Body measuring device and controlling method for the same

Abstract

A body measurement device is provided that includes a camera that captures an image including an RGB image and a depth image for conveniently measuring body size, a display, a source image including a user's body contour image and a skeleton image based on the captured image, a body size including body height based on the source image, and a control unit that controls the display to display the body size.

Description

Technology Field

[0001] The present invention relates to a body measurement device and a control method thereof, wherein the device captures an image of a user, acquires a body contour image and a skeleton image based on the captured image, and measures the user's body size based on the acquired body contour and skeleton images. Background Technology

[0002] When purchasing clothing, buyers need to accurately determine their body size. In particular, when buying online, one cannot try on clothes; conversely, when buying offline, trying on items may be difficult, cause inconvenience, potentially damage the clothing, and delay the process.

[0003] Among the technologies developed to solve these problems, commercially available ones include those that use 3D scanners or set up multiple digital cameras to measure body size. However, these technologies require a specific location and expensive equipment, making them difficult for consumers to use and resulting in low utility.

[0004] In addition, conventional cameras had limitations in that they could only measure body size when the user was naked or in a tight-fitting state. The problem to be solved

[0005] One embodiment of the present invention aims to provide a body measurement device and a control method thereof, wherein when a user assumes a specific pose, an image of the user is captured, a body contour and skeleton image of the user are obtained based on the captured image, and the user's body size is measured based on the obtained body contour and skeleton image.

[0006] Another embodiment of the present invention aims to provide a body measurement device and a control method thereof, which extracts body length from body contour and skeleton images and determines the back waist circumference length by multiplying the body length by a specific parameter when the circumference is obtained.

[0007] Another embodiment of the present invention aims to obtain accurate body measurements by correcting for noise caused by hair, clothes, or shoes, when a user in a loose fit state takes measurements.

[0008] The technical problems to be solved by the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0009] According to one embodiment of the present invention, a body measurement device is provided comprising a camera that captures an image including an RGB image and a depth image, a display, and a control unit that acquires a source image including a user's body contour image and a skeleton image based on the captured image, acquires a body size including a body height based on the source image, and controls the display to display the body size.

[0010] According to another embodiment of the present invention, a method for controlling a body measurement device is provided, comprising the steps of: capturing an image including an RGB image and a depth image; obtaining a source image including a user's body contour image and a skeleton image based on the captured image; obtaining a body size including a body height based on the source image; and controlling a display to display the body size. Effects of the invention

[0011] According to one embodiment of the present invention, when a user assumes a specific pose, an image of the user is captured, a body contour and skeleton image of the user are generated from the captured image, and the user's body size is measured based on the generated body contour image, so that an image of the user can be captured and an accurate body size can be measured with a simple motion.

[0012] According to another embodiment of the present invention, a skeleton image can be extracted from a captured image, and a user's body contour image can be generated by combining the skeleton image, RGB image, and depth image, thereby allowing the user's body size to be measured more accurately.

[0013] According to another embodiment of the present invention, a first length is extracted from a body contour image, and a rear waist circumference length is determined by multiplying the first length by a specific parameter, so that the waist circumference length not visible in the front image can be accurately measured.

[0014] According to another embodiment of the present invention, even when the user's hair volume is included in the body height measurement, it can be corrected to obtain an accurate body height.

[0015] According to another embodiment of the present invention, even when an image is captured while the user is wearing shoes, it can be corrected to obtain an accurate body height.

[0016] According to another embodiment of the present invention, even when an image is captured while the user is wearing loose clothing, it can be corrected to obtain accurate circumference values.

[0017] The effects obtainable from the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0018] FIG. 1 is a diagram illustrating the configuration of a body measurement device according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating a control method for a body measurement device according to one embodiment of the present invention. FIG. 3 is a drawing showing a user standing in front of a body measurement device according to one embodiment of the present invention. FIG. 4 is a drawing showing a user standing in front of a body measurement device according to one embodiment of the present invention. FIG. 5 is a flowchart illustrating a control method for a body measurement device according to one embodiment of the present invention. FIG. 6 is a drawing illustrating an input image according to an embodiment of the present invention. FIG. 7 is a drawing illustrating the generation of a body contour image according to one embodiment of the present invention. FIG. 8 is a drawing illustrating the results of measuring a user's body size according to one embodiment of the present invention. FIG. 9 is a diagram illustrating a body measurement algorithm according to one embodiment of the present invention. FIG. 10 is a drawing illustrating a user's pose according to one embodiment of the present invention. FIG. 11 is a drawing illustrating a skeleton image according to one embodiment of the present invention. FIG. 12 is a diagram illustrating target segmentation using RGB images and depth images according to an embodiment of the present invention. FIG. 13 is a drawing illustrating a body size measurement target according to one embodiment of the present invention. FIG. 14 is a drawing illustrating the measurement of a user's height according to an embodiment of the present invention. FIG. 15 is a diagram illustrating two methods for measuring a user's waist circumference according to one embodiment of the present invention. FIG. 16 is a drawing illustrating the measurement of waist circumference length according to one embodiment of the present invention. FIG. 17 is a drawing illustrating the measurement of the front waist circumference, side waist circumference, and rear waist circumference according to an embodiment of the present invention. FIG. 18 is a diagram illustrating data regarding a user's actual key and a measured key according to an embodiment of the present invention. FIG. 19 is a drawing showing data on a user's actual chest circumference and measured chest circumference according to an embodiment of the present invention. FIG. 20 is a drawing illustrating data on the actual underburst circumference of a user and the measured underburst circumference according to an embodiment of the present invention. FIG. 21 is a drawing illustrating data on a user's actual waist circumference and measured waist circumference according to an embodiment of the present invention. FIG. 22 is a drawing illustrating data on a user's actual hip circumference and measured hip circumference according to an embodiment of the present invention. FIG. 23 is a diagram illustrating data regarding the error and accuracy of user body size measurement data according to one embodiment of the present invention. FIG. 24 is a diagram illustrating data regarding the error and accuracy of user body size measurement data according to one embodiment of the present invention. FIG. 25 is a flowchart relating to a control method of a body measurement device according to one embodiment of the present invention. FIG. 26 is a conceptual diagram showing face information for obtaining a hair volume value according to one embodiment of the present invention. FIG. 27 is a conceptual diagram relating to another method for obtaining hair volume values ​​according to one embodiment of the present invention. FIG. 28 is a skeleton image and a partial enlarged view for body height correction according to one embodiment of the present invention. FIG. 29 is an experimental value regarding body height estimation according to one embodiment of the present invention. FIG. 30 is a diagram illustrating the structure of a camera of a body measurement device according to one embodiment of the present invention. FIG. 31 is a drawing illustrating an example of use of a body measurement device according to one embodiment of the present invention. FIG. 32 is a drawing illustrating the execution screen of a body measurement device according to one embodiment of the present invention. FIG. 33 is a drawing illustrating an execution screen of a body measurement device according to one embodiment of the present invention. FIG. 34 is a drawing illustrating an execution screen of a body measurement device according to one embodiment of the present invention. Specific details for implementing the invention

[0019] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are assigned the same reference number regardless of drawing symbols, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles.

[0020] In addition, when describing the embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description is omitted.

[0021] In addition, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification, and the technical concept disclosed in this specification is not limited by the attached drawings; it should be understood that all modifications, equivalents, and substitutions included within the concept and technical scope of the present invention are included.

[0022] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0023] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0024] A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0025] In this application, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0026] FIG. 1 is a diagram illustrating the configuration of a body measurement device according to one embodiment of the present invention.

[0027] Referring to FIG. 1, the body measurement device (100) includes a camera (110), a display (120), a control unit (130), a sensor unit (140), and a memory (150).

[0028] The camera (110) captures a first image including an RGB image and a depth image.

[0029] The camera (110) includes a depth camera (110) capable of capturing a depth image, an RGB camera capable of acquiring an RGB color image, and an IR camera capable of creating a depth image.

[0030] The depth camera (110) includes an SL camera. The SL camera is a camera that uses a structured light method (SL, Structured Light), which emits a signal of a specific pattern and calculates depth by analyzing the degree to which the pattern is deformed according to the surface of an object. For example, when the SL camera is activated, an infrared pattern consisting of 30,000 dots is projected onto the face from a component called a dot projector. When this pattern is projected onto a flat surface such as a desk, it is reflected back to the infrared camera without distortion, but a human face is not flat. There are countless curves depending on the position and size of the eyes, nose, and mouth. Therefore, as the light emitted from the dot projector is reflected onto the face, a certain pattern of distortion is formed. The SL camera recognizes the user by reading the change in this pattern.

[0031] A depth image refers to an image that includes a depth map. A depth map is a single image in 3D computer graphics that contains information related to the distance from a viewpoint to the surface of an object.

[0032] The camera (110) includes a depth camera and an RGB camera. Additionally, the camera (110) may include a depth camera and an RGB camera individually. The camera (110) may include at least one depth camera and at least one RGB camera.

[0033] The display (120) displays a graphic image according to a control command from the control unit (130).

[0034] The control unit (130) estimates the user's pose based on the first image, and if the user's pose is the first pose, controls the camera (110) to capture a second image including an image of the user's body in front, generates a body contour image of the user based on the captured second image, measures the user's body size based on the generated body contour image, and controls the display (120) to display the user's body size.

[0035] The control unit (130) may be implemented in the physical form of a chipset. The control unit (130) may be implemented as a single chipset or as multiple chipsets. The control unit (130) may perform the role of a processor and may be, for example, a system-on-chip (SOC) or an application processor.

[0036] The sensor unit (140) senses the distance to the user according to a control command from the control unit (130).

[0037] The memory (150) stores the first image and the second image according to a control command from the control unit (130).

[0038] FIG. 2 is a flowchart illustrating a control method for a body measurement device according to an embodiment of the present invention. The present invention is performed by a control unit (130).

[0039] Referring to FIG. 2, a first image including an RGB image and a depth image is captured through a camera (110) (S210).

[0040] The user's pose is estimated based on the first image (S220).

[0041] When the user's pose is the first pose, the camera (110) is controlled to capture a second image including an image of the user's body in front (S230).

[0042] Based on the captured second image, a body contour image of the user is generated (S240). According to one embodiment of the present invention, the control unit (130) extracts a skeleton image based on the first image.

[0043] The user's body size is measured based on the generated body contour image (S250). Additionally, the control unit (130) can measure the body size based on the body contour image and the skeleton image.

[0044] Controls the display (120) to display the user's body size (S260).

[0045] FIG. 3 is a drawing showing a user standing in front of a body measurement device according to one embodiment of the present invention.

[0046] Referring to FIG. 3, a user (200) is standing in front of a body measurement device (100). Here, the camera (110) includes a depth camera.

[0047] First, the location of the camera (110) is explained.

[0048] Since the leg portion is not captured by the camera (110), it is set to a fixed value to measure the height. The position of the camera (110) changes depending on the range of the object to be measured (in the case of height, 140 cm to 200 cm). The distance between the camera (110) and the user (200) is determined by considering the size of the object being measured and the field of view (FOV) of the camera (110). It is determined by considering the accuracy of the distance value of the depth camera (110). That is, it is important to select a position with high accuracy, and if the distance between the camera (110) and the user (200) is too close or too far, the accuracy of the depth decreases.

[0049] The position of the camera (110) can be 1.3 m from the ground.

[0050] The distance between the camera (110) and the user can be 1.0 to 1.5 m.

[0051] The resolution of the camera (110) is explained. The higher the resolution of the camera, the higher the accuracy of MPP (Meter Per Pixel), so it can be measured more accurately.

[0052] The input image of the camera (110) is described.

[0053] The input image can be i) a depth image, ii) a depth image, an RGB image, or iii) a depth image, an IR RAW image.

[0054] In the case of a depth image, the control unit (130) can obtain distance and length information for body measurement from the depth image. If there are no additional RGB images or IR raw images, the control unit (130) can extract a skeleton image from the depth image.

[0055] In the case of RGB images and IR raw images, the control unit (130) extracts a pose for automatic shooting, etc. from the RGB images and IR raw images.

[0056] Additionally, it can be used for additional applications. For example, the control unit (130) can perform gender recognition through face recognition based on RGB images and IR raw images, weight prediction through deep-learning or a weight prediction algorithm, and back circumference prediction using deep-learning (when only the user's front image is captured).

[0057] In addition, the control unit (130) can improve accuracy by executing a fusion algorithm based on information obtained from a depth camera and information obtained from deep-learning using RGB images and IR-raw images as inputs.

[0058] FIG. 4 is a drawing showing a user standing in front of a body measurement device according to one embodiment of the present invention.

[0059] Unlike the embodiment of FIG. 3, images of the user can be captured from multiple angles using multiple cameras. When photographing the user using two cameras, one can be positioned to photograph the front of the user and the other to photograph the side of the user.

[0060] The distance from the camera to the user can be set to be the same as in the case of Fig. 3. It is preferable that the two cameras and the user be located within a space of 2.0 m x 2.0 m or a space of 3.0 m x 3.0 m.

[0061] The front camera can be positioned at a height of about 1.4 m from the ground, and the side camera at a height of about 0.7 m from the ground. The reason for the height difference between the two cameras is that the user is filmed while standing upright, and therefore multiple images spaced apart along the length help to obtain accurate information.

[0062] FIG. 5 is a flowchart illustrating a control method for a body measurement device according to an embodiment of the present invention. The present invention is performed by a control unit (130).

[0063] Referring to FIG. 5, a first image including a depth image and an RGB image is captured (S410).

[0064] The user's pose is estimated based on the first image (S420).

[0065] If the user's pose is the first pose (S430), a second image including the user's body image is captured (S440). The control unit (130) controls the camera (110) to take a single shot or continuous shots. Here, the first pose can be a single pose or multiple poses. If the first pose is multiple poses, there is an advantage in that the accuracy of body size measurement can be improved.

[0066] If the user's pose is not the first pose (S440), the user's pose is estimated again based on the first image (S420). The control unit (130) measures the tilt of the user's pose through the sensor unit (140), and if the user's pose is tilted further to the left or right than the first pose by a predetermined range, the control unit (130) displays a warning message on the screen. The control unit (130) can control a speaker (not shown) to output a voice message saying, "Your posture is tilted to the right. Please stand up straight."

[0067] According to the present invention, if the user's pose is tilted more than the first pose, a warning message is displayed, allowing the user to assume the correct posture and thus enabling accurate measurement of body size.

[0068] Combine the RGB image and the depth image (S450).

[0069] A body contour image is generated based on an RGB image and a depth image (S460). According to one embodiment of the present invention, the control unit (130) can generate a body contour image and a skeleton image based on an RGB image and a depth image.

[0070] Height, chest circumference, and waist circumference are measured based on a body contour image (S470).

[0071] Displays the measured body size (S480)

[0072] FIG. 6 is a drawing illustrating an input image according to an embodiment of the present invention. FIG. 6 includes FIG. 6(a) and FIG. 6(b).

[0073] Figure 6(a) represents an RGB image among the input images.

[0074] An RGB image refers to an image to which the RGB model has been applied. The RGB model is the most basic color model, which considers color as a combination of three components: Red, Green, and Blue. In the RGB model, black is represented as R=G=B=0, white as R=G=B=255, red as R=255, G=B=0, and yellow as R=G=255, B=0. When R=G=B, the color becomes gray, which is an achromatic color.

[0075] Since R, G, and B can each have values ​​between 0 and 255, using the RGB color model allows for a total of 256 x 256 x 256 = 16,777,216 different colors to be expressed.

[0076] Figure 6(b) represents a depth image among the input images.

[0077] A depth image refers to an image of the same scene captured from different angles to generate a depth map. A depth map is an image that represents the relative distances between pixels using a grayscale. In the case of a depth map, closer areas are represented by bright pixels, while distant areas are represented by dark pixels.

[0078] FIG. 7 is a drawing illustrating the generation of a body contour image according to an embodiment of the present invention. FIG. 7 includes FIG. 7(a), FIG. 7(b) and FIG. 7(c).

[0079] FIG. 7(a) represents a drawing illustrating a skeleton image extracted from a second image. FIG. 7(b) represents a drawing illustrating an image combining an RGB image and a depth image. FIG. 7(c) represents a user's body contour image.

[0080] Referring to FIG. 7(a), the control unit (130) extracts a skeleton image (610) based on the second image.

[0081] Referring to FIG. 7(b), the control unit (130) combines the RGB image and the depth image to generate a combined image (620).

[0082] Referring to FIG. 7(c), the control unit (130) combines the extracted skeleton image (610), RGB image, and depth image (620) to generate a user's body contour image (630).

[0083] The control unit (130) measures the user's height based on the camera center (632). A detailed explanation of this is provided later in FIG. 14.

[0084] FIG. 8 is a drawing illustrating the results of measuring a user's body size according to one embodiment of the present invention.

[0085] Referring to FIG. 8, the control unit (130) displays the results of the user's body size measurement on the screen.

[0086] For example, the results of the body size measurement include height (71), chest circumference (72), waist circumference (73), and hip circumference (74). Height (71) can be 172 cm, chest circumference (72) can be 100 cm, waist circumference (73) can be 94 cm, and hip circumference (74) can be 102 cm.

[0087] FIG. 9 is a diagram illustrating a body measurement algorithm according to one embodiment of the present invention.

[0088] Referring to FIG. 9, a depth image (810) and an RGB image (820) are first received as input.

[0089] The control unit (130) estimates the user's pose based on the RGB image (820) and generates pose key point coordinates (840).

[0090] The control unit (130) generates a target depth map (830) based on the depth image (810) and pose key point coordinates (840).

[0091] The control unit (130) generates a target segmented image (850) based on the target depth map (830) and the RGB image (820).

[0092] The control unit (130) performs body size measurement (860) based on the target segmented image (850).

[0093] The control unit (130) can measure the height, chest circumference, under-chest circumference, waist circumference, and hip circumference by measuring the target's body size.

[0094] FIG. 10 is a drawing illustrating a user's pose according to an embodiment of the present invention. FIG. 10 includes FIG. 10(a) and FIG. 10(b).

[0095] FIG. 10(a) is a drawing illustrating a user's pose. Referring to FIG. 10(a), the control unit (130) estimates the user's pose based on the ear image, nose image, eye image, shoulder, heel, and buttock image included in the RGB image.

[0096] The control unit (130) controls the camera (110) to automatically capture a second image including an image of the user's body in front when the user's pose is a first pose. Here, the first pose may be a key pose for executing a measurement algorithm.

[0097] A key pose can be defined as follows: It is a pose in which the user stands at attention, looking at the camera, with their arms extended to the sides at a specific angle. Here, the specific angle can be 30 degrees. Here, the key pose is not limited to a single posture and various postures are possible.

[0098] When a user strikes a key pose, the camera captures multiple frontal images of the user. This is because capturing multiple frontal images of the user allows data to accumulate, which can reduce the error rate compared to a single image.

[0099] Key poses can be identified from the key points of the pose. For example, when a user is standing directly in front of the camera, the key points could be the ankles, elbows, ears, eyes, nose, neck, shoulders, hips, etc.

[0100] FIG. 10(b) is a diagram illustrating a method for specifying a user to measure body size when there are multiple users.

[0101] Referring to FIG. 10(b), the sensor unit (140) senses the distance between the body measurement device (100) and the user.

[0102] When there are multiple users, the control unit (130) senses the user (200) closest to the body measurement device (100) among the multiple users through the sensor unit (140), and if the user's pose is the first pose, controls the camera (110) to capture a second image including a body image of the user (200) in front.

[0103] When multiple second images are captured, the control unit (130) independently performs all body size measurements based on the second images. Therefore, the present invention has the advantage of being able to reduce the error rate.

[0104] FIG. 11 is a drawing illustrating a skeleton image according to one embodiment of the present invention.

[0105] Referring to FIG. 11, the control unit (130) extracts a skeleton image (1010) based on a first image and estimates the user's pose based on the extracted skeleton image (1010).

[0106] The control unit (130) can estimate the user's pose using pose key points. The key points may be ears (17, 18), eyes (15, 16), nose (0), neck (0 to 1), shoulders (2, 5), elbows (3, 6), and hips (9, 12). These key points are used to select a region of interest for user body measurements. Additionally, these key points are used to identify key poses, and the key points can be used to estimate body measurement sizes.

[0107] The control unit (180) can estimate the user's pose based on the skeleton image (1010). Additionally, the control unit (180) can measure the user's arm length, leg length, waist circumference, chest circumference, and hip circumference based on the skeleton image (1010).

[0108] FIG. 12 is a diagram illustrating target segmentation using RGB images and depth images according to an embodiment of the present invention.

[0109] Referring to Fig. 12, target segmentation is performed using RGB images and depth images.

[0110] Target splitting is executed as follows.

[0111] The control unit (130) combines an RGB image and a depth image, and generates a body contour image based on the combined RGB image and depth image. This will be explained in more detail.

[0112] First, the control unit (130) filters the depth map guided by the RGB image. The depth image includes the depth map.

[0113] The control unit (130) performs trimap generation from the depth map. The control unit matts the depth map.

[0114] Explain the output of the target split.

[0115] The control unit (130) generates a segmented map of the target subject. Next, the control unit (130) performs the operation of removing noise and smoothing the edges on the segmented map.

[0116] According to the present invention, the segmentation map can be used as a guide for measuring body size.

[0117] Explain the target segmentation map.

[0118] As illustrated in FIG. 12, the body contour image can be divided into a black area (1110) and a white area (1120). The control unit (130) places the black area (1110) at the very front, viewing it as the foreground, and places the white area (1120) behind the black area (1110), viewing it as the background. Here, in the image, the black area (1110) may represent a person of interest to the user, and the white area (1120) may represent the background behind the person.

[0119] Accordingly, the control unit (130) generates a user's body contour image (1100) based on the captured second image.

[0120] FIG. 13 is a drawing illustrating a body size measurement target according to one embodiment of the present invention.

[0121] Referring to FIG. 13, based on a body contour image (1200), height (1210), chest circumference (1220), underbust circumference (1230), waist circumference (1240), and hip circumference (1250) can be measured.

[0122] Specifically, when measuring body size, the following data is required. Specifically, a depth map, pose key point coordinates, and a body segmentation mask are required. The control unit (130) generates a body contour image (1200) based on the depth map, pose key point coordinates, and body segmentation mask.

[0123] Next, the body size measurement targets are as follows. The control unit (130) measures at least one of the user's height (1210), arm length, leg length, chest circumference (1220), waist circumference (1240), and hip circumference (1250) based on the generated body contour image (1200).

[0124] Here, all measurement areas can be measured based on the contours of the body contour image (1200). In addition, for lengths, if they correspond to two points, for example, arm length and leg length can also be measured.

[0125] Next, we will divide the body proportions and explain how to determine the chest and waist.

[0126] The control unit (130) divides the body contour image (1200) into a predetermined ratio, determines the protruding part of the body contour image (1200) as the chest, and determines the indented part of the body contour image (1200) as the waist.

[0127] For example, the height (1210) means the distance from the ground (1212) to the top of the head (1214).

[0128] The range between the neck and hips of the body contour image (1200) is primarily used for body measurements. This part can be considered as [0, 1] in the upper body part. The upper body part includes the chest, waist, and hips. That is, when the body contour image is divided into a predetermined ratio, the neck means 0 and the hips mean 1.

[0129] The chest area refers to the [0.15 - 0.25] part of the upper body.

[0130] The underburst part refers to the [0.35 - 0.45] part of the upper body.

[0131] The waist area refers to the [0.65 - 0.75] part of the upper body.

[0132] The hip area refers to the [0.95 - 1.00] part of the upper body.

[0133] The above values ​​are not fixed values, and the control unit (130) may be set differently depending on the body contour image by referring to the memory (150). The memory (150) includes various body contour images and stores body ratios corresponding to each body contour image.

[0134] FIG. 14 is a drawing illustrating the measurement of a user's height according to an embodiment of the present invention.

[0135] Referring to FIG. 14, the body contour image (1300) includes a camera center (1310). The control unit (130) can measure the user's height using the camera center (1310).

[0136] The camera center (1310) represents the camera height as the distance between the ground and the camera lens.

[0137] For example, if the camera height is 175 cm, the control unit (130) controls the sensor unit (140) to measure the camera height. Based on the camera height, the control unit (130) measures the user's height.

[0138] If the camera height is shorter than the user's height, the control unit (130) determines the user's height by adding a) the camera height and b) the distance between the camera and the user's head.

[0139] For example, if the camera height is 175 cm and the distance between the camera and the user's head is 5 cm, the control unit (130) adds 175 cm and 5 cm to determine 180 cm as the user's height.

[0140] If the camera height is greater than the user's height, the control unit (130) determines the user's height by subtracting a) the camera height and b) the distance between the camera height and the user's head.

[0141] For example, if the camera height is 175 cm and the distance between the camera and the user's head is 5 cm, the control unit (130) subtracts 5 cm from 175 cm to determine the user's height as 170 cm.

[0142] FIG. 15 is a diagram illustrating two methods for measuring a user's waist circumference according to one embodiment of the present invention. FIG. 15 includes FIG. 15(a) and FIG. 15(b).

[0143] FIG. 15(a) is a drawing capturing a front image of the user. FIG. 15(b) is a drawing capturing a rear image of the user.

[0144] The first method is when both the front image and the rear image are captured. In this case, the control unit (130) can measure the circumference most accurately based on the depth map. Additionally, since the ratio of body size and depth information can be known based on the skeleton image, the control unit (130) can measure the circumference more accurately.

[0145] The second method involves capturing only the front image. The back circumference can be calculated using an estimation method. In other words, the back circumference is related to the straight-line length of the body. Based on experimental data, the back circumference can be calculated by multiplying the straight-line length of the body by a specific parameter.

[0146] A detailed explanation regarding this will be provided later in FIGS. 16 and FIGS. 17.

[0147] FIG. 16 is a drawing illustrating the measurement of waist circumference length according to an embodiment of the present invention. FIG. 16 includes FIG. 16(a) and FIG. 16(b).

[0148] In the case of the present invention, only the measurement of waist circumference is described as one embodiment, but chest circumference and hip circumference can also be measured in the same way.

[0149] FIG. 16(a) is a drawing showing the waist circumference length area of ​​the user in the overall user body contour image. FIG. 16(b) is a drawing showing the waist circumference length area of ​​the user in detail.

[0150] Referring to Fig. 16 (a), the waist circumference length (1510) = front waist circumference length (10) + side waist circumference length (20, 40) + back waist circumference length (30).

[0151] As shown in FIG. 16(b), an enlarged image including the front waist circumference length (10), the side waist circumference lengths (20, 22), and the rear waist circumference length (30) is explained.

[0152] In the case of the front waist circumference length (10), the control unit (130) separates the front waist circumference into n straight lines and adds the lengths of the n straight lines to obtain the front waist circumference length (10).

[0153] The control unit (130) uses data obtained from the depth camera in which the noise is greater than a predetermined range for the calculation of the side waist circumference length (20). The control unit (130) uses data in which the noise is less than a predetermined range for the calculation of the front waist circumference length (10).

[0154] In the case of the present invention, data with noise below a predetermined range is clear data and can be used for the front waist circumference length, and data with noise above a predetermined range is unclear data and can be used to measure the side waist circumference length, thereby allowing for more accurate measurement of body size.

[0155] For the left side waist circumference length (20), the length of the longest side (22) is obtained using a right triangle, and the length of the obtained side (22) is multiplied by two to obtain the left side waist circumference length (20).

[0156] The circumference of the right side of the waist (40) is calculated in the same way as the circumference of the left side of the waist (20).

[0157] According to the present invention, in the case of a depth camera, noise occurs and the boundary line of the body is not accurate. In such cases, there is an advantage in that the lateral circumference length can be predicted using a right triangle.

[0158] The control unit (130) extracts a first length (32) from a body contour image and determines the back waist circumference length (30) by multiplying the extracted first length (32) by a specific parameter.

[0159] For example, the rear perimeter length (30) is calculated using the first length xk parameter. Here, the k parameter represents experimental data based on actual measurements.

[0160] The k-parameter varies depending on the body part. For example, the k-parameter of the waist is smaller than that of the chest.

[0161] The first length (32) refers to the length when the starting point and the ending point of the front waist circumference length (10) are connected in a straight line.

[0162] The control unit (130) determines the user's front waist circumference length (10), side waist circumference length (20, 40), and back waist circumference length (30) based on the generated body contour image, and measures the user's body size, specifically the waist circumference length, by combining the front waist circumference length (10), side waist circumference length (20, 40), and back waist circumference length (30).

[0163] According to the present invention, chest circumference and hip circumference can also be measured in the same way as above.

[0164] FIG. 17 is a drawing illustrating the measurement of front waist circumference, side waist circumference, and rear waist circumference according to an embodiment of the present invention. FIG. 17 includes FIG. 17(a) and FIG. 17(b).

[0165] FIG. 17(a) is a drawing specifically illustrating a method for calculating the side waist circumference. Referring to FIG. 17(a), the control unit (130) calculates the side waist circumference based on a body contour image (1610).

[0166] First, the method for determining the side waist circumference (20) is explained. The side waist circumference is determined using a right triangle (1620). The right triangle (1620) includes a first side (1621), a opposite side (1622), and a first angle (1623).

[0167] The main point is to determine the actual length of the opposite side (1622) of the right triangle (1620). First, the control unit (130) uses data obtained from the depth camera in which the noise is greater than a predetermined range for the calculation of the side waist circumference length (20). That is, the side waist circumference length is determined based on the noise.

[0168] For example, if the noise is greater than a predetermined range, the control unit (130) calculates the length of the first side (1621). The length of the first side (1621) can be 10 pixels.

[0169] The control unit (130) can estimate the first angle (1623) based on experimental data. For example, the first angle can be 15 degrees.

[0170] The control unit (130) can calculate the actual length of the side (1622) based on the first side (1621) and the first angle (1623). The actual length of the side (1622) is obtained by multiplying the side length (number of pixels) by the pixel length per meter (PPM, Pixel per meter).

[0171] FIG. 17(b) is a detailed drawing of the waist circumference length area by cutting the user's waist. Referring to FIG. 17(b), the control unit (130) calculates the side waist circumference length (20, 40) based on the body contour image (1610). The waist circumference length (1630) = front waist circumference length (10) + side waist circumference length (20, 40) + rear waist circumference length (30).

[0172] A method for calculating the back waist circumference (30) is explained further.

[0173] The control unit (130) extracts a first length (32) from a body contour image and determines the back waist circumference length (30) by multiplying the extracted first length (32) by a specific parameter.

[0174] For example, the back circumference length (30) is calculated using the first length xk parameter. Here, the k parameter refers to experimental data based on actual measurement data. The first length (32) refers to the length when the starting point and the ending point of the front waist circumference length (10) are connected by a straight line.

[0175] The back waist circumference length (30) can be considered as half the circumference length of the ellipse. The circumference length of the ellipse is proportional to the length of the major axis. Also, the circumference length of the ellipse is proportional to the length of the minor axis. Half of the back waist circumference length (30) can be the product of the first length (32) and the k parameter. Here, the k parameter is determined differently depending on the body part. For example, the k parameter of the waist part is smaller than the k parameter of the chest part. Also, the k parameter is experimental data obtained based on actual measurement data.

[0176] FIG. 18 is a diagram illustrating data regarding a user's actual key and a measured key according to an embodiment of the present invention.

[0177] Referring to Fig. 18, if the user is CHC, the actual height is 1.78 m. The actual height refers to the value measured using a measuring tape.

[0178] The first experimental value is 1.78202 m.

[0179] The second experimental value is 1.77908 m.

[0180] The third experimental value is 1.76101 m.

[0181] The fourth experimental value is 1.79096 m.

[0182] The fifth experimental value is 1.79234 m.

[0183] In this way, up to the 10th experimental value can be measured, and if the average of the error from the 1st experimental value to the 10th experimental value is calculated, it becomes 0.015 m.

[0184] Users can also measure UEN, ZF, WJU, and PSO in the same way.

[0185] FIG. 19 is a drawing showing data on a user's actual chest circumference and measured chest circumference according to an embodiment of the present invention.

[0186] Referring to Fig. 19, if the user is CHC, the actual chest circumference is 0.98 m. The actual chest circumference refers to the value measured using a measuring tape.

[0187] The first experimental value is 1.04009 m.

[0188] The second experimental value is 1.02241 m.

[0189] The third experimental value is 1.00679 m.

[0190] The fourth experimental value is 1.01789 m.

[0191] The fifth experimental value is 1.01635 m.

[0192] In this way, up to the 10th experimental value can be measured, and if the average of the error from the 1st experimental value to the 10th experimental value is calculated, it becomes 0.032 m.

[0193] Users can also measure UEN, ZF, WJU, and PSO in the same way.

[0194] FIG. 20 is a drawing illustrating data on the actual underburst circumference of a user and the measured underburst circumference according to an embodiment of the present invention.

[0195] Referring to Fig. 20, if the user is CHC, the actual underburst perimeter length is 0.88 m. The actual underburst perimeter length refers to the value measured using a measuring tape.

[0196] The first experimental value is 0.959572 m.

[0197] The second experimental value is 0.960445 m.

[0198] The third experimental value is 0.885358 m.

[0199] The fourth experimental value is 0.869253 m.

[0200] The fifth experimental value is 0.903299 m.

[0201] In this way, up to the 10th experimental value can be measured, and if the average of the error from the 1st experimental value to the 10th experimental value is calculated, it becomes 0.040 m.

[0202] Users can also measure UEN, ZF, WJU, and PSO in the same way.

[0203] FIG. 21 is a drawing illustrating data on a user's actual waist circumference and measured waist circumference according to an embodiment of the present invention.

[0204] Referring to Fig. 21, if the user is CHC, the actual waist circumference is 0.92 m. The actual waist circumference refers to the value measured using a tape measure.

[0205] The first experimental value is 0.985915 m.

[0206] The second experimental value is 0.939380 m.

[0207] The third experimental value is 0.929100 m.

[0208] The fourth experimental value is 0.910563 m.

[0209] The fifth experimental value is 0.914214 m.

[0210] In this way, up to the 10th experimental value can be measured, and if the average of the error from the 1st experimental value to the 10th experimental value is calculated, it becomes 0.025 m.

[0211] Users can also measure UEN, ZF, WJU, and PSO in the same way.

[0212] FIG. 22 is a drawing illustrating data on a user's actual hip circumference and measured hip circumference according to an embodiment of the present invention.

[0213] Referring to Fig. 22, if the user is CHC, the actual hip circumference is 0.99 m. The actual hip circumference refers to the value measured using a measuring tape.

[0214] The first experimental value is 1.09757 m.

[0215] The second experimental value is 1.06528 m.

[0216] The third experimental value is 1.06060 m.

[0217] The fourth experimental value is 1.04748 m.

[0218] The fifth experimental value is 1.04226 m.

[0219] In this way, up to the 10th experimental value can be measured, and if the average of the error from the 1st experimental value to the 10th experimental value is calculated, it becomes 0.039 m.

[0220] Users can also measure UEN, ZF, WJU, and PSO in the same way.

[0221] FIG. 23 is a diagram illustrating data regarding the error and accuracy of user body size measurement data according to one embodiment of the present invention.

[0222] Referring to Fig. 23, the error rate is explained.

[0223] In the case of the key, it becomes 0.6%.

[0224] In the case of chest circumference, it becomes 6.1%.

[0225] For the underburst circumference, it becomes 4.9%.

[0226] In the case of waist circumference, it becomes 2.3%.

[0227] In the case of hip circumference, it becomes 2.0%.

[0228] Explain accuracy. Accuracy refers to the value obtained by subtracting the error rate from 100%.

[0229] In the case of height, it becomes 99.4%.

[0230] In the case of chest circumference, it becomes 93.9%.

[0231] For the underburst circumference, it becomes 95.1%.

[0232] In the case of waist circumference, it becomes 97.7%.

[0233] In the case of hip circumference, it becomes 98.0%.

[0234] FIG. 24 is a diagram illustrating data regarding the error and accuracy of user body size measurement data according to one embodiment of the present invention.

[0235] Referring to Fig. 24, if the user is UEK, the actual height is 1.53 m. The actual height refers to the value measured using a measuring tape.

[0236] The first experimental value is 1.52192 m.

[0237] The second experimental value is 1.53040 m.

[0238] The third experimental value is 1.54128 m.

[0239] The fourth experimental value is 1.53899 m.

[0240] The fifth experimental value is 1.54272 m.

[0241] Calculating the average of the errors from the first experimental value to the fifth experimental value yields 0.00933063 m. Shoulder length and arm length can also be measured in the same way as above.

[0242] Measurements can be taken in the same way even if the user is KBR, ZU, or CHC.

[0243] Next, the error rate will be explained.

[0244] In the case of the key, it becomes 0.2%.

[0245] In the case of shoulder length, it becomes 1.6%.

[0246] In the case of arm length, it becomes 2.5%.

[0247] Explain accuracy. Accuracy refers to the value obtained by subtracting the error rate from 100%.

[0248] In the case of *key, it becomes 99.8%.

[0249] In the case of shoulder length, it becomes 98.4%.

[0250] In the case of arm length, it becomes 97.5%.

[0251] FIG. 25 is a flowchart relating to a control method of a body measurement device according to one embodiment of the present invention.

[0252] Since the body measurement device is based on RGB images and depth images, it is difficult to accurately measure body size when the user is wearing clothes, especially when wearing loose clothes that do not cling to the body (hereinafter referred to as a loose fit state).

[0253] In addition to clothing, it is also difficult to obtain the user's actual height while taking these noises into account when the user has hair volume at a significant height and is wearing shoes.

[0254] Therefore, it is necessary to correct the body size before correction, that is, the body size simply obtained from RGB images and depth images, to obtain the body size after correction, that is, the actual body size.

[0255] As described above, a process is required to correct the acquired body height to the actual body height, a process to correct the acquired body circumference to the actual body circumference, or a process capable of directly estimating the actual body height or actual body circumference from other factors.

[0256] As shown in the process S210 to S240 of FIG. 2, a user's body contour image can be generated (S2510 to S2540). As described above, if the user's pose estimated based on the first image is the first pose, a second image is captured, and the body size is estimated using the captured second image.

[0257] The body contour image obtained through processes S2510 to S2540 corresponds to the body size before correction, which includes noise caused by hair volume, shoe height, loose clothing, etc. Therefore, a process to remove such noise, or a process to derive a body size that does not require correction based on a noise-free body size, is necessary.

[0258] RGB images and depth images, and furthermore, body contour images and skeleton images derived from these images, may be used to correct body size. The body contour images and skeleton images are defined as source images. Based on the source images, body sizes including the user's actual height and actual circumference can be obtained. The actual body size may be obtained using both the body contour images and the skeleton images, or by using only one of them.

[0259] As shown in FIG. 2, a first image may be obtained through a camera and a second image may be obtained when it corresponds to a key pose to measure body size, but if necessary, body size may be measured using the first image already obtained when it corresponds to a key pose.

[0260] FIG. 26 is a conceptual diagram showing face information (410) for obtaining a hair volume value (421) according to one embodiment of the present invention.

[0261] The control unit (130) obtains the user's hair volume value (421) from the source image and can correct the body height before correction to the body height after correction using the obtained hair volume value (421).

[0262] Obtaining the hair volume value (421) means that the user's crown (422) point can be identified. That is, the body height after correction represents the height to the crown (422) obtained by subtracting the hair volume value (421) from the body height before correction.

[0263] The hair volume value (421) can be obtained based on the face information (410) of the source image. This means that the height of the crown of the head (422) can be determined probabilistically through the obtained face information (410).

[0264] In particular, the face information (410) from which the hair volume value (421) can be obtained can be calculated based on the distance from the nose (423) to the ear (424). When a point on the nose (423) that meets the horizontal line passing through the ear (424) is denoted as G, statistically, the distance R from G to the crown of the head (422) can be the value obtained by multiplying the distance r from G to the ear (424) by a specific constant. For example, the specific constant is 1.35.

[0265] The coordinates of the nose (423) and ears (424) can be obtained by the control unit (130) through an RGB image, a depth image, or a skeleton image. At this time, the coordinates of the nose (423) and ears (424) may be required that the face information (410) of the source image be frontal face information.

[0266] FIG. 27 is a conceptual diagram of another method for obtaining a hair volume value (421) according to one embodiment of the present invention.

[0267] As another method for determining the hair volume value (421), the face shape (432) may be used as a basis for determining the hair volume. Statistically, if the face shape (432) is determined, there is a high probability that the shape and length up to the height of the crown of the head (422) will also be fixed. In this case, the face shape (432) can be determined by the jawline shape (431).

[0268] Accordingly, when the acquired face information (410) includes the user's jawline shape (431), the control unit (130) determines which of the plurality of previously stored jawline shapes (431) the jawline shape (431) corresponds to, as shown in FIG. 27(b), and can acquire a hair volume value (421) by reflecting the height of the crown (422) of the face shape having the corresponding jawline shape (431).

[0269] Fig. 27(b) shows nine face types (432) as examples, but depending on the case, fewer or more samples may be stored.

[0270] Furthermore, the height of the crown of the head (422) according to the jawline shape (431) may take into account the size of the face. Even if the jawline shape (431) or the face shape corresponds to previously stored information, the height of the crown of the head (422) may differ if the face size is different. Therefore, the height value of the crown of the head (422) can be obtained by reflecting the face size factor. For example, if the ratio of the face size obtained from the source image to the previously stored corresponding face size is 0.8:1.0, the actual height value of the crown of the head (422) can be obtained by multiplying the previously stored height of the crown of the head (422) by 0.8.

[0271] The ratio of face size may also be calculated as the ratio of the distance values ​​from the nose (423) to the ear (424) in Fig. 26 described above.

[0272] FIG. 28 is a skeleton image (441) and a partial enlarged view for body height correction according to one embodiment of the present invention.

[0273] As described above, body height can be adjusted by taking into account shoe height (445).

[0274] The shoe height value (445) can be obtained specifically based on the skeleton image (441). The control unit (130) can obtain the shoe height value (445) through the difference between the distance (446) from the lateral point (442) of the skeleton image (441) to the floor and the distance (444) from the lateral point (442) to the sole of the foot (443).

[0275] The actual user's body height can be estimated by correcting the body height value before correction by reflecting the hair volume value (421) and shoe height value (445) obtained through the methods described above.

[0276] FIG. 29 is an experimental value regarding body height estimation according to an embodiment of the present invention. FIG. 29(a) shows the error of the factor used in body height estimation regarding hair volume and the estimated value, and FIG. 29(b) is an explanation of the factor used in body height estimation.

[0277] In addition to the method of correcting body height through hair volume value (421) and shoe height value, body height can also be estimated through specific factors among the acquired pre-correction body size.

[0278] That is, the control unit (130) can obtain the body height after correction by applying the actual measurement value obtained from the source image to a specific function.

[0279] The relationship to be substituted is as follows.

[0280] y i = X i T β + ε i

[0281] (y i : Actual height , X i T : Actual measured value, β: Specific gravity, ε i : Noise )

[0282] For example, the control unit (130) can obtain an actual height value by substituting the length value from the shoulder to the center of the hand of the skeleton image as an actual measurement value. The actual measurement value may be the user's chest value obtained from the source image, neck-shoulder-elbow-waist value, back-shoulder-neck intersection value, neck-to-gluteal value, natural waist value, maximum hip value, natural waist increase value, upper arm value, and waist value. Therefore, the actual height can be estimated by substituting at least one of the above values ​​into the above relationship.

[0283] Empirically, it can be confirmed that the error is smallest when the length value from the shoulder to the center of the hand is included.

[0284] Refer again to FIG. 25. Body height can be obtained through the method described above. And length-related values, that is, length values ​​among body sizes, can be obtained with almost the same value regardless of whether clothes are worn. This can be obtained through the source images described above, namely RGB images, depth images, body contour images, and skeleton images. Length values ​​may include, for example, shoulder length, arm length, and leg length.

[0285] On the other hand, values ​​related to circumference—specifically, circumference values ​​among body measurements—have a large margin of error when obtained from source images when clothing is worn. Therefore, circumference values ​​can be estimated from relatively accurate length values.

[0286] Circumference values ​​include chest circumference, waist circumference, and hip circumference among body sizes.

[0287] The derivation of circumference values ​​from length values ​​can be estimated using a probabilistic method or a deep learning method. For example, the circumference value corresponding to the highest probability of the input length value in the accumulated information can be derived. This accumulated information may be stored in advance. Alternatively, in the case of a deep learning method, a result circumference value is derived based on the input of length values, and this can be identified as a body size.

[0288] FIG. 30 is a diagram illustrating the structure of a camera of a body measurement device according to one embodiment of the present invention.

[0289] Referring to FIG. 30, the camera (110) includes an RGB camera (111), a first IR camera (112), an IR light (113), and a second IR camera (114).

[0290] Here, the camera (110) is installed vertically on the body measurement device (100). The camera (110) must be installed vertically so that when capturing a person image, a depth image can be generated with greater three-dimensionality.

[0291] When using a camera (110), three-dimensional depth information in the form of a map can be output, and compared to stereo vision technology, there is less noise due to changes in lighting, and since it is not a result of image processing, there is an advantage of not having textureless or occlusion.

[0292] FIG. 31 is a drawing illustrating an example of use of a body measurement device according to an embodiment of the present invention. FIG. 31 includes FIG. 31(a) and FIG. 31(b).

[0293] FIG. 31(a) is a drawing illustrating a body measurement device (100) mounted on a stylus and a user (200) standing in front of the body measurement device (100). As shown in FIG. 31(a), the camera (110) of the body measurement device (100) captures a full-body image of the user and measures the body size.

[0294] FIG. 31(b) is a drawing illustrating a body measuring device (100) mounted on a tailor bot in a clothing store, with a user (200) standing in front of the body measuring device (100). Referring to FIG. 31(b), the camera (110) of the body measuring device (100) captures an image of the user's upper body or the entire body to measure the user's body size.

[0295] The tailor bot is a guide robot that recommends clothes to the user (200) and performs virtual try-ons, etc., and can measure body size and perform virtual fitting using an avatar corresponding to the user on the screen.

[0296] FIG. 32 is a drawing illustrating an execution screen of a body measurement device according to an embodiment of the present invention. FIG. 32 includes FIG. 32(a) and FIG. 32(b).

[0297] FIG. 32(a) is a diagram illustrating the execution screen of a body measurement device. Referring to FIG. 32(a), when the body measurement device (100) is executed, the control unit (130) displays the current time 2019. 01. 14, 03:45 PM and an important meeting schedule.

[0298] FIG. 32(b) is a diagram showing the execution of a body measurement process when a user is in front of a body measurement device (100).

[0299] When the user stands in front of the body measurement device (100), the control unit (130) displays text (2620) containing instructions for the user to stand upright and spread their arms. Additionally, the control unit (130) displays an image (2430) indicating the execution steps of the body measurement process, for example, 64%. Here, 0% means that the body measurement process has started, and 100% means that the body measurement process has been completed.

[0300] FIG. 33 is a drawing illustrating an execution screen of a body measurement device according to one embodiment of the present invention.

[0301] Referring to FIG. 33, the control unit (130) displays an avatar (10) having the same body size as the user's body size on the screen.

[0302] For example, the user's body size is height 172 cm, chest circumference 101 cm, waist circumference 94 cm, and hip circumference 102 cm. Here, the clothing can be pants.

[0303] The control unit (130) determines the most suitable clothing size based on the user's body size and displays text (2710) recommending size M clothing to the user.

[0304] The size can be L, M, or S, and when the user selects a specific size icon (20), the control unit (130) displays an avatar (10) wearing clothes corresponding to the size icon (20) selected by the user. Here, the clothes can be pants.

[0305] When the control unit (130) receives an input (50) from the user selecting a fitting icon (40), the control unit displays an image of the user wearing pants. This is described later in FIG. 34.

[0306] FIG. 34 is a drawing illustrating an execution screen of a body measurement device according to one embodiment of the present invention.

[0307] FIG. 34 is a drawing illustrating the wearing state when a user wears pants of a specific size. The control unit (130) displays a progressive bar (2810) indicating the wearing state. For example, the specific size may be size M.

[0308] The progressive bar (2810) can change from a first color to a second color, where the first color indicates that the clothing worn is tight on the user and the second color indicates that the clothing worn is loose on the user. The color change from the first color to the second color can change as a gradation effect.

[0309] The control unit (130) displays a graphic image (20) indicating the wearing status for each body part on the avatar image (10) when the user avatar (10) wears pants. The graphic image (20) may appear differently depending on the wearing status for each body part.

[0310] For example, if the user avatar (10) is wearing pants, a graphic image (20) containing a color indicating looseness can be displayed on the upper part of the pants, and a graphic image (30) containing a color indicating tightness can be displayed on the lower part of the pants.

[0311] According to the present invention, when a user avatar (10) wears clothes of a specific size, a graphic image showing the wearing status for each body part is displayed so that the user can intuitively know whether the clothes are tight or loose without wearing them directly, thereby improving user convenience.

[0312] According to one embodiment of the present invention, when a user assumes a specific pose, an image of the user is captured, and an RGB image and a depth image are combined from the captured image to generate a body contour image of the user, and the user's body size can be measured based on the generated body contour image, so that the user's image can be captured and the body size can be accurately measured with a simple motion, thereby improving user convenience.

[0313] According to another embodiment of the present invention, a skeleton image can be extracted from a captured image, and a user's body contour image can be generated by combining the skeleton image, RGB image, and depth image, thereby allowing the user's body size to be measured more accurately and thus improving user convenience.

[0314] According to another embodiment of the present invention, a first length can be extracted from a body contour image and a rear waist circumference length can be determined by multiplying the first length by a specific parameter, thereby accurately measuring the waist circumference length that is not visible in the front image, thus improving user convenience.

[0315] The body measurement device and the control method according to the present invention are not limited to the configurations and methods of the embodiments described above; rather, all or part of each embodiment may be selectively combined to allow for various modifications to be made.

[0316] Meanwhile, the method of operation of the body measurement device of the present invention can be implemented as a processor-readable code on a processor-readable recording medium equipped in the body measurement device. A processor-readable recording medium includes all types of recording devices in which data that can be read by a processor is stored. Examples of processor-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc., and also include implementation in the form of a carrier wave, such as transmission via the Internet. Furthermore, the processor-readable recording medium may be distributed across networked computer systems, so that the processor-readable code can be stored and executed in a distributed manner.

[0317] Furthermore, although preferred embodiments of the present invention have been illustrated and described above, the present invention is not limited to the specific embodiments described above. It is understood that various modifications can be made by those skilled in the art without departing from the essence of the invention as claimed in the claims, and such modifications should not be understood individually from the technical spirit or perspective of the present invention. Explanation of the symbols

[0318] 410: Face Information 421: Hair volume value 422: Crown of the head 423: Co 424: Ear 431: Jawline shape 432: Face shape 441: Skeleton Image 442: Lateral copy point 443: Soles of the feet 444: Distance from the lateral malleolus to the sole of the foot 445: Shoe height 446: Distance from the lateral radiant point to the floor 451: Floor

Claims

Claim 1 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires a user’s hair volume value based on frontal face information of the source image, and corrects the body height before correction to the body height after correction through the hair volume value, wherein the frontal face information includes coordinate information of the nose and ears, and the control unit acquires the hair volume value as a distance value from the nose to the ears. Claim 2 A body measurement device according to claim 1, wherein the hair volume value is obtained by multiplying the distance value from the nose to the ear by a specific constant. Claim 3 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires a user’s hair volume value based on face information of the source image, and corrects the body height before correction to the body height after correction through the hair volume value, wherein the face information includes the user’s jawline shape, and the control unit acquires the hair volume value based on the user’s jawline shape. Claim 4 In claim 3, the control unit determines which of the nine previously stored jawline shapes the user's jawline shape corresponds to, and the body measurement device obtains the hair volume value based on the information of the corresponding jawline shape. Claim 5 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires a user’s shoe height value from the skeleton image among the source images and corrects the body height before correction to the body height after correction through the shoe height value. Claim 6 In claim 5, the control unit is a body measurement device that obtains the shoe height value through the difference between the distance from the lateral point of the skeleton image to the floor and the distance from the lateral point to the sole of the foot. Claim 7 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires length values ​​including shoulder length, arm length, and leg length among the body sizes from the source image, acquires circumference values ​​including the user’s chest circumference, waist circumference, and hip circumference from the acquired length values ​​and the body height after correction, and acquires the circumference values ​​based on a list of circumference values ​​according to the length values. Claim 8 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires length values ​​including shoulder length, arm length, and leg length among the body sizes from the source image, and acquires circumference values ​​including the user’s chest circumference, waist circumference, and hip circumference from the acquired length values ​​and the body height after correction, wherein the acquired circumference values ​​are based on deep learning results of circumference values ​​according to length values. Claim 9 A body measurement device comprising: a camera for capturing an image including an RGB image and a depth image; a display; and a control unit for acquiring a source image including a user’s body contour image and a skeleton image based on the captured image, acquiring a body size including a body height based on the source image, and controlling the display to display the body size, wherein the body height includes a body height before correction and a body height after correction, and the control unit acquires a length value from the user’s shoulder to the center of the hand from the source image and corrects the body height after correction through the length value from the shoulder to the center of the hand. Claim 10 In claim 9, the control unit obtains the user's chest value, neck-shoulder-elbow-waist value, back-shoulder-neck cross value, neck-to-gluteal value, natural waist value, maximum hip value, natural waist increase value, upper arm value, and waist value from the source image, and corrects the body height after correction through the obtained values ​​and the length value from the shoulder to the center of the hand. Claim 11 delete Claim 12 delete Claim 13 delete Claim 14 delete Claim 15 delete Claim 16 delete Claim 17 delete Claim 18 delete Claim 19 delete Claim 20 delete

Images