Image generation device, virtual try-on system, image generation method, program, and storage medium

Abstract

To provide an image generation device capable of visually confirming feelings of a future clothing size.SOLUTION: An image generation device (200) includes deformation unit (250) for comparing a future body size of a person with a clothing size to change the scale of at least one of the clothing image and a person image, and an image generation unit (260) for overlapping the clothing image on the person image to generate a virtual try-on image.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an image generating device, a virtual try-on system, an image generating method, a program, and a storage medium. [Background technology]

[0002] When considering purchasing clothes for their children, customers want to try them on to determine whether they fit their child's body type and look good. However, having a child try on multiple clothes takes time and effort, which can be a burden for customers. In addition, when purchasing clothes for their children, customers want to be able to estimate how long the clothes will last before making a purchase.

[0003] 2. Description of the Related Art A virtual try-on system is known that displays a state in which a customer who is considering purchasing clothing virtually tries on the clothing he or she is considering, thereby assisting the customer in making a selection.

[0004] Patent Document 1 discloses an image generating device that generates a clothing image by changing the scale of a clothing image according to the body dimensions of a human body image, and displays a composite image in which the clothing image is superimposed on the human body image. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2022-106923 Summary of the Invention [Problem to be solved by the invention]

[0006] However, the technique disclosed in Patent Document 1 allows one to check the current "feeling of clothing length" of a child, but does not allow one to visually check the future "feeling of clothing length."

[0007] Therefore, an object of the present invention is to provide an image generating device that makes it possible to check the future feeling of length of a garment. [Means for solving the problem]

[0008] An image generating device according to one aspect of the present invention includes an image transformation unit that changes the scale of at least one of a clothing image or a person image by comparing a person's future body dimensions with clothing dimensions, and an image generation unit that overlays the clothing image on the person image to generate a virtual fitting image.

[0009] Other objects and features of the present invention will be described in the following embodiments. Effect of the Invention

[0010] According to the present invention, it is possible to provide an image generating device that allows the user to visually confirm the future expected length of a garment. [Brief description of the drawings]

[0011] [Figure 1] 1 is a block diagram of a virtual try-on system according to an embodiment of the present invention. [Diagram 2] FIG. 11 is a schematic diagram showing a result of estimation by a body part estimation unit in the present embodiment. [Diagram 3] 13 is a flowchart showing a processing procedure of a post-elapsed body measurement acquisition unit in the present embodiment. [Figure 4] 11 is a schematic diagram of body dimensions used in the processing of the post-elapsed body dimension acquisition unit in the embodiment. FIG. [Diagram 5] 13 is a flowchart showing a processing procedure of a post-lapse body measurement acquisition unit in the case where a past body measurement is used in the present embodiment. [Figure 6] 5A to 5C are explanatory diagrams of the processing of an image transformation unit in the present embodiment. [Figure 7] 5A to 5C are explanatory diagrams of the processing of an image generating unit in the present embodiment. [Figure 8] 4 is a flowchart showing a processing procedure of the virtual try-on system in the present embodiment. [Figure 9] 4 is a schematic diagram of a virtual fitting image in the present embodiment. FIG. [Figure 10] 5A to 5C are schematic diagrams of virtual try-on images in each embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0013] First, the configuration of a virtual try-on system 1000 in this embodiment will be described with reference to Fig. 1(a) and (b). Fig. 1(a) is a block diagram showing the overall configuration of the virtual try-on system 1000. The virtual try-on system 1000 is configured to include a user terminal 100, an image generating device 200, and a database 300.

[0014] The database 300 includes a clothing database 320 that holds data on a plurality of clothing images, different sizes of each clothing image, and clothing dimensions for each size, and a statistical database 310 that includes a probability distribution of body dimensions for both men and women from birth to a predetermined age. In this embodiment, the predetermined age is, for example, 20 years old, and the statistical database 310 has data on body dimensions for the age at which a child's growth stops. In this embodiment, the clothing dimensions refer to at least one of the following scales used to measure the dimensions of clothing: sleeve length, length, inseam, chest circumference, waist circumference, or hip circumference. In this embodiment, the clothing database 320 that records the above-mentioned clothing dimensions for each garment is held.

[0015] The user terminal 100 has a camera (image capture device) 110, an input unit 120, a clothing selection unit 130, and a display unit 140. The camera 110 captures an image of a subject person based on a user's instruction, and outputs an image of the subject person obtained by capturing the image. The input unit 120 is an operation means through which the user inputs. The clothing selection unit 130 allows the user to select one of a plurality of clothing images in the clothing database 320, and select one size of the selected clothing image. The display unit 140 displays a virtual try-on image generated by the image generation unit 260, which will be described later. The camera 110 may allow the user to capture a person other than the user as the subject person, or may capture the user himself as the subject person.

[0016] The image generating device 200 includes a distance information acquiring section 210, a body part estimating section 220, a subject person's current body dimension acquiring section 230, a subject person's future body dimension acquiring section (future body dimension acquiring section) 240, an image transforming section 250, and an image generating section 260. The distance information acquiring section 210 acquires three-dimensional distance information of the subject person based on the subject person image captured by the camera 110 and the shooting information. The body part estimating section 220 estimates a plurality of body parts of the subject person in the subject person image. In this embodiment, the body parts are at least one of the head, shoulders, hands, waist, crotch, and soles, but are not limited thereto. The current body dimension acquiring section 230 acquires the person's current body dimensions. The current body dimensions can be acquired by a calculation method based on the plurality of body parts estimated by the body part estimating section 220 and the three-dimensional distance information of the subject person acquired by the distance information acquiring section 210, but are not limited thereto.

[0017] 1(b) is a block diagram partially showing the input unit 120, the post-mortem body dimension acquisition unit 240, the image transformation unit 250, and the statistical database 310. The input unit 120 has a predicted date input unit 121, a photographed person's current body dimension input unit 122, an age input unit 123, a gender input unit 124, a past body dimension input unit 125, a post-mortem body dimension input unit (future body dimension input unit) 126, and a premature birth information input unit 127.

[0018] The predicted date input unit 121 inputs the date for which the user wishes to predict the growth of the photographed person. The current body dimension input unit 122 allows the user to input the current body dimensions of the photographed person. The age input unit 123 and gender input unit 124 allow the user to input the age and gender of the photographed person, respectively. In this embodiment, age may include age in months. The past body dimension input unit 125 is configured to allow the user to input at least one combination of the past body dimension of the photographed person and the age at that time. The post-lapse body dimension input unit 126 allows the user to input any body dimension (future body dimension). The premature birth information input unit 127 allows the user to input whether the photographed person is a premature baby or not, and if so, the expected date of birth.

[0019] In this embodiment, the post-elapsed body dimension acquisition unit 240 estimates the post-elapsed body dimensions (future body dimensions) of the photographed person after the predicted years and months have passed by at least one of two methods. One is a method of estimating the post-elapsed body dimensions from the current body dimensions, the age and predicted years and months acquired from the respective input units 120, and the probability distribution of the body dimensions of the photographed person's gender acquired from the gender input unit 124. The current body dimensions may be the current body dimensions calculated by the current body dimension acquisition unit 230, or may be the current body dimensions acquired from the current body dimension input unit 122. The other acquisition method is a method of estimating the post-elapsed body dimensions using the past body dimensions acquired from the past body dimension input unit 125 and the age at that time in addition to the current body dimensions.

[0020] The image transformation unit 250 changes the scale of at least one of the clothing image and the photographed person image based on a result of comparing the post-elapsed body dimensions with the clothing dimensions of the size of the clothing image selected by the user in the clothing selection unit 130. The post-elapsed body dimensions may be post-elapsed body dimensions estimated by the post-elapsed body dimension acquisition unit 240 or may be post-elapsed body dimensions acquired by the post-elapsed body dimension input unit 126. The image generation unit 260 generates a virtual fitting image by superimposing a clothing image on the person image based on the body part estimated by the body part estimation unit 220. In this embodiment, the current body dimensions and the post-elapsed body dimensions are at least one of height, sleeve length, back length, inseam, chest circumference, waist circumference, and hip circumference, but the present invention is not limited to these.

[0021] Next, the details of the individual blocks in FIG. 1(a) and (b) will be described. The distance information acquisition unit 210 acquires three-dimensional distance information of the subject person from the image captured by the camera 110 and the image capture information. Although the details of the acquisition method are not specified, for example, there is a method of estimating the three-dimensional distance of the subject person by machine learning. This can be realized, for example, by using images such as images and videos captured by the camera 110, optical flows obtained from these images, and sensor information such as a gyro sensor and an acceleration sensor built into the camera 110 as input values. Then, the estimation can be performed by using a learning device trained with a model in which the three-dimensional distance information of the subject is an output value. In addition, the defocus amount obtained from a focus detection means built into the camera 110 may be used. Alternatively, a sensor such as a LiDAR (Light Detection and Ranging) or a ToF (Time of Flight) sensor may be mounted on the user terminal 100 and the three-dimensional distance information from the sensor may be used. Also, a compound eye camera may be used as the camera 110 to acquire the three-dimensional distance information of the subject person from the triangulation result.

[0022] Next, the body part estimation unit 220 will be described with reference to FIG. 2. FIG. 2 is a schematic diagram showing the results of estimation by the body part estimation unit 220, showing the body parts of a subject person. The body part estimation unit 220 estimates a plurality of body parts of the subject person in the subject person image. For example, a method of estimating the body parts using a machine-learned learning device with a person image as an input value and a body part as an output value is considered, but is not limited to this. In this embodiment, as shown in FIG. 2, the body part estimation unit 220 estimates the positions of the head, shoulders, hands, waist, crotch, and soles in the subject person image.

[0023] The current body dimension acquisition unit 230 calculates the current body dimensions from the estimated body parts and the three-dimensional distance information of the photographed person. For example, the height, which is the current body dimension, is calculated from the distance information from the head to the soles of the feet of the photographed person.

[0024] Next, a method for estimating post-elapsed body dimensions by the post-elapsed body dimension acquisition unit 240 will be described in detail. First, a method for estimating post-elapsed body dimensions based on current body dimensions, age, predicted year and month, and probability distribution of body dimensions will be described with reference to Fig. 3 and Figs. 4(a) and (b). Fig. 3 is a flowchart showing the procedure for the post-elapsed body dimension acquisition unit 240 to estimate post-elapsed body dimensions, and corresponds to step S320 in Fig. 8 relating to the processing method of the virtual try-on system 1000 described later. Figs. 4(a) and (b) are schematic diagrams of body dimensions used in the processing of the post-elapsed body dimension acquisition unit 240, showing the probability distribution of body dimensions stored in the statistical database 310.

[0025] First, in step S321 in Fig. 3, the post-lapse body dimension acquisition unit 240 calls up from the statistical database 310 a probability distribution of body dimensions that match the age and sex of the photographed person input by the user via the input unit 120. Next, in step S322, the post-lapse body dimension acquisition unit 240 acquires the probability of the current body dimension in the called probability distribution. The method used here is not limited, but for example, in Fig. 4(a) where the horizontal axis is the body dimension and the vertical axis is the probability density, the area from the smallest body dimension to the current body dimension is calculated to acquire the probability.

[0026] Next, in step S323, the post-elapsed body dimension acquisition unit 240 calls up the probability distribution of the post-elapsed age obtained by adding the predicted years and months to the age and the body dimensions that match the gender of the photographed person from the statistical data. Next, in step S324, the post-elapsed body dimension acquisition unit 240 calculates the body dimensions at the probability calculated in step S322 in the probability distribution called up in step S323. For example, if the probability is calculated in step S322 as shown in FIG. 4(a), the body dimensions that result in the probability calculated in step S322 are calculated from the smallest body dimensions as shown in FIG. 4(b). Then, in step S324, the post-elapsed body dimension acquisition unit 240 outputs the calculated body dimensions as the post-elapsed body dimensions.

[0027] In this embodiment, the body dimensions calculated in step 323 are output as is in step 324, but this is not limiting, and any correction information for body dimensions from the user may be reflected. For example, the probability distribution to be called up may be changed in order to upwardly revise the growth prediction in accordance with the individual's tendencies, or a correction value for directly correcting each body dimension may be input via the input unit 120, and various body dimensions are corrected in response to the input (waist circumference +5%, inseam +4 cm, etc.).

[0028] Next, a method for estimating a post-elapsed body dimension using a past body dimension will be described with reference to Fig. 5. Fig. 5 is a flowchart showing the processing procedure in the post-elapsed body dimension acquisition unit 240 when using a past body dimension. Fig. 5 shows the procedure for estimating a post-elapsed body dimension using a past body dimension in the processing of step S320 in Fig. 8.

[0029] 3, the after-time body dimension acquisition unit 240 also calls up a probability distribution that the age and gender of the photographed person match at the time of each past body dimension acquired from the past body dimension input unit 125. When multiple past body dimensions are input, the probability distributions of all the corresponding ages are called up.

[0030] Next, in step S326, the post-lapse body dimension acquisition unit 240 acquires the probability of past body dimensions in the probability distribution of each age, similar to Figures 4(a) and (b), in addition to the process of step S322. When there are multiple combinations of past body dimensions and ages, as in the example shown in step S322, the probability of past body dimensions is acquired for the probability distribution of each age.

[0031] Next, in step S327, the post-elapsed body dimension acquisition unit 240 obtains a probability obtained by averaging the probability of the current body dimension obtained in step S326 and the probability of the past body dimension. Next, in step S328, the post-elapsed body dimension acquisition unit 240 calls the probability distribution of the body dimension that matches the post-elapsed age and gender from the statistical database 310, as in step S323. Next, in step S329, the post-elapsed body dimension acquisition unit 240 obtains the body dimension at the probability obtained in step S327 in the probability distribution called in step S328, as in step S324. Then, the post-elapsed body dimension acquisition unit 240 outputs the body dimension obtained in step S329 as the post-elapsed body dimension. Note that in this embodiment, the average probability is obtained in step S327, but this is not limited to this. Also, in step S329, correction based on correction information from the user may be performed as in step S324.

[0032] The current age and the age corresponding to the past body dimension used in the above-mentioned processing of the post-elapsed body dimension acquisition unit 240 are input by the user based on the birthday of the photographed person. However, when there is an input to the premature birth information input unit 127, the post-elapsed body dimension acquisition unit 240 calculates the corrected age by using the expected birth date as the birthday, and replaces the age input to the age input unit 123 and the past body dimension input unit 125 with the corrected age. For example, when the current age acquired from the age input unit 123 is 0 years and 8 months, and the expected birth date acquired from the premature birth information input unit 127 is two months after the birthday, it is replaced with the current corrected age of 0 years and 6 months. In addition, the age corresponding to the past body dimension and the post-elapsed age are calculated by subtracting 2 months, which is the difference between the expected birth date and the birthday. The post-elapsed body dimension acquisition unit 240 estimates the post-elapsed body dimension using the corrected age replaced in this way.

[0033] The above-mentioned method for estimating post-elapsed body dimensions uses a probability distribution of body dimensions to estimate post-elapsed body dimensions, but this embodiment is not limited to this. For example, estimation using machine learning may be used. Furthermore, the input unit 120 is not limited to this, and a weight input unit for inputting the weight of the subject person may be provided, and the post-elapsed body dimensions may be estimated using the weight as well.

[0034] Next, the processing of the image transformation unit 250 will be described in detail with reference to Fig. 6. Fig. 6 is an explanatory diagram of the processing of the image transformation unit 250. The image transformation unit 250 changes the scale of at least one of the clothing image or the photographed person image. Fig. 6 describes a case where the scale of the clothing image is changed. When the image transformation unit 250 changes the scale of the clothing image, the scale factor is calculated by the following formula (1) using the post-lapse body dimensions, the image size of the photographed person image, the clothing dimensions, and the image size of the clothing image.

[0035] Scale factor = (Clothing size × Image size of subject person image) / (Clothing image size × Body size after progression) (1) The image size is the size of the image itself, and may be expressed, for example, by the number of pixels. The above calculation will be explained using the specific example of FIG. 6. The upper left of FIG. 6 is an image of a T-shirt, which is clothing, and the upper right is an image of a person, and both images are displayed at the same scale as image sizes. In the example of FIG. 6, the clothing dimensions for calculating the scale factor are the length of the T-shirt (length of clothing dimensions), and the image size of the clothing image is obtained as an image size equivalent to the length of the T-shirt (image size equivalent to the length of the clothing image). In addition, the post-elapsed body dimensions are the height (post-elapsed height) of the post-elapsed body dimensions of the person, and the image size of the person image is obtained as an image size equivalent to the height of the person (image size equivalent to the height of the person image). Therefore, the scale factor is calculated by the following formula (2).

[0036] Scale factor = (length of clothing × image size equivalent to height of person image) / (image size equivalent to length of clothing image × height after progression) (2) When the image transformation unit 250 calculates the scale factor, any type of clothing dimensions or post-elapsed dimensions may be selected. However, the image size of the clothing image must be the image size of the part corresponding to the type selected in the clothing dimensions, and the image size of the subject person image must be the image size of the part corresponding to the type selected in the post-elapsed body dimensions. The scale factor calculated in this way can be used to change the scale of the clothing image. In other words, the post-elapsed image size of the clothing image is calculated using the following formula (3).

[0037] Image size of clothing image after transformation = Image size of clothing image × Scale factor (3) The scale factor may be calculated by one representative type of clothing size and post-experience body size, or may be calculated by associating the clothing size with the post-experience body size type as shown in the lower table of FIG. 6. In an example of calculating the scale factor by one representative type of clothing size and post-experience body size, the image size of the entire clothing image is changed using the scale factor calculated by the above formula (2). On the other hand, in an example of calculating the scale factor by associating the clothing size with the post-experience body size type, the scale factor calculated by the clothing size length and the body size back length, which are in a corresponding relationship as shown in the lower table of FIG. 6, is used only for changing the image size equivalent to the length of the clothing image. The calculation of the scale factor using formula (2) and the process of changing the image size using formula (3) are performed on the other items, sleeve length, inseam, chest circumference, waist circumference, and hip circumference.

[0038] The above example describes the case where the image size of the clothing image is changed, but the image size of the subject person image may also be changed. In this case, the calculation formula for the scale factor is expressed as the following formula (4) by switching the denominator and numerator in formula (1).

[0039] Scale ratio = (image size of clothing image × body size after elapsed time) / (clothing size × image size of subject person image) (4) The transformed image size of the subject person image can be calculated using the scale factor of equation (4) according to the following equation (5).

[0040] Image size after transformation of subject person image=image size of subject person image×scale ratio (5) Next, the processing of the image generating unit 260 will be described in detail with reference to Fig. 7. Fig. 7 is an explanatory diagram of the processing of the image generating unit 260. As shown in Fig. 7, the image generating unit 260 generates a composite image by overlaying a clothing image on a subject person image based on the positions of body parts. For example, a T-shirt is composited to match the position of the subject person's shoulders, pants are composited to match the position of the crotch, and a skirt is composited to match the position of the waist.

[0041] Next, a processing method of the virtual try-on system 1000 will be described with reference to Fig. 8. Fig. 8 is a flowchart showing the processing procedure of the virtual try-on system 1000.

[0042] First, in step S101, the virtual try-on system 1000 photographs the subject person with the camera 110 and acquires a subject person image. Next, in step S102, the clothing selection unit 130 selects clothing. Here, the user selects clothing from the clothing database 320. Next, in step S103, the clothing selection unit 130 selects one size from among the sizes of the clothing selected by the user in step S102. Next, in step S104, the virtual try-on system 1000 determines whether or not a current body measurement has been input in the current body measurement input unit 122 of the user terminal 100. If it is determined that a current body measurement has been input, the process proceeds to step S105. On the other hand, if it is determined that a current body measurement has not been input, the process proceeds to step S210.

[0043] In step S210, the distance information acquisition unit 210 acquires three-dimensional distance information of the subject person. Then, in step S220, the body part estimation unit 220 estimates the body part of the subject person in the subject person image. Then, in step S230, the current body size acquisition unit 230 calculates (acquires) the current body size. Then, the process proceeds to step S105.

[0044] In step S105, the virtual try-on system 1000 judges whether the age input to the age input unit 123 is equal to or less than a predetermined age. If it is judged that the input age is equal to or less than the predetermined age, the process proceeds to step S106. On the other hand, if it is judged that the input age is not equal to or less than the predetermined age, the process proceeds to step S420.

[0045] In step S106, the virtual try-on system 1000 judges whether or not the post-metabolism body dimension has been input in the post-metabolism body dimension input unit 126. If it is judged that the post-metabolism body dimension has been input, the process proceeds to step S107. On the other hand, if it is judged that the post-metabolism body dimension has not been input, the process proceeds to step S310.

[0046] In step S310, the virtual try-on system 1000 judges whether or not a predicted date has been input to the predicted date input unit 121. If it is judged that a predicted date has been input, the process proceeds to step S320. In step S320, the post-elapsed body dimension acquisition unit 240 estimates the post-elapsed body dimensions of the subject person after the predicted date has passed from the present. Next, in step S107, the image transformation unit 250 calculates the scale factor using formula (1) or formula (4), and the process proceeds to step S108.

[0047] On the other hand, if it is determined in step S310 that the predicted date has not been input to the predicted date input unit 121, the process proceeds to step S420. If the predicted date has not been input, it is assumed that the user wishes to check the state of the current photographed person virtually trying on the clothing. In step S420, the virtual try-on system 1000 assumes that the predicted date is zero, replaces the post-elapsed body dimensions in formula (1) or formula (4) with the current body dimensions, calculates the scale factor, and proceeds to step S108.

[0048] In step S108, the image transformation unit 250 changes the image size of the photographed person image or the clothing image using formula (3) or formula (5) based on the scale factor calculated in step S107 or step S420. Then, in step S109, the image generation unit 260 generates a virtual try-on image by combining the photographed person image with the clothing image. Then, in step S110, the virtual try-on system 1000 displays the virtual try-on image on the display unit 140. Then, in step S111, the virtual try-on system 1000 determines whether or not there has been a change in input or selection in the input unit 120 or the clothing selection unit 130. If it is determined that there has been a change in input or selection, the process proceeds to step S101. On the other hand, if it is determined that there has been no change in input or selection, the process ends.

[0049] Next, the effect of the virtual try-on system 1000 will be described with reference to Figs. 9(a) and (b). Figs. 9(a) and (b) are schematic diagrams of virtual try-on images generated by the virtual try-on system 1000. In the virtual try-on system 1000, when a virtual try-on image is generated via a route that passes through step S420 in Fig. 8, the try-on image will be as shown in Fig. 9(a). When an input is made to the predicted date input unit 121 and a virtual try-on image is generated via a route that passes through step S107, the virtual try-on image will be as shown in Fig. 9(b). In the example of this figure, the image that is scaled by the image transformation unit 250 will be described as a clothing image.

[0050] First, Fig. 9(a) will be described. As described with reference to Fig. 8, if the subject person is not below the predetermined age in step S105, the display unit 140 will display as shown in Fig. 9(a). If the subject person is not below the predetermined age, it is determined that there is no need for growth prediction, and a virtual try-on image is generated using only the current body dimensions of the current subject person and presented to the user. Similarly, if there is no input from the predicted date input unit 121 in step S310, the display will be as shown in Fig. 9(a). This allows the user to check the state of the current subject person virtually trying on the clothing.

[0051] Next, FIG. 9(b) will be described. In the calculation of the scale factor using the formula (1) in the image transformation unit 250, the scale factor calculated using the current body size and the scale factor calculated using the body size after the lapse of time are larger than the current body size. Therefore, the scale factor when the body size after the lapse of time is used is smaller. Therefore, the clothing image in FIG. 9(b) is smaller than that in FIG. 9(a). Conventionally, there has been a technology such as that in FIG. 9(a), and the user was able to confirm the current "feeling of length" of the current subject person. On the other hand, in the virtual try-on system 1000 in this embodiment, the user can visually confirm what the feeling of length of the clothing will be after the subject person has passed the predicted years and months, as shown in FIG. 9(b). Furthermore, correction can be made by user input so that the subject person's appearance after the predicted years and months can be more accurately imagined.

[0052] The virtual try-on images in each embodiment will be described below. EXAMPLES

[0053] First, the virtual try-on image in the first embodiment will be described with reference to Fig. 10(a). Fig. 10(a) is an explanatory diagram of a display method on the display unit 140 of the virtual try-on system 1000 in this embodiment. In a calculation unit (not shown) of the virtual try-on system 1000, in the case of a route passing through step S107 in Fig. 8, a post-travel body dimension is calculated from the clothing dimension, and in the case of a route passing through step S420, a dimension is calculated by subtracting the current body dimension from the clothing dimension, and the result is displayed on the display unit 140. By displaying in this manner, the user can confirm how big the subject person will look when wearing clothing of the size selected in the clothing selection unit 130, and this can be used as a basis for deciding whether to purchase the clothing. EXAMPLES

[0054] Next, the virtual try-on image in the second embodiment will be described with reference to FIG. 10(b). FIG. 10(b) is an explanatory diagram of a display method on the display unit 140 of the virtual try-on system 1000 in this embodiment. When the sleeve length of the clothing image is longer than the sleeve length of the subject person, the image transformation unit 250 trims the cuff so that the sleeve length of the clothing image is the same as the sleeve length of the subject person. The display unit 140 also displays the measurement obtained by subtracting the sleeve length of the subject person from the sleeve length of the clothing image to inform the user how long the sleeve length is. Similarly, when the inseam of the clothing is longer than the inseam of the subject person, the hem is trimmed so that the inseam of the clothing is the same as the inseam of the subject person. The display unit 140 then displays the measurement obtained by subtracting the inseam length of the subject person from the inseam of the clothing to inform the user how long the hem is. By displaying in this manner, the user can confirm to what extent the cuffs and hems should be folded up when the photographed person wears clothing of the size selected in the clothing selection section 130, and can use this information as a basis for deciding whether to purchase clothing.

[0055] In each embodiment, the subject person image captured by the camera 110 is preferably a still image or a video, and more preferably a real-time video. When the virtual try-on system 1000 captures the subject person with the camera 110, the user may be guided by voice or text displayed on the display unit 140 so that the subject person in the subject person image is captured within a predetermined range, facing a predetermined direction, and taking a predetermined pose. This makes it easier to obtain the subject person's body dimensions and to generate a virtual try-on image. In addition, guidance may be given to the subject person to guide a predetermined movement, and a video may be captured by the camera 110, and a still image or video when the subject person is facing a desired direction may be cut out. In this case, although it is not a real-time video, it is easier to generate a virtual try-on image in the image generating device 200.

[0056] In each embodiment, preferably, the clothing image in the clothing database 320 is a clothing image configured as a three-dimensional model, and when the clothing image is synthesized with the subject person, a physics engine is used to configure the clothing to follow the movement of the subject person. This allows the user to confirm how the clothing will deform when the subject person moves. For example, when a T-shirt is selected with the clothing selection unit 130, it is possible to confirm whether the hem of the T-shirt will rise up and reveal the stomach when the subject person raises their arms. (Other Examples) The present invention can also be realized by a process in which a program for implementing one or more of the functions of the above-mentioned embodiments is supplied to a system or device via a network or a storage medium, and one or more processors in a computer of the system or device read and execute the program. It can also be realized by a circuit (e.g., ASIC) that implements one or more of the functions.

[0057] According to each embodiment, it is possible to provide an image generating device, a virtual try-on system, an image generating method, a program, and a storage medium that enable visual confirmation of the future garment length.

[0058] The disclosure of each embodiment includes the following configurations and methods. (Configuration 1) an image transformation unit that changes the scale of at least one of the clothing image and the person image by comparing the person's future body dimensions and clothing dimensions; and an image generating unit that generates a virtual fitting image by superimposing the clothing image on the person image. (Configuration 2) the person image is obtained by photographing the person with an imaging device, the future body dimensions are estimated using the person's current body dimensions; 2. The image generating device according to configuration 1, wherein the clothing dimensions correspond to the clothing image obtained from a clothing database. (Configuration 3) 3. The image generating device according to configuration 1 or 2, further comprising a current body dimension acquiring unit for acquiring current body dimensions of the person. (Configuration 4) The image generating device according to configuration 3, further comprising a distance information acquiring unit that acquires three-dimensional distance information of the person. (Configuration 5) a body part estimation unit that estimates a body part of the person in the person image, The image generating device described in configuration 4, wherein the current body dimension acquisition unit acquires the current body dimensions based on the body part estimated by the body part estimation unit and the distance information acquired by the distance information acquisition unit. (Configuration 6) The image generating device according to configuration 5, wherein the body parts include at least one of the person's head, shoulders, hands, waist, crotch, or soles of the feet. (Configuration 7) 7. The image generating device according to configuration 5 or 6, wherein the image generating unit generates the virtual fitting image by superimposing the clothing image on the person image based on the body part. (Configuration 8) The image generating device according to any one of configurations 2 to 7, further comprising a future body size acquisition unit that estimates the future body size based on the current body size, the age of the person, and the predicted year and month of the person. (Configuration 9) 9. The image generating device according to any one of configurations 1 to 8, wherein the future body dimensions are at least one of the person's height, sleeve length, back length, inseam, chest circumference, waist circumference, and hip circumference. (Configuration 10) 10. The image generating device according to any one of configurations 1 to 9, wherein the clothing dimensions are at least one of sleeve length, length, inseam, chest circumference, waist circumference, and hip circumference. (Configuration 11) An image generating device according to any one of configurations 1 to 10; and an imaging device for photographing the person and outputting the person image. (Configuration 12) 12. The virtual try-on system according to configuration 11, further comprising a current body measurement input unit for inputting the person's current body measurements. (Configuration 13) 13. The virtual try-on system according to configuration 11 or 12, further comprising a gender input unit for acquiring the gender of the person. (Configuration 14) an age input unit for acquiring the age of the person; A predicted date input unit for acquiring a predicted date of the person; a future body size acquisition unit that estimates the future body size based on the person's current body size, the person's age, and the person's predicted year and month; a statistical database including a probability distribution of body dimensions for each sex from birth to a given age; The virtual try-on system according to configuration 13, wherein the future body size acquisition unit estimates the future body size based on the current body size, the age, the predicted year and month, and a probability distribution of the body size of the person's gender. (Configuration 15) the future body dimension acquisition unit acquires a probability of the current body dimension based on a probability distribution of the body dimension of the age and the sex of the person; The virtual try-on system according to configuration 14, characterized in that it outputs the future body dimensions as the elapsed age obtained by adding the predicted years and months to the age and the body dimensions at the time of the probability in the probability distribution of the body dimensions for the gender. (Configuration 16) a past body dimension input unit for inputting a combination of the person's past body dimension and his / her age at that time, The virtual try-on system according to configuration 14 or 15, wherein the future body measurement acquisition unit estimates the future body measurements using the past body measurements acquired when there is input to the past body measurement input unit and the age at that time. (Configuration 17) Further comprising a premature birth information input unit for acquiring information regarding whether the person is a premature baby and, if the person is a premature baby, an expected date of birth; the future body size acquisition unit calculates a corrected age using the expected delivery date as a birthday when an input is made to the premature birth information input unit, 17. The virtual try-on system according to any one of configurations 14 to 16, wherein the age is replaced with the corrected age. (Configuration 18) The apparatus further includes a future body dimension input unit into which the future body dimensions of the person are input, 18. The virtual try-on system according to any one of configurations 11 to 17, wherein when there is an input to the future body dimension input section, the input body dimension is set as the future body dimension. (Configuration 19) 19. The virtual try-on system according to any one of configurations 11 to 18, further comprising a display unit for displaying the virtual try-on image. (Configuration 20) 20. The virtual try-on system according to any one of configurations 11 to 19, further comprising a clothing database including data relating to a plurality of clothing images and clothing dimensions corresponding to the plurality of clothing images. (Configuration 21) a clothing selection unit that selects one of the clothing images from the clothing database and selects one size of the selected clothing image; 21. The virtual try-on system according to claim 20, wherein the image transformation unit changes the scale of the image using the garment dimensions of the selected size. (Configuration 22) The image transformation unit is if it is determined that the sleeve length of the clothing measurement is longer than the sleeve length of the person, trimming the cuff so that the sleeve length of the clothing measurement is the same as the sleeve length of the person; The virtual fitting system described in configuration 21, characterized in that if it is determined that the inseam of the clothing dimension is longer than the inseam of the person, the hem is trimmed so that the inseam of the clothing dimension is the same as the inseam of the person. (Configuration 23) The display unit is If it is determined that the sleeve length of the clothing dimensions is longer than the sleeve length of the person, a measurement obtained by subtracting the sleeve length of the person from the sleeve length of the clothing dimensions is displayed; The virtual fitting system described in configuration 19, characterized in that if the inseam of the clothing dimension is determined to be longer than the inseam of the person, the dimension obtained by subtracting the inseam of the person from the inseam of the clothing dimension is displayed. (Configuration 24) The virtual try-on system according to configuration 19, wherein the display unit displays measurements obtained by subtracting future body measurements or current body measurements from the clothing measurements. (Method 1) Comparing the person's future body dimensions to the clothing dimensions to rescale at least one of the clothing image and the person image; and generating a virtual try-on image by superimposing the clothing image on the person image. (Configuration 25) A program for causing a computer to execute the image generating method according to method 1. (Configuration 26) A computer-readable storage medium storing the program according to configuration 25.

[0059] Although the preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various modifications and changes are possible within the scope of the gist of the present invention. [Explanation of symbols]

[0060] 200 Image generating device 250 Image transformation section 260 Image Generation Unit

Claims

1. an image transformation unit that changes the scale of at least one of the clothing image and the person image by comparing the person's future body dimensions and clothing dimensions; and an image generating unit that generates a virtual fitting image by superimposing the clothing image on the person image.

2. the person image is obtained by photographing the person with an imaging device, the future body dimensions are estimated using the person's current body dimensions; The image generating apparatus of claim 1 , wherein the clothing dimensions correspond to the clothing image obtained from a clothing database.

3. 2. The image generating apparatus according to claim 1, further comprising a current body dimension acquiring unit for acquiring current body dimensions of the person.

4. The image generating apparatus according to claim 3 , further comprising a distance information acquiring unit for acquiring three-dimensional distance information of the person.

5. a body part estimation unit that estimates a body part of the person in the person image, The image generating device according to claim 4, characterized in that the current body dimension acquisition unit acquires the current body dimensions based on the body part estimated by the body part estimation unit and the distance information acquired by the distance information acquisition unit.

6. 6. The image generating apparatus according to claim 5, wherein the body parts include at least one of the person's head, shoulders, hands, waist, crotch, and soles.

7. The image generating device according to claim 5 , wherein the image generating unit generates the virtual fitting image by superimposing the clothing image on the person image based on the body part.

8. 3. The image generating apparatus according to claim 2, further comprising a future body size acquiring section for estimating the future body size based on the current body size, the person's age, and the person's predicted years and months.

9. 2. The image generating apparatus according to claim 1, wherein the future body dimensions are at least one of the person's height, sleeve length, back length, inseam, chest circumference, waist circumference, and hip circumference.

10. 2. The image generating device according to claim 1, wherein the clothing dimensions are at least one of sleeve length, length, inseam, chest circumference, waist circumference, and hip circumference.

11. An image generating device according to any one of claims 1 to 10; and an imaging device for photographing the person and outputting the person image.

12. 12. The virtual try-on system according to claim 11, further comprising a current body size input section for inputting the current body size of the person.

13. The virtual try-on system according to claim 11, further comprising a gender input unit for acquiring the gender of the person.

14. an age input unit for acquiring the age of the person; A predicted date input unit for acquiring a predicted date of the person; a future body size acquisition unit that estimates the future body size based on the person's current body size, the person's age, and the person's predicted year and month; a statistical database including a probability distribution of body dimensions for each sex from birth to a given age; The virtual try-on system according to claim 13, characterized in that the future body size acquisition unit estimates the future body size based on the current body size, the age, the predicted year and month, and a probability distribution of the body size of the person's gender.

15. the future body dimension acquisition unit acquires a probability of the current body dimension based on a probability distribution of the body dimension of the age and the sex of the person; The virtual try-on system according to claim 14, characterized in that the future body dimensions are output as the elapsed age obtained by adding the age to the predicted years and months and the body dimensions at the probability in the probability distribution of the body dimensions for the gender.

16. a past body dimension input unit for inputting a combination of the person's past body dimension and his / her age at that time, The virtual try-on system according to claim 14, characterized in that the future body measurement acquisition unit estimates the future body measurements using the past body measurements acquired when there is input to the past body measurement input unit and the age at that time.

17. Further comprising a premature birth information input unit for acquiring information regarding whether the person is a premature baby and, if the person is a premature baby, an expected date of birth; the future body size acquisition unit calculates a corrected age using the expected delivery date as a birthday when an input is made to the premature birth information input unit, The virtual try-on system according to claim 14, wherein the age is replaced with the corrected age.

18. The apparatus further includes a future body dimension input unit into which the future body dimensions of the person are input, The virtual try-on system according to claim 11, wherein, when there is an input to the future body size input section, the input body size is set as the future body size.

19. 12. The virtual try-on system according to claim 11, further comprising a display unit for displaying the virtual try-on image.

20. 12. The virtual try-on system of claim 11, further comprising a clothing database including data relating to a plurality of clothing images and clothing dimensions corresponding to the plurality of clothing images.

21. a clothing selection unit that selects one of the clothing images from the clothing database and selects one size of the selected clothing image; 21. The virtual try-on system of claim 20, wherein an image transformation unit changes the scale of the image using the garment dimensions of the selected size.

22. The image transformation unit is if it is determined that the sleeve length of the clothing measurement is longer than the sleeve length of the person, trimming the cuff so that the sleeve length of the clothing measurement is the same as the sleeve length of the person; The virtual try-on system of claim 21, wherein if it is determined that the inseam of the clothing dimension is longer than the inseam of the person, the hem is trimmed so that the inseam of the clothing dimension is the same as the inseam of the person.

23. The display unit is If it is determined that the sleeve length of the clothing dimensions is longer than the sleeve length of the person, a measurement obtained by subtracting the sleeve length of the person from the sleeve length of the clothing dimensions is displayed; The virtual try-on system according to claim 19, characterized in that if the inseam of the clothing dimension is determined to be longer than the inseam of the person, the inseam of the clothing dimension minus the inseam of the person is displayed.

24. 20. The virtual try-on system according to claim 19, wherein the display unit displays measurements obtained by subtracting future body measurements or current body measurements from the clothing measurements.

25. Comparing the person's future body dimensions to the clothing dimensions to rescale at least one of the clothing image and the person image; and generating a virtual try-on image by superimposing the clothing image on the person image.

26. A program causing a computer to execute the image generating method according to claim 25.

27. 27. A computer-readable storage medium storing the program according to claim 26.

Images