Clothing detail feature intelligent fusion method and system based on fashion trend

Abstract

The invention discloses a clothing detail feature intelligent fusion method and system based on a fashion trend, belongs to the technical field of intelligent clothing design, and can reduce the dependence on the personal experience and subjective judgment of a designer. Comprising the steps of obtaining a plurality of garment images conforming to a fashion trend; identifying clothes type information and style type information corresponding to each clothes image, classifying and marking a corresponding first classification label; obtaining target clothes type information; obtaining a candidate clothing image set from the classified clothing images based on the first classification label and the target clothing type information; extracting a plurality of key features of each clothing image in the candidate clothing image set to obtain a plurality of to-be-fused key features marked with second classification labels; based on the second classification label, combining the plurality of to-be-fused key features to obtain a plurality of costume design schemes; and analyzing the aesthetic score of each costume design scheme, and outputting the costume design scheme of which the aesthetic score is greater than a preset threshold value.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent clothing design, and in particular to a method and system for intelligently fusing clothing detail features based on fashion trends. Background Art

[0002] In today's rapidly developing clothing industry, keeping up with fashion trends and innovatively designing unique clothing styles have become the key to brand competition. However, the traditional clothing design process often relies on the designer's personal experience and subjective judgment, which is particularly insufficient in today's ever-changing fashion environment. Designers are usually limited by their own aesthetics and experience when creating, and it is difficult to fully and accurately capture the popular elements and clothing details in the current market.

[0003] The disclosure of the above background technology content is only used to assist in understanding the concept and technical solution of the present invention. It does not necessarily belong to the prior art of this patent application. In the absence of clear evidence that the above content has been disclosed on the filing date of this patent application, the above background technology should not be used to evaluate the novelty and creativity of the present application. Summary of the invention

[0004] The present application provides a method and system for intelligently integrating clothing detail features based on fashion trends, which can reduce the reliance on the designer's personal experience and subjective judgment.

[0005] To achieve the above objectives, the present application discloses the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a method for intelligently fusing clothing detail features based on fashion trends, comprising the following steps:

[0007] Obtain multiple clothing images that match fashion trends;

[0008] Using image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classifying multiple clothing images according to a preset classification rule, and marking a first classification label of the clothing type and style type information to which each clothing image belongs;

[0009] Get target clothing type information;

[0010] extracting a plurality of clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set;

[0011] Using image recognition technology to extract multiple key features of each clothing image in the candidate clothing image set, multiple key features to be fused are obtained, and a second classification label of the clothing type to which each key feature belongs is marked;

[0012] Based on the second classification label, multiple key features to be fused are combined to obtain multiple clothing design solutions;

[0013] Analyze the aesthetic score of each clothing design scheme and output clothing design schemes whose aesthetic score is greater than a preset threshold.

[0014] In the embodiment of the present application, first, the method establishes a rich data foundation by acquiring multiple clothing images that conform to fashion trends. These images provide an objective basis for analysis and reduce the subjective bias of designers in the source of inspiration. Next, the clothing images are classified using image recognition technology to mark the clothing type and style type information. This process ensures the standardization and accuracy of the data, thereby overcoming the errors that may be caused by manual classification.

[0015] After obtaining the target clothing type information, the system extracts clothing images that match the target type from the classified images, ensuring the pertinence of the design scheme. By extracting multiple key features of the candidate clothing images, a feature set to be fused is formed. These features are based on objective analysis and can fully reflect the popular design details, avoiding omissions that may be caused by designers' personal aesthetics and experience limitations. Subsequently, the system combines multiple key features to be fused to generate multiple clothing design schemes. This step relies on a data-driven approach to ensure that the generated design scheme can integrate different popular elements and increase the diversity and innovation of the design.

[0016] Finally, by analyzing the aesthetic score of each clothing design solution, the system outputs solutions with scores higher than the preset threshold. This step not only improves design efficiency, but also ensures that the output design solution is aesthetically competitive in the market. In summary, this technical solution reduces the reliance on designers' personal experience and subjective judgment through intelligent and systematic methods, and improves the innovation and adaptability of the design.

[0017] In some possible implementations of the first aspect, the steps of using image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classifying the multiple clothing images according to a preset classification rule include:

[0018] Acquire pixel data of the clothing image, and pre-process the pixel data to enhance the image quality;

[0019] The preprocessed clothing image is input into a pre-trained deep learning model based on a convolutional neural network. The input clothing image is feature extracted through the convolution layer, and local features in the image are captured using multiple convolution kernels of different sizes. The feature map output by the convolution layer is then downsampled through the pooling layer to reduce the dimension of the feature map. The pooled features are then input into the fully connected layer for integration and classification, and finally the recognition results of clothing type and style type are output.

[0020] Based on the recognition results, multiple clothing images are classified according to the preset classification rules. In this way, the accuracy and efficiency of image recognition can be improved, the type and style of clothing can be determined more accurately, and a reliable basis can be provided for subsequent design processing.

[0021] In some possible implementations of the first aspect, the target clothing type information is a top and / or a coat; and the step of extracting multiple key features of each clothing image in the candidate clothing image set by using image recognition technology to obtain multiple key features to be fused includes:

[0022] Each clothing image in the candidate clothing image set is converted into a grayscale image, and the clothing main body area image is separated using a grayscale threshold segmentation algorithm;

[0023] Execute the Canny edge detection algorithm on the main body area image to extract a continuous edge contour curve, determine the contour shape of the clothing according to the geometric features of the contour curve, and mark the contour shape data as the first feature data;

[0024] Based on the clothing main body area image, the neckline area is determined by using the region growing algorithm to obtain the neckline area image;

[0025] In the neckline area image, the corner point features of the neckline edge are extracted by the Harris corner point detection algorithm, and the neckline shape category and decorative detail features are analyzed according to the distribution, quantity and relative position relationship of the corner points, and marked as the second feature data, and the extraction range of the second feature data is limited to the neckline area image within the overall outline framework of the clothing determined by the first feature data;

[0026] Based on the clothing main body area image and the first feature data, a shape template matching algorithm is used to match a plurality of pre-set cuff shape templates with each part of the main body area image to obtain a cuff area image;

[0027] In the cuff area image, the SIFT algorithm is used to extract the cuff texture features and shape features, obtain the cuff design data and mark it as the third feature data; the first feature data, the second feature data and the third feature data constitute multiple key features to be fused. In this way, the detailed features of the clothing design can be captured more accurately, providing more abundant construction elements for subsequent design solutions, and enhancing the diversity and aesthetics of the design.

[0028] In some possible implementations of the first aspect, the method for intelligently fusing clothing detail features based on fashion trends further includes the following steps:

[0029] Get the target style type;

[0030] Each clothing design scheme with an aesthetic score greater than a preset threshold is compared and analyzed with the target style type, the similarity value is calculated, and the clothing design scheme with a similarity value greater than the preset threshold is output. In this way, it can ensure that the output design scheme is more in line with the specific target style type and improve the matching degree between the design scheme and the expected style.

[0031] In some possible implementations of the first aspect, a plurality of key features to be fused are combined using a permutation and combination algorithm to obtain a plurality of clothing design solutions. In this way, a large number of design combination possibilities can be quickly generated, the diversity and innovation of the design can be increased, the number of designs that a designer can conceive within a limited time can be broken, and more creative options can be provided.

[0032] In some possible implementations of the first aspect, multiple clothing images that conform to fashion trends are collected from multiple preset fashion information platforms through web crawler technology. In this way, the problem that designers have limited channels to obtain fashion trend information and are not timely and comprehensive is solved, which is conducive to making designs keep up with fashion trends.

[0033] In a second aspect, the embodiment of the present application provides a clothing detail feature intelligent fusion system based on fashion trends, including:

[0034] A first acquisition module is used to acquire a plurality of clothing images that conform to fashion trends;

[0035] A first recognition module is used to use image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classify multiple clothing images according to a preset classification rule, and mark each clothing image with a first classification label of the clothing type and style type information to which it belongs;

[0036] The second acquisition module is used to acquire target clothing type information;

[0037] A first extraction module is used to extract multiple clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set;

[0038] The second extraction module is used to extract multiple key features of each clothing image in the candidate clothing image set by using image recognition technology, obtain multiple key features to be fused, and mark each key feature with a second classification label of the clothing type to which it belongs;

[0039] A first fusion module is used to combine multiple key features to be fused based on the second classification label to obtain multiple clothing design solutions;

[0040] The first scoring module is used to analyze the aesthetic score of each clothing design scheme and output the clothing design schemes whose aesthetic scores are greater than a preset threshold.

[0041] In some possible implementations of the second aspect, the first recognition module is specifically used to: obtain pixel data of a clothing image, and preprocess the pixel data to enhance image quality; input the preprocessed clothing image into a pre-trained deep learning model based on a convolutional neural network, perform feature extraction on the input clothing image through a convolution layer, and use multiple convolution kernels of different sizes to capture local features in the image, and then downsample the feature map output by the convolution layer through a pooling layer to reduce the dimension of the feature map, and then input the pooled features into a fully connected layer for integration and classification, and finally output recognition results of clothing types and style types; based on the recognition results, classify multiple clothing images according to preset classification rules.

[0042] In some possible implementations of the second aspect, the target clothing type information is a top and / or a coat; the second extraction module is specifically used to: convert each clothing image in the candidate clothing image set into a grayscale image, and separate the clothing main area image by using a grayscale threshold segmentation algorithm; execute a Canny edge detection algorithm on the main area image to extract a continuous edge contour curve, determine the contour shape of the clothing according to the geometric features of the contour curve, and mark the contour shape data as the first feature data; based on the clothing main area image, determine the neckline area by using a regional growing algorithm to obtain a neckline area image; in the neckline area image, extract the neckline edge corner point features by using a Harris corner point detection algorithm, and The neckline shape category and decorative detail features are analyzed according to the distribution, quantity and relative position relationship of corner points, and marked as the second feature data, and the extraction range of the second feature data is limited to the neckline area image within the overall clothing outline framework determined by the first feature data; based on the clothing main area image and the first feature data, a shape template matching algorithm is used to match a plurality of pre-set cuff shape templates with each part of the main area image to obtain the cuff area image; within the cuff area image, the SIFT algorithm is used to extract the cuff texture features and shape features, and the cuff design data is obtained and marked as the third feature data; the first feature data, the second feature data and the third feature data constitute a plurality of key features to be fused.

[0043] In some possible implementations of the second aspect, the clothing detail feature intelligent fusion system based on fashion trends also includes a style comparison module, which is used to obtain a target style type; compare and analyze each output clothing design scheme whose aesthetic score is greater than a preset threshold with the target style type, calculate the similarity value, and output the clothing design scheme whose similarity value is greater than the preset threshold.

[0044] In some possible implementations of the second aspect, the first fusion module is specifically used to combine multiple key features to be fused using a permutation and combination algorithm to obtain multiple clothing design solutions. In some possible implementations of the second aspect, the first acquisition module is specifically used to collect multiple clothing images that conform to fashion trends from multiple preset fashion information platforms using web crawler technology.

[0045] In a third aspect, an embodiment of the present application provides an electronic device, comprising one or more processors; a storage device on which one or more programs are stored; when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any technical solution of the first aspect.

[0046] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any technical solution of the first aspect is implemented.

[0047] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method described in any technical solution of the first aspect.

[0048] Among them, the technical effects brought about by any design method in the second to fifth aspects can refer to the technical effects brought about by different design methods in the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0050] Figure 1 A schematic diagram of a flow chart of a method for intelligently fusing clothing detail features based on fashion trends provided in some embodiments of the present application;

[0051] Figure 2A schematic diagram of the structure of a clothing detail feature intelligent fusion system based on fashion trends provided in some embodiments of the present application;

[0052] Figure 3 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present application. DETAILED DESCRIPTION

[0053] Specific embodiments of the present invention will now be mentioned in detail. Although the present invention is described in conjunction with these specific embodiments, it should be appreciated that it is not intended to limit the present invention to these specific embodiments. On the contrary, these embodiments are intended to cover substitutions, changes or equivalent embodiments that may be included in the spirit and scope of the invention defined by the claims. In the following description, a large number of specific details are set forth in order to provide a comprehensive understanding of the present invention. The present invention may be implemented without some or all of these specific details.

[0054] When used in conjunction with "including," "methods comprising," or similar language in this specification and the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0055] Application Overview:

[0056] In today's rapidly developing clothing industry, keeping up with fashion trends and innovatively designing unique clothing styles have become the key to brand competition. However, the traditional clothing design process often relies on the designer's personal experience and subjective judgment, which is particularly insufficient in today's ever-changing fashion environment. Designers are usually limited by their own aesthetics and experience when creating, and it is difficult to fully and accurately capture the popular elements and clothing details in the current market.

[0057] In response to the above technical problems, the overall idea of ​​the technical solution provided by the present application is as follows: a method for intelligent fusion of clothing detail features based on fashion trends is provided, comprising the following steps: obtaining multiple clothing images that conform to fashion trends; using image recognition technology to identify the clothing type information and style type information corresponding to each clothing image, and classifying the multiple clothing images according to preset classification rules, and marking each clothing image with a first classification label of the clothing type and style type information to which it belongs; obtaining target clothing type information; extracting multiple clothing images that match the target clothing type information from the classified clothing images based on the first classification label to obtain a set of candidate clothing images; using image recognition technology to extract multiple key features of each clothing image in the candidate clothing image set to obtain multiple key features to be fused, and marking each key feature with a second classification label of the clothing type to which it belongs; based on the second classification label, combining the multiple key features to be fused to obtain multiple clothing design schemes; analyzing the aesthetic score of each clothing design scheme, and outputting clothing design schemes whose aesthetic scores are greater than a preset threshold.

[0058] This method builds a rich data base by acquiring multiple clothing images that match fashion trends. These images provide an objective basis for analysis and reduce the subjective bias of designers in the source of inspiration. Next, image recognition technology is used to classify clothing images and mark clothing type and style type information. This process ensures the standardization and accuracy of the data, thereby overcoming the errors that may be caused by manual classification.

[0059] After obtaining the target clothing type information, the system extracts clothing images that match the target type from the classified images, ensuring the pertinence of the design scheme. By extracting multiple key features of the candidate clothing images, a feature set to be fused is formed. These features are based on objective analysis and can fully reflect the popular design details, avoiding omissions that may be caused by designers' personal aesthetics and experience limitations. Subsequently, the system combines multiple key features to be fused to generate multiple clothing design schemes. This step relies on a data-driven approach to ensure that the generated design scheme can integrate different popular elements and increase the diversity and innovation of the design.

[0060] Finally, by analyzing the aesthetic score of each clothing design solution, the system outputs solutions with scores higher than the preset threshold. This step not only improves design efficiency, but also ensures that the output design solution is aesthetically competitive in the market. In summary, this technical solution reduces the reliance on designers' personal experience and subjective judgment through intelligent and systematic methods, and improves the innovation and adaptability of the design.

[0061] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced in conjunction with the accompanying drawings. Figure 1 The embodiment of the present application provides a method for intelligently fusing clothing detail features based on fashion trends, comprising the following steps:

[0062] S101: Acquire multiple clothing images that conform to fashion trends;

[0063] Specifically, in some embodiments, multiple clothing images that conform to fashion trends can be collected from multiple preset fashion information platforms through web crawler technology. This solves the problem that designers have limited channels to obtain fashion trend information and are not timely and comprehensive enough, which is conducive to keeping designs in line with fashion trends.

[0064] Of course, the present application is not limited thereto. In other embodiments, multiple clothing images that conform to fashion trends and are manually imported by a user may be acquired through a user interaction interface.

[0065] S102: Using image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classifying multiple clothing images according to preset classification rules, and marking each clothing image with a first classification label of its clothing type and style type information.

[0066] Exemplarily, the preset classification rules may be classified and stored according to preset classification rules such as clothing type (such as tops, pants, skirts, etc.), style (such as casual, business, sports, etc.), applicable groups (such as men, women, children), etc.

[0067] Specifically, in some embodiments, the following steps may be used to use image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classify multiple clothing images according to preset classification rules:

[0068] The first step is to obtain the pixel data of the clothing image and preprocess the pixel data to enhance the image quality;

[0069] In the second step, the preprocessed clothing image is input into a pre-trained deep learning model based on a convolutional neural network. The input clothing image is feature extracted through the convolution layer, and local features in the image are captured using multiple convolution kernels of different sizes. The feature map output by the convolution layer is then downsampled through the pooling layer to reduce the dimension of the feature map. The pooled features are then input into the fully connected layer for integration and classification, and finally the recognition results of clothing type and style type are output.

[0070] The third step is to classify multiple clothing images according to the preset classification rules based on the recognition results. This can improve the accuracy and efficiency of image recognition, determine the type and style of clothing more accurately, and provide a reliable basis for subsequent design processing.

[0071] S103: Obtain target clothing type information;

[0072] Exemplarily, in some embodiments, the target clothing type information is a coat and / or a jacket. Of course, the present application is not limited thereto. In other embodiments, the target clothing type information can also be a dress, pants, etc.

[0073] S104: extracting a plurality of clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set;

[0074] S105: using image recognition technology to extract multiple key features of each clothing image in the candidate clothing image set, obtaining multiple key features to be fused, and marking a second classification label of the clothing type to which each key feature belongs;

[0075] Specifically, in some embodiments, multiple key features of each clothing image in the candidate clothing image set can be extracted using image recognition technology through the following steps to obtain multiple key features to be fused:

[0076] The first step is to convert each clothing image in the candidate clothing image set into a grayscale image, and use the grayscale threshold segmentation algorithm to separate the clothing main area image;

[0077] The second step is to perform the Canny edge detection algorithm on the main area image, extract the continuous edge contour curve, determine the contour shape of the clothing according to the geometric features of the contour curve, and mark the contour shape data as the first feature data;

[0078] The third step is to determine the neckline area based on the clothing main area image using the region growing algorithm to obtain the neckline area image;

[0079] The fourth step is to extract the corner point features of the neckline edge in the neckline area image by using the Harris corner point detection algorithm, analyze the neckline shape category and decorative detail features according to the distribution, quantity and relative position relationship of the corner points, and mark them as the second feature data, and the extraction range of the second feature data is limited to the neckline area image within the overall outline framework of the clothing determined by the first feature data;

[0080] Step 5: Based on the clothing main body area image and the first feature data, a shape template matching algorithm is used to match a plurality of pre-set cuff shape templates with each part of the main body area image to obtain a cuff area image;

[0081] In the sixth step, the SIFT algorithm is used to extract the texture and shape features of the cuffs in the cuff area image, and the cuff design data is obtained and marked as the third feature data; the first feature data, the second feature data and the third feature data constitute multiple key features to be fused. In this way, the detailed features of the clothing design can be captured more accurately, providing more abundant construction elements for subsequent design solutions, and enhancing the diversity and aesthetics of the design.

[0082] S106: Based on the second classification label, multiple key features to be fused are combined to obtain multiple clothing design solutions;

[0083] Specifically, in some embodiments, a plurality of key features to be fused can be combined using a permutation and combination algorithm to obtain a plurality of clothing design solutions. In this way, a large number of design combination possibilities can be quickly generated, the diversity and innovation of the design can be increased, the number of designs that a designer can conceive within a limited time can be broken, and more creative options can be provided.

[0084] Of course, the present application is not limited to this. In other embodiments, the attention mechanism can also be used to analyze different detail features in the clothing design scheme, obtain visual attention data of each detail feature, determine the visual importance weights of different detail features based on the visual attention data, and combine and adjust the detail features according to the visual importance weights.

[0085] S107: Analyze the aesthetic score of each clothing design scheme, and output clothing design schemes whose aesthetic scores are greater than a preset threshold.

[0086] Preferably, in some embodiments, the method for intelligently fusing clothing detail features based on fashion trends further includes the following steps:

[0087] The first step is to obtain the target style type;

[0088] In the second step, each output clothing design scheme with an aesthetic score greater than a preset threshold is compared and analyzed with the target style type, the similarity value is calculated, and the clothing design scheme with a similarity value greater than the preset threshold is output. In this way, it can ensure that the output design scheme is more in line with the specific target style type and improve the matching degree between the design scheme and the expected style.

[0089] See also Figure 2 Based on the same inventive concept as the method for intelligently fusing clothing detail features based on fashion trends in the aforementioned embodiment, the embodiment of the present application provides an intelligent fusing system for clothing detail features based on fashion trends, including:

[0090] A first acquisition module 201 is used to acquire a plurality of clothing images that conform to fashion trends;

[0091] The first recognition module 202 is used to use image recognition technology to identify the clothing type information and style type information corresponding to each clothing image, and classify the multiple clothing images according to a preset classification rule, and mark each clothing image with a first classification label of the clothing type and style type information to which it belongs;

[0092] The second acquisition module 203 is used to acquire target clothing type information;

[0093] A first extraction module 204 is used to extract multiple clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set;

[0094] The second extraction module 205 is used to extract multiple key features of each clothing image in the candidate clothing image set by using image recognition technology, obtain multiple key features to be fused, and mark each key feature with a second classification label of the clothing type to which it belongs;

[0095] A first fusion module 206 is used to combine multiple key features to be fused based on the second classification label to obtain multiple clothing design solutions;

[0096] The first scoring module 207 is used to analyze the aesthetic score of each clothing design scheme and output clothing design schemes whose aesthetic scores are greater than a preset threshold.

[0097] In some embodiments, the first recognition module 202 is specifically used to: obtain pixel data of a clothing image, and preprocess the pixel data to enhance image quality; input the preprocessed clothing image into a pre-trained deep learning model based on a convolutional neural network, extract features of the input clothing image through a convolution layer, and use multiple convolution kernels of different sizes to capture local features in the image, and then downsample the feature map output by the convolution layer through a pooling layer to reduce the dimension of the feature map, and then input the pooled features into a fully connected layer for integration and classification, and finally output recognition results of clothing types and style types; based on the recognition results, classify multiple clothing images according to preset classification rules.

[0098] In some embodiments, the target clothing type information is a top and / or a coat; the second extraction module 205 is specifically used to: convert each clothing image in the candidate clothing image set into a grayscale image, and separate the clothing main area image by using a grayscale threshold segmentation algorithm; perform a Canny edge detection algorithm on the main area image to extract a continuous edge contour curve, determine the contour shape of the clothing according to the geometric features of the contour curve, and mark the contour shape data as the first feature data; based on the clothing main area image, determine the neckline area by using a regional growing algorithm to obtain a neckline area image; in the neckline area image, extract the neckline edge corner point features by using a Harris corner point detection algorithm, and The collar shape category and decorative detail features are analyzed by distribution, quantity and relative position relationship, and marked as the second feature data, and the extraction range of the second feature data is limited to the collar area image within the overall clothing outline framework determined by the first feature data; based on the clothing main area image and the first feature data, a shape template matching algorithm is used to match a plurality of pre-set cuff shape templates with each part of the main area image to obtain the cuff area image; within the cuff area image, the SIFT algorithm is used to extract the cuff texture features and shape features, and the cuff design data is obtained and marked as the third feature data; the first feature data, the second feature data and the third feature data constitute a plurality of key features to be fused.

[0099] In some embodiments, the clothing detail feature intelligent fusion system based on fashion trends also includes a style comparison module 208, which is used to obtain a target style type; each clothing design scheme whose aesthetic score is greater than a preset threshold is compared and analyzed with the target style type, the similarity value is calculated, and the clothing design scheme whose similarity value is greater than the preset threshold is output.

[0100] In some embodiments, the first fusion module 206 is specifically used to combine multiple key features to be fused using a permutation and combination algorithm to obtain multiple clothing design solutions. In some embodiments, the first acquisition module is specifically used to collect multiple clothing images that meet fashion trends from multiple preset fashion information platforms using web crawler technology.

[0101] It is understandable that the modules recorded in the intelligent fusion system of clothing details features based on fashion trends are similar to those in the reference Figure 1 The steps in the method for intelligently integrating clothing detail features based on fashion trends correspond to each other. Therefore, the operations, features and beneficial effects described above for the method are also applicable to the intelligent system for intelligently integrating clothing detail features based on fashion trends and the modules contained therein, and will not be repeated here.

[0102] See also Figure 3Based on the inventive concept of a method for intelligently integrating clothing detail features based on fashion trends in the aforementioned embodiment, an embodiment of the present application provides an electronic device. The electronic device may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), etc., and fixed terminals such as digital TVs, desktop computers, etc. The electronic device includes a processing device 301 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a ROM 302 (read-only memory) or a program loaded from a storage device 308 into a RAM 303 (random access memory). In RAM 303, various programs and data required for the operation of the electronic device are also stored. The processing device 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An input / output interface (i.e., an I / O interface 305) is also connected to the bus 304.

[0103] Typically, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data.

[0104] In particular, according to some embodiments of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present application are executed.

[0105] It should be noted that the computer-readable medium recorded in some embodiments of the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present application, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In some embodiments of the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0106] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperTextTransferProtocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an adhoc peer-to-peer network), as well as any currently known or future developed network.

[0107] The computer-readable medium may be included in the electronic device; or it may exist independently without being assembled into the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains multiple clothing images that conform to the fashion trend; uses image recognition technology to identify the clothing type information and style type information corresponding to each clothing image, and classifies the multiple clothing images according to the preset classification rules, and marks each clothing image with a first classification label of the clothing type and style type information to which it belongs; obtains target clothing type information; extracts multiple clothing images that match the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set; uses image recognition technology to extract multiple key features of each clothing image in the candidate clothing image set to obtain multiple key features to be fused, and marks each key feature with a second classification label of the clothing type to which it belongs; based on the second classification label, combines multiple key features to be fused to obtain multiple clothing design schemes; analyzes the aesthetic score of each clothing design scheme, and outputs clothing design schemes with aesthetic scores greater than a preset threshold.

[0108] Computer program code for performing the operations of some embodiments of the present application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).

[0109] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0110] The modules described in some embodiments of the present application may be implemented by software or hardware. The modules described may also be set in a processor: for example, they may be described as: a first acquisition module, a first recognition module, a second acquisition module, a first extraction module, a second extraction module, a first fusion module, a first scoring module, and a style comparison module. The names of these modules do not, in some cases, constitute limitations on the modules themselves. For example, the first acquisition module may also be described as a "clothing image acquisition module."

[0111] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0112] Some embodiments of the present application also provide a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for intelligently fusing clothing detail features based on fashion trends.

[0113] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. An intelligent fusion method of clothing detail features based on fashion trends, characterized in that: The following steps are involved: Obtain multiple clothing images that match fashion trends; Using image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classifying multiple clothing images according to a preset classification rule, and marking a first classification label of the clothing type and style type information to which each clothing image belongs; Get target clothing type information; extracting a plurality of clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set; Using image recognition technology to extract multiple key features of each clothing image in the candidate clothing image set, multiple key features to be fused are obtained, and a second classification label of the clothing type to which each key feature belongs is marked; Based on the second classification label, multiple key features to be fused are combined to obtain multiple clothing design solutions; Analyze the aesthetic score of each clothing design scheme and output clothing design schemes whose aesthetic score is greater than a preset threshold.

2. The method for intelligently integrating clothing detail features based on fashion trends according to claim 1 is characterized in that: The steps of using image recognition technology to identify clothing type information and style type information corresponding to each clothing image, and classifying the plurality of clothing images according to a preset classification rule include: Acquire pixel data of the clothing image, and pre-process the pixel data to enhance the image quality; The preprocessed clothing image is input into a pre-trained deep learning model based on a convolutional neural network. The input clothing image is feature extracted through the convolution layer, and local features in the image are captured using multiple convolution kernels of different sizes. The feature map output by the convolution layer is then downsampled through the pooling layer to reduce the dimension of the feature map. The pooled features are then input into the fully connected layer for integration and classification, and finally the recognition results of clothing type and style type are output. Based on the recognition result, the multiple clothing images are classified according to preset classification rules.

3. The method for intelligently integrating clothing detail features based on fashion trends according to claim 1 is characterized in that: The target clothing type information is a top and / or a coat; the steps of extracting multiple key features of each clothing image in the candidate clothing image set by using image recognition technology to obtain multiple key features to be fused include: Each clothing image in the candidate clothing image set is converted into a grayscale image, and the clothing main body area image is separated using a grayscale threshold segmentation algorithm; Execute the Canny edge detection algorithm on the main body area image to extract a continuous edge contour curve, determine the contour shape of the clothing according to the geometric features of the contour curve, and mark the contour shape data as the first feature data; Based on the clothing main body area image, a neckline area is determined by using a region growing algorithm to obtain a neckline area image; In the neckline area image, the corner point features of the neckline edge are extracted by the Harris corner point detection algorithm, and the neckline shape category and decorative detail features are analyzed according to the distribution, quantity and relative position relationship of the corner points, and marked as second feature data, and the extraction range of the second feature data is limited to the neckline area image within the overall outline framework of the clothing determined by the first feature data; Based on the clothing main body area image and the first feature data, a shape template matching algorithm is used to match a plurality of pre-set cuff shape templates with each part of the main body area image to obtain a cuff area image; In the cuff area image, the SIFT algorithm is used to extract the cuff texture features and shape features, and the cuff design data is obtained and marked as the third feature data; the first feature data, the second feature data and the third feature data constitute a plurality of key features to be fused.

4. The method for intelligently integrating clothing detail features based on fashion trends according to claim 1 is characterized in that: The following steps are also included: Get the target style type; Each output clothing design scheme having an aesthetic score greater than a preset threshold is compared and analyzed with the target style type, a similarity value is calculated, and clothing design schemes having a similarity value greater than the preset threshold are output.

5. The method for intelligently integrating clothing detail features based on fashion trends according to claim 1 is characterized in that: The permutation and combination algorithm is used to combine multiple key features to be fused to obtain multiple clothing design solutions.

6. The method for intelligently fusing clothing detail features based on fashion trends according to claim 1 is characterized in that: Through web crawler technology, multiple clothing images that conform to fashion trends are collected from multiple preset fashion information platforms.

7. An intelligent fusion system of clothing detail features based on fashion trends, characterized in that: include: A first acquisition module is used to acquire a plurality of clothing images that conform to fashion trends; A first recognition module is used to use image recognition technology to identify the clothing type information and style type information corresponding to each clothing image, and classify the multiple clothing images according to a preset classification rule, and mark each clothing image with a first classification label of the clothing type and style type information to which it belongs; The second acquisition module is used to acquire target clothing type information; A first extraction module is used to extract a plurality of clothing images matching the target clothing type information from the classified clothing images based on the first classification label to obtain a candidate clothing image set; The second extraction module is used to extract multiple key features of each clothing image in the candidate clothing image set by using image recognition technology, obtain multiple key features to be fused, and mark each key feature with a second classification label of the clothing type to which it belongs; A first fusion module, configured to combine a plurality of key features to be fused based on the second classification label to obtain a plurality of clothing design solutions; The first scoring module is used to analyze the aesthetic score of each clothing design scheme and output the clothing design schemes whose aesthetic scores are greater than a preset threshold.

8. An electronic device, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processing device, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processing device, the method according to any one of claims 1 to 6 is implemented.

Images