Image processing method, device and storage medium

Abstract

The application provides an image processing method, device and storage medium, the method comprising: collecting a first image of a user and obtaining a second image of the user, wherein the first image and the second image are images of the user at different age stages; fusing the first image and the second image to obtain a fused image; determining a clothing image corresponding to the fused image; determining a dressing image according to the fused image and the clothing image, and displaying the dressing image. The image processing method, device and storage medium provided by the application are used to improve the diversity of functions of an electronic device for the elderly.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to an image processing method, device and storage medium. Background Technology

[0002] Currently, with the deepening of population aging in my country, in order to meet the living needs of middle-aged and elderly people and make their lives more convenient, many electronic devices for the elderly have appeared on the market, such as washing machines and mobile phones specially designed for middle-aged and elderly people.

[0003] In related technologies, for example, washing machines designed specifically for middle-aged and elderly people can only assist them in washing clothes; similarly, mobile phones designed specifically for middle-aged and elderly people can only meet their calling needs and some entertainment needs, making the functions of electronic devices for the elderly relatively limited. Summary of the Invention

[0004] This application provides an image processing method, device, and storage medium to enhance the versatility of functions in electronic devices for the elderly.

[0005] In a first aspect, embodiments of this application provide an image processing method, including:

[0006] A first image of the user is captured, and a second image of the user is obtained, wherein the first image and the second image are images of the user at different age stages;

[0007] The first image and the second image are fused to obtain a fused image;

[0008] Identify the clothing image corresponding to the fused image;

[0009] Based on the merged image and the clothing image, determine the outfit image and display the outfit image.

[0010] In one possible design, determining the clothing image corresponding to the fused image includes: extracting the user's feature information from the fused image; determining the clothing type based on the feature information; and determining the clothing image corresponding to the fused image based on the clothing type.

[0011] In one possible design, determining the clothing type based on feature information includes: receiving scene information input by the user; and determining the clothing type based on the feature information and the scene information.

[0012] In one possible design, clothing type is determined based on feature information and scene information, including: inputting feature information and scene information into a pre-trained clustering network to obtain clothing type.

[0013] In one possible design, determining the clothing image corresponding to the fused image based on the clothing type includes: displaying multiple clothing images corresponding to the clothing type; and, in response to a touch operation input by the user, determining the clothing image corresponding to the fused image from the multiple clothing images.

[0014] In one possible design, fusing the first image and the second image to obtain a fused image includes: inputting the first image and the second image into a pre-trained first VGG network to obtain a first feature vector corresponding to the first image and a second feature vector corresponding to the second image; concatenating the first feature vector and the second feature vector to obtain a first fused feature vector, which is used to represent the fused features of the first image and the second image; and inputting the first fused feature vector into a first generative adversarial network to obtain the fused image.

[0015] In one possible design, determining the outfit image based on the fused image and the clothing image includes: inputting the fused image and the clothing image into a pre-trained second VGG network to obtain a third feature vector corresponding to the fused image and a fourth feature vector corresponding to the clothing image; concatenating the third feature vector and the fourth feature vector to obtain a second fused feature vector, which is used to represent the fused features of the fused image and the clothing image; and inputting the second fused feature vector into a second generative adversarial network to obtain the outfit image.

[0016] Secondly, embodiments of this application provide an image processing apparatus, including: an acquisition module, an acquisition module, a processing module, and a display module; wherein,

[0017] The acquisition module is used to acquire the user's first image;

[0018] The acquisition module is used to acquire a second image of the user, wherein the first image and the second image are images of the user at different age stages;

[0019] The processing module is used to fuse the first image and the second image to obtain a fused image;

[0020] The processing module is also used to determine the clothing image corresponding to the fused image;

[0021] The processing module is also used to determine the outfit image based on the fused image and the clothing image;

[0022] The display module is also used to display outfit images.

[0023] In one possible design, the processing module is further used to: extract user feature information from the fused image; determine the clothing type based on the feature information; and determine the clothing image corresponding to the fused image based on the clothing type.

[0024] In one possible design, the processing module is further used to: receive scene information input by the user; and determine the clothing type based on the feature information and the scene information.

[0025] In one possible design, the processing module is specifically used to: input feature information and scene information into a pre-trained clustering network to obtain clothing type.

[0026] In one possible design, the processing module is further configured to: display multiple clothing images corresponding to the clothing type; and, in response to a user's touch input, determine the clothing image corresponding to the fused image from the multiple clothing images.

[0027] In one possible design, the processing module is further configured to: input the first image and the second image into a pre-trained first VGG network to obtain a first feature vector corresponding to the first image and a second feature vector corresponding to the second image; concatenate the first feature vector and the second feature vector to obtain a first fused feature vector, which is used to represent the fused features of the first image and the second image; and input the first fused feature vector into a first generative adversarial network to obtain a fused image.

[0028] In one possible design, the processing module is further configured to: input the fused image and the clothing image into a pre-trained second VGG network to obtain a third feature vector corresponding to the fused image and a fourth feature vector corresponding to the clothing image; concatenate the third feature vector and the fourth feature vector to obtain a second fused feature vector, which is used to represent the fused features of the fused image and the clothing image; and input the second fused feature vector into a second generative adversarial network to obtain an outfit image.

[0029] Thirdly, embodiments of this application provide an electronic device, including: a processor and a memory;

[0030] The memory stores instructions that the computer executes;

[0031] The processor executes computer execution instructions stored in memory, causing the processor to perform the method as described in any of the first aspects.

[0032] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the method as described in any of the first aspects.

[0033] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the method as described in any of the first aspects.

[0034] This application provides an image processing method, apparatus, and storage medium. The method includes: acquiring a first image of a user and obtaining a second image of the user, wherein the first image and the second image are images of the user at different age stages; fusing the first image and the second image to obtain a fused image; determining a clothing image corresponding to the fused image; and determining and displaying an outfit image based on the fused image and the clothing image. In the method provided by this application, fusing the first image and the second image to obtain a fused image enables the elderly care electronic device to have the function of transforming older images into younger images. Furthermore, determining the clothing image corresponding to the fused image and determining and displaying the outfit image based on the fused image and the clothing image enables the elderly care electronic device to have a clothing try-on function, thereby improving the versatility of the elderly care electronic device's functions. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram illustrating an application scenario of the image processing method provided in the embodiments of this application;

[0037] Figure 2 Flowchart of the image processing method provided in the embodiments of this application Figure 1 ;

[0038] Figure 3 Flowchart of the image processing method provided in the embodiments of this application Figure 2 ;

[0039] Figure 4 A schematic diagram of the structure of a training VGG 19 network provided in an embodiment of this application;

[0040] Figure 5 A schematic diagram of the structure of the jointly trained VGG 19 network and CGAN provided for embodiments of this application;

[0041] Figure 6 This is a schematic diagram of the structure for obtaining outfit images provided in this application;

[0042] Figure 7 A system structure diagram of the virtual try-on system provided in the embodiments of this application;

[0043] Figure 8 This is a schematic diagram of the structure of the image processing apparatus provided in the embodiments of this application;

[0044] Figure 9 A hardware schematic diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0046] In existing technologies, the functions of electronic devices for the elderly are relatively limited. For example, washing machines designed specifically for middle-aged and elderly people can only help them wash clothes, and mobile phones designed specifically for middle-aged and elderly people can only meet their calling needs and some entertainment needs.

[0047] To enhance the functionality of electronic devices for elderly care, the inventors conceived of adding a clothing try-on function. In this application, to achieve this function, the inventors proposed generating a fused image from two images of different age groups, matching the fused image with clothing images, and then generating an outfit image based on the fused image and the clothing images. This allows the electronic device to have a clothing try-on function, thereby increasing its functionality.

[0048] The following is combined Figure 1 The application scenarios of the image processing method provided in this application are described.

[0049] Figure 1 This is a schematic diagram illustrating an application scenario of the image processing method provided in the embodiments of this application. For example... Figure 1 As shown, it includes: a first image, a second image, a merged image, a clothing image, and an outfit image.

[0050] Optionally, the first image can be an image of the user when they were young, and the second image can be an image of the user currently.

[0051] The fused image is the image obtained by fusing the first image and the second image.

[0052] The outfit image is the result of fusing the blended image and the clothing image. In the outfit image, the person in the outfit image is the same person in the blended image, and the person in the outfit image is wearing the clothing from the clothing image.

[0053] In this application, the person in the clothing image is a person in the fused image, and the person in the clothing image is wearing the clothing in the clothing image, so that the elderly care electronic device has a clothing try-on function, thereby improving the diversity of functions of the elderly care electronic device.

[0054] The technical solutions of this application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0055] Figure 2 Flowchart of the image processing method provided in the embodiments of this application Figure 1 .like Figure 2 As shown, the method includes:

[0056] S201. Acquire a first image of the user and obtain a second image of the user, wherein the first image and the second image are images of the user at different age stages.

[0057] Optionally, the execution subject of this application embodiment is an elderly care electronic device, or it can be an image processing device installed in the elderly care electronic device. The image processing device can be implemented through a combination of software and / or hardware. The software includes, but is not limited to, a virtual fitting system installed in the elderly care electronic device.

[0058] Alternatively, the elderly care electronic device can be a mobile phone (or "cellular" phone) and a computer with a mobile terminal, or a care device, etc.

[0059] Optionally, the elderly care electronic device is equipped with a camera that can capture a first image of the user. The first image is the user's current image. The second image is a pre-stored image. Both the first and second images include an image of the user's human body (including the head, face, limbs, torso, etc.).

[0060] S202. The first image and the second image are fused to obtain a fused image.

[0061] In one possible design, a pre-trained deep learning model is used to fuse the first image and the second image to obtain a fused image. The deep learning model is trained using a set of first sample images corresponding to multiple users. Each set of first sample images includes a first sample image and a second sample image. The first and second sample images represent images of users at different age stages.

[0062] Optionally, the user's age group in the fused image (e.g., adult) is younger than the user's age group in the first image (e.g., elderly) and older than the user's age group in the second image (e.g., teenager). The method of training a deep learning model using a set of first sample images is described below.

[0063] Step 1: Input the first sample image of the i-th group into the learning model after the (i-1)th update to obtain the i-th target image corresponding to the first sample image of the i-th group.

[0064] Step 2: Determine the similarity between the label image corresponding to the i-th target image and the first sample image of the i-th group.

[0065] The label image corresponding to the first sample image in group i is the image of the user in the first age stage.

[0066] Optionally, the first age stage can be images of the user at different age stages, including the first and second sample images in the i-th group of first sample images.

[0067] Step 3: Determine whether the similarity is greater than or equal to the threshold. If yes, then determine the learning model after the (i-1)th update as a deep learning model. If no, then update the learning model after the (i-1)th update based on the similarity to obtain the learning model after the i-th update, and repeat steps 1, 2, and 3 until the similarity is greater than or equal to the threshold.

[0068] S203. Determine the clothing image corresponding to the fused image.

[0069] In one possible design, determining the clothing image corresponding to the fused image includes: extracting the user's feature information from the fused image; determining the clothing type based on the feature information; and determining the clothing image corresponding to the fused image based on the clothing type.

[0070] The user's characteristics include age, body type, height, weight, etc.

[0071] Alternatively, the fused image can be processed using a multi-head YOLOv3 network to extract the user's feature information.

[0072] Alternatively, the feature information can be processed using the k-means clustering algorithm to determine the type of clothing.

[0073] Optionally, an image database is pre-configured, which includes at least one image of clothing corresponding to each of multiple clothing types. This image database can be located on a server or in an electronic device for elderly care.

[0074] Alternatively, the clothing image corresponding to the fused image can be determined using the following two methods.

[0075] Method 1: When the image database is set in the server, send the clothing type to the server; receive at least one clothing image belonging to the clothing type from the server; and determine the at least one clothing image belonging to the clothing type as the clothing image corresponding to the fused image.

[0076] Method 2: When the image database is set in the elderly care electronic device, at least one clothing image of the clothing type in the image database is determined as the clothing image corresponding to the fused image.

[0077] S204. Based on the fused image and the clothing image, determine the outfit image and display the outfit image.

[0078] exist Figure 2 In this embodiment, the first image and the second image are fused to obtain a fused image. The user's age stage (e.g., middle-aged) in the fused image is younger than the user's age stage (e.g., elderly) in the first image, but older than the user's age stage (e.g., teenager) in the second image. This enables the elderly care electronic device to transform older images into younger images. Furthermore, a clothing image corresponding to the fused image is determined, and an outfit image is determined based on the fused image and the clothing image. The outfit image is then displayed, which enables the elderly care electronic device to have a clothing try-on function, thereby improving the diversity of the elderly care electronic device's functions.

[0079] Currently, with the deepening of population aging, the spiritual life of middle-aged and elderly people is not rich enough. In order to enrich the spiritual life of middle-aged and elderly people, the elderly care electronic device provided in this application can enrich the spiritual life of middle-aged and elderly people due to its diverse functions.

[0080] In the above Figure 2 Based on this, the following will combine Figure 3 For a further detailed description of the image processing method provided in this application, please refer to [link / reference needed]. Figure 3 Example.

[0081] Figure 3 Flowchart of the image processing method provided in the embodiments of this application Figure 2 .like Figure 3 As shown, the method includes:

[0082] S301. Acquire a first image of the user and obtain a second image of the user, wherein the first image and the second image are images of the user at different age stages.

[0083] Specifically, the execution method of S301 is the same as that of S201, and the execution process of S301 will not be described again here.

[0084] S302. Input the first image and the second image into the pre-trained first VGG network respectively to obtain the first feature vector corresponding to the first image and the second feature vector corresponding to the second image.

[0085] Optionally, the first VGG network can be obtained by training a VGG 19 network using the aforementioned multiple sets of first sample images. It should be noted that for an explanation of the VGG 19 network training process, please refer to [link to documentation / reference]. Figure 4 and Figure 5 Example.

[0086] Optionally, both the first feature vector and the second feature vector are 1-dimensional vectors with 2048 dimensions.

[0087] S303. The first feature vector and the second feature vector are concatenated to obtain the first fused feature vector, which is used to represent the fused features of the first image and the second image.

[0088] When both the first feature vector and the second feature vector are 1-dimensional vectors with dimensions of 2048, the first fused feature vector is a 1-dimensional vector with dimensions of 4096.

[0089] S304. Input the first fusion feature vector into the first generative adversarial network to obtain the fusion image.

[0090] Optionally, when the first VGG network is obtained by training a VGG 19 network, the first generative adversarial network can be a pre-obtained network after training a Conditional Generative Adversarial Network (CGAN).

[0091] Optionally, the first generative adversarial network can also be a network corresponding to CGAN obtained after jointly training CGAN and VGG 19 networks. For details on jointly training CGAN and VGG 19 networks, please refer to [link to documentation]. Figure 5 Example.

[0092] S305. Extract user feature information from the fused image.

[0093] S306. Receive scene information input by the user.

[0094] Scene information may include at least one of the following: singing, square dancing, performance, old classmates reunion, leisure, travel, etc.

[0095] In one possible design, the elderly care electronic device can display multiple scene information corresponding to each control. The user can click on any control corresponding to any scene information, enabling the elderly care electronic device to receive scene information.

[0096] S307. Determine the type of clothing based on feature information and scene information.

[0097] In one possible design, feature information and scene information are input into a pre-trained clustering network to obtain clothing types. The clustering network can be a k-means clustering algorithm.

[0098] S308. Display multiple clothing images corresponding to the clothing type.

[0099] Optionally, multiple clothing images corresponding to the clothing type are determined from the image database, and the multiple clothing images corresponding to the clothing type are displayed. The method for determining multiple clothing images corresponding to the clothing type from the image database can be found in methods 1 and 2 in S203.

[0100] S309. In response to a user's touch input, determine the clothing image corresponding to the fused image from multiple clothing images.

[0101] Optionally, a user can click on at least one of multiple clothing images to input a touch operation into the elderly care electronic device. Further, in response to the user's touch operation, the clothing image clicked by the user is determined to be the clothing image corresponding to the fused image.

[0102] S310. Input the fused image and the clothing image into the pre-trained second VGG network respectively to obtain the third feature vector corresponding to the fused image and the fourth feature vector corresponding to the clothing image.

[0103] When the number of clothing images is 1, execute S310 to S312 in sequence.

[0104] When the number of clothing images is greater than 1, for example, 2 clothing images, then S310 to S312 are executed sequentially for each clothing image to obtain the outfit image corresponding to each clothing image.

[0105] The second VGG network can also be obtained by training the VGG 19 network using multiple sets of second samples. Each set of second samples includes a fused image and a clothing image.

[0106] It should be noted that the process of obtaining the second VGG network is similar to that of obtaining the first VGG network, and will not be described in detail here.

[0107] S311. The third feature vector and the fourth feature vector are concatenated to obtain the second fusion feature vector, which is used to represent the fusion features of the fused image and the clothing image.

[0108] S312. Input the second fused feature vector into the second generative adversarial network to obtain the outfit image.

[0109] It should be noted that the second generative adversarial network is similar to the first generative adversarial network, and the method for obtaining the second generative adversarial network is similar to the method for obtaining the first generative adversarial network, which will not be elaborated here.

[0110] This application also provides a structural schematic diagram of obtaining an outfit image according to the methods in S305 to S312. For details, please refer to [link / reference needed]. Figure 6 Example.

[0111] exist Figure 3 In the image processing method provided in the embodiment, a first image and a second image are respectively input into a pre-trained first VGG network to obtain a first feature vector corresponding to the first image and a second feature vector corresponding to the second image. The first feature vector and the second feature vector are concatenated to obtain a first fused feature vector. The first fused feature vector is then input into a first generative adversarial network to obtain a fused image, enabling the elderly care electronic device to transform older images into younger images. Further, the fused image and a clothing image are respectively input into a pre-trained second VGG network to obtain a third feature vector corresponding to the fused image and a fourth feature vector corresponding to the clothing image. The third feature vector and the fourth feature vector are concatenated to obtain a second fused feature vector. The second fused feature vector is then input into a second generative adversarial network to obtain a clothing image, enabling the elderly care electronic device to have a clothing try-on function and improving the versatility of the elderly care electronic device's functions.

[0112] Figure 4 This is a schematic diagram of the structure of a trained VGG 19 network provided in an embodiment of this application. Figure 4 As shown, the first VGG network is obtained by performing multiple iterations on the VGG 19 network as follows.

[0113] For each iteration, the first sample image and the second sample image included in a set of first samples are input into the VGG 19 network respectively. This yields the first sample feature vector corresponding to the first sample image and the second sample feature vector corresponding to the second sample image. The similarity between the first and second sample feature vectors is determined, and the difference between a preset value and the similarity is calculated. The model parameters of the VGG 19 network are updated based on this difference. The similarity can be cosine similarity. The preset value is 1, indicating that the first and second sample images are approximately identical.

[0114] When a round of iterations meets the termination condition, the final VGG 19 network is determined as the first VGG network. The termination condition can be at least one of the following: the number of iterations reaches a preset number, the difference is less than or equal to a preset difference, or the model parameters of the VGG19 network converge.

[0115] Figure 5 A schematic diagram illustrating the structure of the jointly trained VGG 19 network and CGAN provided in this application embodiment. Figure 5 As shown, the first VGG network and the first generative adversarial network are obtained by performing multiple iterations on the VGG 19 network and CGAN as follows.

[0116] For each iteration, the first sample image and the second sample image included in a set of first samples are input into the VGG19 network to obtain the first sample feature vector corresponding to the first sample image and the second sample feature vector corresponding to the second sample image. The first sample feature vector and the second sample feature vector are concatenated to obtain the sample fusion feature vector. The sample fusion feature vector is input into CGAN to obtain the sample fusion image. The similarity between the sample fusion image and the label image corresponding to a set of first samples is determined, and the VGG19 network and CGAN are updated according to the similarity.

[0117] When an iteration meets the termination condition, the final VGG 19 network is determined as the first VGG network, and CGAN is determined as the first generative adversarial network. The termination condition can be at least one of the following: the number of iterations reaches a preset number, the similarity is greater than or equal to a preset similarity, or the model parameters of both the VGG 19 network and CGAN converge.

[0118] Figure 6 This is a structural diagram illustrating the structure of the clothing images provided in this application. (Example) Figure 6 As shown, the fused image is processed by a multi-head YOLOv3 network to extract the user's feature information. The scene information and feature information input by the user are processed by the k-means clustering algorithm to obtain the clothing type, and then the clothing image corresponding to the fused image is determined. The fused image and the clothing image are input into the second VGG network to obtain the third feature vector corresponding to the fused image and the fourth feature vector corresponding to the clothing image. The third feature vector and the fourth feature vector are concatenated to obtain the second fused feature vector. The second fused feature vector is input into the second generative adversarial network to obtain the outfit image.

[0119] Based on the above embodiments, this application also provides a virtual try-on system. For details, please refer to [link to relevant documentation]. Figure 7 Example.

[0120] Figure 7 This is a system structure diagram of the virtual try-on system provided in an embodiment of this application. Figure 7 As shown, for example, a virtual fitting system can have three functions: time reversal function, outfit recommendation function, and reverse time outfit recommendation function.

[0121] The time reversal function is used to capture a first image of the user and obtain a second image of the user, wherein the first image and the second image are images of the user at different age stages; the first image and the second image are then merged to obtain a merged image.

[0122] Optionally, the user's age group (e.g., adult) in the merged image is younger than the user's age group (e.g., elderly) in the first image, but older than the user's age group (e.g., teenager) in the second image.

[0123] Optionally, multiple first images of the user can be captured, and each first image and second image can be merged to obtain multiple merged images. The multiple merged images can be played in sequence to obtain a dynamic display process.

[0124] It should be noted that when using a first VGG network trained on the VGG 19 network to fuse each first and second image separately to obtain multiple fused images, the time required to obtain the fused images is relatively long and the speed is slow, resulting in an unsmooth dynamic display process. In this application, to make the dynamic display process smoother, the convolutional layers in the VGG 19 network are replaced with depthwise separable convolutions. Then, a first VGG network trained on the VGG 19 network including depthwise separable convolutions is used to fuse each first and second image separately to obtain multiple fused images, thereby reducing the time required to fuse images, increasing the speed of obtaining fused images, and making the dynamic display process smoother.

[0125] In this application, a first image of the user in an older age group is fused with a second image of the user in an younger age group to obtain a fused image. This allows, for example, an elderly person to see an image of their prime age, thereby increasing the versatility of the functions of elderly care electronic devices equipped with virtual fitting systems.

[0126] The outfit recommendation function is used to identify the clothing image corresponding to the first image; based on the first image and the clothing image, it determines and displays an outfit image. The person in the outfit image is the same person in the first image, and the person in the outfit image is wearing the clothing from the clothing image. Optionally, the first image can be an image of the user at any age (including middle-aged, elderly, etc.).

[0127] It should be noted that the method for determining the clothing image corresponding to the first image is similar to the method for determining the clothing image corresponding to the fused image, and will not be described in detail here.

[0128] In this application, the outfit recommendation function can recommend clothing to users of any age, thereby increasing the diversity of functions of elderly care electronic devices equipped with virtual fitting systems.

[0129] The reverse-time outfit recommendation function implements the image processing method provided in this application.

[0130] It should be noted that when the virtual fitting system is set up on the elderly care electronic device, the corresponding controls for the time reversal function, outfit recommendation function, and reverse time outfit recommendation function can be displayed on the elderly care electronic device. When the user clicks on the control, the corresponding function is realized.

[0131] exist Figure 7 In the embodiments, the elderly care electronic device may have time reversal function, outfit recommendation function, and reverse time outfit recommendation function, which improves the diversity of functions of the elderly care electronic device.

[0132] Figure 8 This is a schematic diagram of the structure of an image processing apparatus provided in an embodiment of this application. Figure 8 As shown, the image processing device 10 includes: an acquisition module 11, an acquisition module 12, a processing module 13, and a display module 14; wherein,

[0133] Acquisition module 11 is used to acquire the user's first image;

[0134] The acquisition module 12 is used to acquire a second image of the user, wherein the first image and the second image are images of the user at different age stages;

[0135] Processing module 13 is used to fuse the first image and the second image to obtain a fused image;

[0136] Processing module 13 is also used to determine the clothing image corresponding to the fused image;

[0137] Processing module 13 is also used to determine the outfit image based on the fused image and the clothing image;

[0138] Display module 14 is used to display outfit images.

[0139] The image processing apparatus 10 provided in this application embodiment can execute the above-described image processing method. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0140] In one possible design, the processing module 13 is further configured to: extract user feature information from the fused image; determine clothing type based on the feature information; and determine clothing image corresponding to the fused image based on the clothing type.

[0141] In one possible design, the processing module 13 is further configured to: receive scene information input by the user; and determine the clothing type based on the feature information and the scene information.

[0142] In one possible design, the processing module 13 is further used to: input feature information and scene information into a pre-trained clustering network to obtain clothing type.

[0143] In one possible design, the processing module 13 is further configured to: display multiple clothing images corresponding to the clothing type; and, in response to a user's touch input, determine the clothing image corresponding to the fused image from the multiple clothing images.

[0144] In one possible design, the processing module 13 is further configured to: input the first image and the second image into a pre-trained first VGG network to obtain a first feature vector corresponding to the first image and a second feature vector corresponding to the second image; concatenate the first feature vector and the second feature vector to obtain a first fused feature vector, which is used to represent the fused features of the first image and the second image; and input the first fused feature vector into a first generative adversarial network to obtain a fused image.

[0145] In one possible design, the processing module 13 is further configured to: input the fused image and the clothing image into a pre-trained second VGG network to obtain a third feature vector corresponding to the fused image and a fourth feature vector corresponding to the clothing image; concatenate the third feature vector and the fourth feature vector to obtain a second fused feature vector, which is used to represent the fused features of the fused image and the clothing image; and input the second fused feature vector into a second generative adversarial network to obtain an outfit image.

[0146] The image processing apparatus 10 provided in this application embodiment can execute the above-described image processing method. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0147] Figure 9 A hardware schematic diagram of the electronic device provided in an embodiment of this application. For example... Figure 9 As shown, the electronic device 20 may include: a transceiver 21, a memory 22, and a processor 23.

[0148] Transceiver 21 may include a transmitter and / or a receiver. The transmitter may also be referred to as a transmitter, transmitter, transmitting port, or transmitting interface, etc. The receiver may also be referred to as a receiver, receiver, receiving port, or receiving interface, etc. Exemplarily, transceiver 21, memory 22, and processor 23 are interconnected via bus 24.

[0149] Memory 22 is used to store computer-executed instructions.

[0150] The processor 23 is used to execute computer execution instructions stored in the memory 22, so that the processor 23 performs the above-described image processing method.

[0151] This application provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the image processing method of any of the above method embodiments.

[0152] This application provides a computer program product, including a computer program that, when executed by a processor, implements the image processing method of any of the above method embodiments.

[0153] All or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable memory. When the program is executed, it performs the steps of the above method embodiments; and the aforementioned memory (storage medium) includes: read-only memory (ROM), RAM, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof.

[0154] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processing unit of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0155] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0156] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0157] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

[0158] In this application, the term "comprising" and its variations can refer to non-limiting inclusion; the term "or" and its variations can refer to "and / or". The terms "first", "second", etc., in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. In this application, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. For example, "A, B, or C" or "A, B, and / or C" means "any one of the following: A; B; C; A and B; A and C; B and C; A, B, and C".

Claims

1. An image processing method applied to an old electronic device, characterized in that, The method comprises: collecting a first image of a user and obtaining a second image of the user, wherein the first image and the second image are images of the user at different age stages; fusing the first image and the second image to obtain a fused image, wherein the age stage of the user in the fused image is smaller than the age stage of the user in the first image and larger than the age stage of the user in the second image; determining a clothing image corresponding to the fused image, determining a dressing image according to the fused image and the clothing image, and displaying the dressing image; the fusing the first image and the second image to obtain a fused image comprises: inputting the first image and the second image into a pre-trained first VGG network respectively to obtain a first feature vector corresponding to the first image and a second feature vector corresponding to the second image; splicing the first feature vector and the second feature vector to obtain a first fused feature vector, the first fused feature vector being used to represent the fused features of the first image and the second image; and inputting the first fused feature vector into a first generative adversarial network to obtain the fused image.

2. The method of claim 1, wherein, The determining a clothing image corresponding to the fused image comprises: extracting feature information of the user from the fused image; determining a clothing type according to the feature information; determining a clothing image corresponding to the fused image according to the clothing type.

3. The method of claim 2, wherein, The determining a clothing type according to the feature information comprises: receiving scene information input by the user; determining the clothing type according to the feature information and the scene information.

4. The method of claim 3, wherein, The determining the clothing type according to the feature information and the scene information comprises: inputting the feature information and the scene information into a pre-trained clustering network to obtain the clothing type.

5. The method according to any one of claims 2 to 4, characterized in that, The determining a clothing image corresponding to the fused image according to the clothing type comprises: displaying a plurality of clothing images corresponding to the clothing type; determining a clothing image corresponding to the fused image from the plurality of clothing images in response to a touch operation input by the user.

6. The method according to any one of claims 1 to 4, characterized in that, The determining a dressing image according to the fused image and the clothing image comprises: inputting the fused image and the clothing image into a pre-trained second VGG network respectively to obtain a third feature vector corresponding to the fused image and a fourth feature vector corresponding to the clothing image; splicing the third feature vector and the fourth feature vector to obtain a second fused feature vector, the second fused feature vector being used to represent the fused features of the fused image and the clothing image; inputting the second fused feature vector into a second generative adversarial network to obtain the dressing image.

7. An electronic device, comprising: The method comprises: a processor and a memory; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory, so that the processor executes the method according to any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer executing instructions, and when the processor executes the computer executing instructions, the method in any one of claims 1 to 6 is implemented.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method in any one of claims 1 to 6.

Images