph 1, The step of obtaining the above body score is, A step of acquiring keypoints for the object included in the plurality of images; A step of determining the attitude of the object based on the above feature points; and A method comprising the step of obtaining the body score corresponding to the determined pose of the object for each image.

3. In any one of paragraphs 1 to 2, The step of acquiring the above base image is, A step of further acquiring an image score, which is a numerical score representing the shooting completeness of multiple images; and A method comprising the step of acquiring the base image among the plurality of images based on the image score and the body score.

4. In Paragraph 3, A method in which the above-mentioned shooting quality is determined based on at least one of the sharpness of the image and the composition of the image.

5. In any one of paragraphs 1 through 4, The step of acquiring the above highlight image is, A step of obtaining a part score, which is a numerical score of the visual fidelity of a body part of an object within an image, for each body part of the object within the plurality of images; and A method comprising the step of acquiring a highlight image of a first body part among a plurality of images based on a part score for a first body part of the object.

6. In Paragraph 5, A method in which the visual completeness is determined based on at least one of the degree to which the body part contributes to the importance of the object's posture, the clarity of the body part, and whether the body part is occluded.

7. In paragraphs 1 through 6, The step of correcting the area of ​​the first body part within the base image is, A step of obtaining a first fragment image relating to the first body part from the above highlight image; and A method comprising the step of correcting the base image by inserting the first fragment image into the area of ​​the first body part within the base image.

8. An input / output interface for receiving user input requesting image processing and outputting the processed image according to the user input; Memory where commands for processing images are stored; and It includes at least one processor, By the above-mentioned at least one processor executing a program stored in the memory or at least one instruction, the electronic device, Acquire multiple images containing an object, and A body score, which is a numerical score representing the importance of the pose of the object included in the plurality of images, is obtained for each image, and Based on the above body score, a base image among the plurality of images is obtained, and One of the plurality of images above is obtained as a highlight image of the first body part of the object, and An electronic device that corrects the area of ​​the first body part within the base image based on the above highlight image.

9. In Paragraph 8, By the above-mentioned at least one processor executing a program stored in the memory or at least one instruction, the electronic device, Obtain keypoints for the object included in the plurality of images above, and Determining the posture of the object based on the above feature points, and An electronic device that acquires the body score for each image based on the determined pose of the object.

10. In any one of paragraphs 8 through 9, By the above-mentioned at least one processor executing a program stored in the memory or at least one instruction, the electronic device, Acquire more image scores, which are numerical scores representing the shooting quality of multiple images, and An electronic device that acquires the base image among the plurality of images based on the image score and the body score.

11. In Paragraph 10, An electronic device in which the above-mentioned shooting quality is determined based on at least one of the sharpness of the image and the composition of the image.

12. In any one of paragraphs 8 through 11, By the above-mentioned at least one processor executing a program stored in the memory or at least one instruction, the electronic device, A part score, which is a numerical score of the visual fidelity of a body part of an object within an image, is obtained for each body part of the object within the plurality of images, and An electronic device that acquires a highlight image of a first body part among a plurality of images based on a part score for a first body part of the object.

13. In Paragraph 12, An electronic device in which the above visual completeness is determined based on at least one of the degree to which the body part contributes to the importance of the object's posture, the clarity of the body part, and whether the body part is occluded.

14. In paragraphs 8 through 13, By the above-mentioned at least one processor executing a program stored in the memory or at least one instruction, the electronic device, A first fragment image regarding the first body part is obtained from the above highlight image, and An electronic device that corrects a base image by inserting a first fragment image into an area of ​​a first body part within the base image.

15. A computer-readable recording medium having a program recorded thereon for performing the method of any one of paragraphs 1 through 7 on a computer.

Images