WO2019237721A1

WO2019237721A1 - Garment dimension data identification method and device, and user terminal

Info

Publication number: WO2019237721A1
Application number: PCT/CN2018/125516
Authority: WO
Inventors: 斯科特马修•罗伯特; 黄鼎隆; 唐颖雯; 王海涵
Original assignee: 深圳码隆科技有限公司
Priority date: 2018-06-14
Filing date: 2018-12-29
Publication date: 2019-12-19
Also published as: CN108805135A

Abstract

A garment dimension data identification method and device, and a user terminal. The method comprises: pre-processing a target image, and determining an overall outline of a garment in the target image (S100); according to the overall outline of the garment, capturing a minimum screenshot comprising the overall outline of the garment in the target image (S200); extracting a style characteristic point in the minimum screenshot (S300); determining whether the style characteristic point in the minimum screenshot comprises a separable characteristic (S400); if yes, separating the minimum screenshot according to the separable characteristic, obtaining a garment separating unit, and using same as a regional image to be identified (S500); and identifying the regional image to be identified, so that the garment dimension data in the regional image to be identified is obtained (S600). According to the method, whether the garment style is separable can be determined based on a computer image identification technology; furthermore, image identification is carried out after separation processing, thereby obtaining the garment dimension data; automatic identification on the garment dimension data comprised in the target image is completed; the measurement speed is rapid; the efficiency is high; and the measurement accuracy is improved.

Description

Clothing size data identification method, device and user terminal

This application is based on a Chinese patent application with an application number of 201810613349.0 and an application date of June 14, 2018, and claims its priority. The entire content of this application is hereby incorporated into this application as a whole.

Technical field

The present application relates to the field of image recognition technology, and more particularly, to a method, a device, and a user terminal for recognizing clothing size data.

Background technique

Clothing style refers to the style of clothing, usually refers to the shape factor, which is one of the styling elements. Good clothing style can bring people a good mood all day. Clothing styles generally consist of three aspects: structure, popular elements, and texture. The structure includes the shape and structural characteristics of the garment, and the corresponding size data of the shape and structural characteristics.

Clothing size data can generally include: bust, waist, hips, arm length, shoulder width, height, neck circumference, upper bust, chest height, lower bust, arm circumference, sleeve cage, wrist circumference, shoulder to waist height, Waist-to-skirt length, chest distance, etc. Different clothing styles correspond to different clothing size data.

At present, in the method for acquiring the size data of the clothing styles in the clothing image, because there are many patterns of clothing that cannot be distinguished by automatic computer recognition, only a manual measurement tool or computer-assisted can be used for the single clothing image. Dimension measurement, and statistical calculation, classification and sorting according to experience. The technical expertise of the measurement personnel of dimensional data is relatively high. The measurement of dimensional data in the picture is slow, inefficient, and low in accuracy, which cannot meet the ever-changing The speed of development of clothing diversification.

Application content

In view of this, the present application provides a clothing size data identification method, device and user terminal to solve the shortcomings of the prior art.

To solve the above problems, the present application provides a method for identifying clothing size data, including:

Preprocess the target image to determine the overall clothing outline in the target image;

Taking a minimum screenshot of the overall outline of the clothing according to the overall outline of the clothing;

Extracting style feature points in the minimum screenshot;

Judging whether the style feature points in the minimum screenshot include a detachable feature;

If the minimum screenshot includes the detachable feature, performing a split processing on the minimum screenshot according to the detachable feature to obtain a clothing split unit as an image of a region to be identified;

Identify the image of the area to be identified, and obtain clothing size data in the image of the area to be identified.

Preferably, the "preprocessing the target image and determining the overall outline of the clothing in the target image" includes:

Performing human body feature positioning on the target image to obtain multiple human body feature points;

Determining the number of model portraits included in the target image according to each human body feature location point, and obtaining a selection box including the human body feature location point for each model portrait separately;

Combining all the selection boxes of each of the model portraits in the target image to obtain a human feature positioning area;

Perform edge detection on the human body feature positioning area to determine the overall outline of the garment including all human body feature points in the human body feature positioning area.

Preferably, the "determining whether the style feature point in the minimum screenshot contains a detachable feature" and the "recognizing the image of the area to be identified and obtaining the clothing in the image of the area to be identified Dimension Data "also includes:

If the minimum screenshot does not include detachable features, the minimum screenshot is used as the image of the area to be identified.

Preferably, the "recognizing the image of the area to be identified and obtaining clothing size data in the image of the area to be identified" includes:

Determining a scale of a model portrait of the target image;

Positioning the ruler feature points of the image of the area to be identified, and obtaining the clothing size data based on the ruler feature points based on the scale.

Preferably, the "determining the scale of the model portrait of the target image" includes:

Obtaining the maximum height value of the overall outline of the clothing;

Positioning a waist region in the overall outline of the garment, and determining a maximum width value of the waist region;

Obtaining a preset height value and a preset waist width value; and obtaining the preset height value and the maximum height value according to a quantitative relationship between the preset waist value and the maximum width value The scale of the target image.

Preferably, the "locating the ruler feature points of the image of the area to be identified, and obtaining the clothing size data based on the ruler feature points based on the scale" includes:

Locate all row ruler feature points in the image of the area to be identified;

Connect each of the ruler feature points to obtain a preliminary connection line between each of the ruler feature points;

Extracting the style feature connecting lines from the preliminary connecting lines according to a preset style feature database;

Obtain the length ratio values of all the style feature connecting lines, and obtain the clothing size data according to the scale bar.

In addition, in order to solve the above problems, the present application also provides a clothing size data identification device, including: a preprocessing module, an interception module, an extraction module, a judgment module, a splitting module, and an acquisition module;

The pre-processing module is configured to pre-process the target image to determine the overall clothing outline in the target image;

The intercepting module is configured to intercept the smallest screenshot of the overall outline of the clothing in the target image according to the overall outline of the clothing;

The extraction module is used to extract the style feature points in the minimum screenshot;

The judging module is configured to judge whether the style feature point in the minimum screenshot includes a detachable feature;

The splitting module is configured to perform splitting processing on the smallest screenshot according to the splittable feature when the splittable feature is included in the smallest screenshot to obtain a clothing splitting unit as an image of a region to be identified;

The acquisition module is configured to identify the image of the area to be identified, and acquire clothing size data in the image of the area to be identified.

In addition, in order to solve the above problem, the present application further provides a user terminal including a memory and a processor, where the memory is used to store a clothing size data identification program, and the processor runs the clothing size data identification program to enable the clothing size data identification program to The user terminal executes the clothing size data identification method as described above.

In addition, in order to solve the above problem, the present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a clothing size data identification program, and the clothing size data identification program is implemented as described above when executed by a processor. The clothing size data identification method.

A clothing size data identification method, device, and user terminal provided by the present application. The method provided in this application includes: pre-processing the target image to determine the overall clothing outline in the target image; taking the minimum overall screenshot of the target image containing the overall clothing outline according to the overall clothing outline; extracting the style in the minimum screenshot Feature points; determine whether the style feature points in the smallest screenshot include detachable features; if the smallest screenshot contains detachable features, the smallest screenshot is split based on the detachable features to obtain a clothing split unit as The image of the area to be identified; identify the image of the area to be identified, and obtain the clothing size data in the image of the area to be identified. The method provided in this application determines the overall outline in the target image to obtain the smallest screenshot, and then determines whether the smallest screenshot contains a detachable feature. If the feature is included, the minimum is performed according to the detachable feature. Split processing of screenshots and recognition of myopia images to obtain clothing size data. The method provided in the present application implements the judgment of whether a clothing style is detachable based on computer image recognition technology, and performs image recognition after the splitting process, thereby obtaining clothing size data and completing the processing of the target image. The automatic statistical calculation, classification and sorting of the included clothing size data, fast measurement speed, high efficiency, and improved measurement accuracy, provide convenience for the measurement of clothing size data in the image, and fully meet the changing clothing diversity Speed of development.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic structural diagram of a hardware operating environment involved in a solution of an embodiment of a method for identifying clothing size data of this application;

FIG. 2 is a schematic flowchart of a first embodiment of a clothing size data identification method of the present application; FIG.

FIG. 3 is a schematic flowchart of a second embodiment of a clothing size data identification method of the present application; FIG.

FIG. 4 is a schematic flowchart of a third embodiment of a clothing size data identification method of the present application; FIG.

5 is a detailed flowchart of steps S610 and S620 of a third embodiment of a method for identifying clothing size data of this application;

FIG. 6 is a functional module diagram of the clothing size data identification device of the present application.

The implementation, functional characteristics and advantages of the purpose of this application will be further described with reference to the embodiments and the drawings.

detailed description

The embodiments of the present application are described in detail below, in which the same or similar reference numerals indicate the same or similar elements or elements having the same or similar functions.

In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality" is two or more, unless it is specifically and specifically defined otherwise.

In this application, the terms "installation," "connected," "connected," and "fixed" should be understood broadly unless otherwise specified and limited, for example, they may be fixed connections or removable connections Or integrated; it can be mechanical or electrical; it can be directly connected or indirectly connected through an intermediate medium; it can be the internal connection of two elements or the interaction between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific situations.

It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

As shown in FIG. 1, FIG. 1 is a schematic structural diagram of a hardware operating environment of a terminal involved in a solution according to an embodiment of the present application.

In the embodiment of the present application, the terminal may be a PC provided in an automatic container machine, or may be a mobile terminal device such as a smart phone, a tablet computer, an e-book reader, an MP3 player, an MP4 player, and a portable computer. In addition, it may be a computer hardware device provided in the automatic container machine itself.

As shown in FIG. 1, the terminal may include a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. The communication bus 1002 is used to implement connection and communication between these components. The user interface 1003 may include a display screen, an input unit such as a keyboard, a remote controller, and the optional user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory, such as a magnetic disk memory. The memory 1005 may optionally be a storage device independent of the foregoing processor 1001.

Optionally, the terminal may further include an RF (Radio Frequency) circuit, a sensor, an audio circuit, a WiFi module, and the like. In addition, the mobile terminal may be configured with other sensors such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, and the like, which will not be repeated here.

Those skilled in the art can understand that the terminal shown in FIG. 1 does not constitute a limitation on the terminal, and may include more or fewer components than those shown in the figure, or combine some components, or arrange different components.

As shown in FIG. 1, the memory 1005 as a computer-readable storage medium may include an operating system, a data interface control program, a network connection program, and a clothing size data identification program.

A clothing size data identification method, device, and user terminal provided by the present application. Wherein, the method realizes judging whether the clothing style is detachable based on computer image recognition technology, and recognizes the image after the splitting process, thereby obtaining clothing size data and completing the Automatic statistical calculation, classification and sorting of clothing size data, fast measurement speed, high efficiency, improved measurement accuracy, facilitate the measurement of clothing size data in the image, and fully meet the rapid development of clothing diversification speed.

Example 1:

Referring to FIG. 2, a first embodiment of the present application provides a clothing size data identification method, including:

Step S100: preprocess the target image to determine the overall clothing outline in the target image;

As described above, the method provided in this embodiment recognizes a target image containing clothing based on a computer image recognition technology. Among them, the target image may be a display picture, a catwalk picture, a life picture of a model wearing clothing, or a pure clothing picture without a model.

As mentioned above, in this embodiment, the source of the target image can be directed to a large number of photos of models wearing fashion clothing taken at places such as Fashion Week, Fashion Week, catwalk events, etc., and the method provided in this embodiment Recognize the above photos in order to extract the size data of the fashion clothes in them, so as to grasp the fashion elements.

In addition, the source of the pictures can be crawled through the web, and the web sites and databases that contain photos of models wearing fashion clothes are captured in batches. For example, the official websites of major clothing brands, etc., and then use the method provided in this embodiment to batch identify the above-mentioned photos, so as to extract the size data of the fashion clothing therein and grasp the fashion elements.

As described above, the pre-processing of the image may include tilt correction for the image, smoothing processing of the contour curve, and edge detection.

For image pre-processing, it can be the adjustment of the image itself, which can include:

Convert the target image to the format corresponding to the video image analysis program, convert it to a grayscale image, and obtain a grayscale distribution histogram;

The grayscale image is subjected to inverse color processing, edge detection is performed, and the overall clothing outline in the grayscale image after inversion is obtained based on the grayscale distribution histogram.

Step S200: Take a minimum screenshot of the overall outline of the clothing according to the overall outline of the clothing;

As described above, after the overall outline of the clothing is obtained from the target image, a screenshot is taken of the target image based on the overall outline, and a screenshot including the overall outline of the clothing is taken. The screenshot may be a screenshot that is suitable for the overall size of the overall outline of the clothing. Take a cut along the outline; it may also be appropriate to expand a certain range in consideration of the accuracy of the overall outline of the garment or the size of the edge, and take a screenshot that includes the overall outline of the garment.

Step S300, extracting style feature points in the minimum screenshot;

As mentioned above, the style feature points are characteristic points that characterize different styles of different garments. For example, skirts include dresses and skirts; pants include shorts, leggings, flared pants, cropped pants, etc. Different styles of clothing have different characteristics, and the style characteristics reflect the differences of different styles of clothing. For example, as a different part of the top, the short skirt may include, but is not limited to, different colors from the top, the width of different regions, the height of different regions, and materials.

Step S400, determine whether the style feature points in the minimum screenshot include a detachable feature;

As mentioned above, the style feature points include detachable feature points and indivisible feature points. Among them, the detachable feature points represent the characteristics of different detachable parts in the clothing. For example, clothes and pants, shirts and short sleeves, coats and inner shirts and trousers, short-sleeved tops and skirts of the lower body, sportswear tops and sportswear pants of different or the same fabric or color, etc., are all removable The sub-units all include different separable features. Among them, the characteristic points can be the characteristics of whether the skirt is waisted, the difference in material and fabric, the difference in color, the size of the legs and the waist.

Step S500: if the minimum screenshot includes the detachable feature, perform a split processing on the minimum screenshot according to the detachable feature to obtain a clothing split unit as an image of a region to be identified;

If a detachable feature is included, the smallest screenshot is split to obtain a split unit, which is used as an image of the area to be identified.

As described above, each target image may include multiple minimum screenshots, and if each minimum screenshot includes detachable features, multiple recognition area images may be obtained after splitting. Labels can be set for different minimal screenshots and different images of the area to be identified, and different identification channels can be established according to the labels, so as to perform image recognition for different areas to be identified.

Step S600: Recognize the image of the area to be identified, and obtain clothing size data in the image of the area to be identified.

As mentioned above, the image of the area to be identified is identified, thereby obtaining clothing size data.

The step S400 "determines whether the style feature point in the minimum screenshot includes a detachable feature" and the step S600 "recognizes the image of the region to be identified and obtains the image in the image of the region to be identified Clothing size data "also includes:

In step S700, if the minimum screenshot does not include a detachable feature, the minimum screenshot is used as the image of the area to be identified.

The method provided in this embodiment obtains the smallest screenshot by determining the overall contour in the target image, and then determines whether the smallest screenshot contains a detachable feature. If the feature is included, an analysis is performed according to the detachable feature. The smallest screenshot is split and processed, and myopia image recognition is performed to obtain clothing size data. The method provided in this embodiment implements the determination of whether a clothing style is detachable based on computer image recognition technology, and performs image recognition after the split processing, thereby obtaining clothing size data and completing the target image. The automatic statistical calculation, classification and sorting of the clothing size data contained in it, the measurement speed is fast, the efficiency is high, and the accuracy of the measurement is improved, which facilitates the measurement of the clothing size data in the image and fully meets the changing clothing Diversified development speed.

Example 2:

Referring to FIG. 3, a second embodiment of the present application provides a clothing size data identification method. Based on the first embodiment shown in FIG. 2 described above, the step S100 is "pre-processing a target image to determine the target image. "Overall clothing outline" includes:

Step S110: localize human features on the target image to obtain multiple human feature points;

As mentioned above, human body feature positioning is to perform feature localization on the human body contained in the image. Human body images include limb features, such as head, neck, torso, and limbs; learn through neural network to locate the head, neck, torso, and limbs to find human features, and also add feet, shoes, gestures, If the features of the facial organs and hair include the aforementioned limb features, a plurality of human feature points are obtained.

Step S120: Determine the number of model portraits included in the target image according to each human feature location point, and obtain a selection frame including the human feature location point for each model portrait;

As mentioned above, the selection box is a rough selection box for the anchor points of human characteristics. It is used to select and divide the area where the human body features are located. The shape of the selection box can be different shapes. For example, the selection box corresponding to the characteristic points of the head can be circular, the selection box corresponding to the characteristic points of the legs can be rectangular, and so on.

As described above, each target image may include a plurality of different model portraits, and the number of model portraits included is identified by positioning points of human characteristics. And, for each model portrait in the target image, a selection frame is obtained separately.

Step S130: Combine all the selection boxes of each of the model portraits in the target image to obtain a human feature positioning area;

In the above, all selection boxes corresponding to each model portrait are combined to become a human body feature positioning area. The positioning area includes images of model portraits.

Step S140: Perform edge detection on the human body feature positioning area to determine the overall outline of the garment including all human body feature points in the human body feature positioning area.

For edge detection of human body feature positioning areas, different edge detection algorithms can be used. In this embodiment, a Canny edge detection algorithm is used, and the steps may include:

Perform gray image conversion on human body feature positioning areas;

Gaussian blurring the grayscale image;

Calculate the image gradient, and calculate the image edge amplitude and angle according to the gradient;

Get edge image data through double-threshold edge connection processing;

The binarization processing is performed on the edge graphic data to obtain a binarized image output result, as the overall outline of the clothing including all human body feature points.

In addition, between the steps of “calculating the image gradient and calculating the edge value and angle of the image based on the gradient” and “obtaining the edge image data through the double-threshold edge connection processing”, the method may further include the following steps:

Perform non-maximum signal suppression processing on the image to achieve thinning of the image edges.

As described above, in this embodiment, multiple human feature points are obtained for the target image, and the number of model portraits in the image is determined. Then, the human feature positioning areas for different model portraits are obtained, and the edge detection is performed to obtain the result. The overall outline of the garment in the target image is described. In this way, the overall outline of clothing of multiple different model portraits in the image is automatically obtained, and the recognition efficiency of the overall outline acquisition is greatly improved.

Example 3:

Referring to FIG. 4, a third embodiment of the present application provides a clothing size data recognition method. Based on the first embodiment shown in FIG. 3 described above, the step S600 "recognizes the image of the area to be identified, and obtains said Clothing size data in the image of the area to be identified "includes:

Step S610: Determine the scale of the model portrait of the target image;

As mentioned above, the scale is the ratio of the length of a line segment on the map to the actual length of the corresponding line segment on the ground. The formula is: scale = ratio of distance on the graph to actual distance. The scale of the target image is the ratio of the width or height of the model portrait to the actual distance in the target image.

Step S620: Locate the feature point of the ruler of the image of the area to be identified, and obtain the clothing size data based on the feature point of the ruler based on the scale.

As mentioned above, the characteristic points of the ruler are characteristic points corresponding to different clothing styles. For example, a skirt can include two or more endpoints representing the waist length, two upper or lower endpoints representing the group height, and the skirt Two or more endpoints of the width and so on. As another example, a shirt can include multiple end points of the neckline to indicate its shape, including multiple end points to indicate the width of the shoulders, end points at the cuffs that indicate the length of the sleeves, and end points at the starting point of the sleeves. Endpoints and so on.

In the foregoing, clothing size data is obtained by locating a plurality of different ruler feature points of the image of the area to be identified. The distance between the characteristic points of the row ruler includes clothing size data.

In step S610, "determining the scale of the model portrait of the target image" includes:

Step S611, obtaining a maximum height value of the overall outline of the clothing;

Step S612, positioning a waist region in the overall outline of the garment, and determining a maximum width value of the waist region;

In the above, the maximum height value and the maximum width value of the model portrait are determined.

Step S613: Obtain a preset height value and a preset waist width value; and, according to a quantitative relationship between the preset height value and the maximum height value, and the preset waist width value and the maximum width value, The scale of the target image is displayed.

As described above, the preset height value and the preset waist width value are preset values that are preset manually, and can be set and modified according to actual conditions. For example, the preset height value is 180cm, and the preset waist width value can be 30cm. In addition, a preset hip width value may also be included, for example, it may be 45 cm. The preset height value and preset waist width value are used to calculate the scale corresponding to the target image, which can be adjusted according to the actual situation to improve the accuracy of the calculation.

The step S620, "locating the ruler feature points of the image of the area to be identified, and obtaining the clothing size data based on the ruler feature points based on the scale" includes:

Step S621: Locate all row ruler feature points in the image of the area to be identified;

Step S622: connect each of the ruler feature points to obtain a preliminary connection line between each of the ruler feature points;

As described above, all the ruler feature points are connected, so as to obtain a preliminary connection line between each ruler feature points. The preliminary connection line is the connection line between all the ruler feature points.

Step S623, extracting the style feature connection lines in the preliminary connection line according to a preset style feature database;

As mentioned above, the connecting lines between all the characteristic points of the row ruler include the connecting lines of the style features that can characterize the clothing style, and also include the extra connecting lines other than the connecting lines of the style features. The preset style feature database is a database containing the position range or interval of connecting lines that can characterize the style features. Through the preset style feature data, all the preliminary connection lines are filtered, and the style feature connection lines are extracted to characterize the clothing style. For example, the line between the two ruler feature points of the cuff to the shoulder can represent the data of the sleeve length, while the line between the cuff to the feature ruler of the waist is a redundant connection line. The data was filtered, and the connecting line from the cuff to the shoulder was extracted as one of the connecting lines of the style feature.

Step S624, obtaining length ratio values of all the connecting lines of the style features, and obtaining the clothing size data according to the scale.

According to the connecting lines of the style characteristics, calculate the ratio between all the connecting lines, and obtain the data of the size of the clothing reagent according to the scale.

In this embodiment, the scale of the target image is determined by the maximum height value and the maximum waist width value of the overall outline of the clothing, and the preset height value and the preset waist width value, and the style feature connection is filtered through the preset style feature database. And obtain the ratio of the length of all the characteristics of the style, and then obtain the clothing size data, which improves the accuracy of the clothing size data acquisition and brings convenience to the image clothing measurement work.

In addition, the present application also provides a clothing size data identification device, including: a preprocessing module 10, an interception module 20, an extraction module 30, a judgment module 40, a splitting module 50, and an acquisition module 60;

The pre-processing module 10 is configured to pre-process a target image and determine the overall clothing outline in the target image;

The intercepting module 20 is configured to intercept the smallest screenshot of the overall outline of the clothing according to the overall outline of the clothing;

The extraction module 30 is configured to extract the style feature points in the minimum screenshot;

The judging module 40 is configured to judge whether the style feature point in the minimum screenshot includes a detachable feature;

The splitting module 50 is configured to perform splitting processing on the smallest screenshot according to the splittable feature when the splittable feature is included in the smallest screenshot to obtain a clothing splitting unit as an image of a region to be identified. ;

The obtaining module 60 is configured to identify the image of the area to be identified, and obtain clothing size data in the image of the area to be identified.

In addition, the present application also provides a user terminal, including a memory and a processor, where the memory is used to store a clothing size data identification program, and the processor runs the clothing size data identification program to cause the user terminal to execute the program as described above. The clothing size data identification method.

In addition, the present application also provides a computer-readable storage medium, where the computer-readable storage medium stores a clothing size data identification program, and the clothing size data identification program implements the clothing size data as described above when executed by a processor. recognition methods.

The above-mentioned serial numbers of the embodiments of the present application are merely for description, and do not represent the superiority or inferiority of the embodiments.

Through the description of the above embodiments, those skilled in the art can clearly understand that the methods in the above embodiments can be implemented by means of software plus a necessary universal hardware platform, and of course, also by hardware, but in many cases the former is better. Implementation. Based on such an understanding, the technical solution of the present application, in essence, or a part that contributes to the existing technology, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM) as described above. , Magnetic disk, optical disc), including a number of instructions to enable a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in the embodiments of the present application.

The above are only preferred embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the description and drawings of the application, or directly or indirectly used in other related technical fields Are included in the scope of patent protection of this application.

Claims

A clothing size data identification method, comprising:

Preprocess the target image to determine the overall clothing outline in the target image;

Taking a minimum screenshot of the overall outline of the clothing according to the overall outline of the clothing;

Extracting style feature points in the minimum screenshot;

Judging whether the style feature points in the minimum screenshot include a detachable feature;

If the minimum screenshot includes the detachable feature, performing a split processing on the minimum screenshot according to the detachable feature to obtain a clothing split unit as an image of a region to be identified;

Identify the image of the area to be identified, and obtain clothing size data in the image of the area to be identified.
The method for identifying clothing size data according to claim 1, wherein the "preprocessing the target image to determine the overall contour of the clothing in the target image" comprises:

Performing human body feature positioning on the target image to obtain multiple human body feature points;

Determining the number of model portraits included in the target image according to each human body feature location point, and obtaining a selection box including the human body feature location point for each model portrait separately;

Combining all the selection boxes of each of the model portraits in the target image to obtain a human feature positioning area;

Perform edge detection on the human body feature positioning area to determine the overall outline of the garment including all human body feature points in the human body feature positioning area.
The method for recognizing clothing size data according to claim 1, wherein the "determining whether the style feature points in the smallest screenshot include a detachable feature" and the "for the image of the region to be identified Performing identification and obtaining clothing size data in the image of the area to be identified "further includes:

If the minimum screenshot does not include detachable features, the minimum screenshot is used as the image of the area to be identified.
The method for recognizing clothing size data according to claim 1, wherein the "recognizing the image of the area to be identified and obtaining the clothing size data in the image of the area to be identified" comprises:

Determining a scale of a model portrait of the target image;

Positioning the ruler feature points of the image of the area to be identified, and obtaining the clothing size data based on the ruler feature points based on the scale.
The method for identifying clothing size data according to claim 4, wherein the "determining the scale of the model portrait of the target image" comprises:

Obtaining the maximum height value of the overall outline of the clothing;

Positioning a waist region in the overall outline of the garment, and determining a maximum width value of the waist region;

Obtaining a preset height value and a preset waist width value; and obtaining the preset height value and the maximum height value according to a quantitative relationship between the preset waist value and the maximum width value The scale of the target image.
The method for recognizing clothing size data according to claim 4, wherein the "locating the feature points of the row ruler of the image of the area to be identified is based on the scale, and the garment size data is obtained according to the feature points of the row ruler "include:

Locate all row ruler feature points in the image of the area to be identified;

Connect each of the ruler feature points to obtain a preliminary connection line between each of the ruler feature points;

Extracting the style feature connecting lines from the preliminary connecting lines according to a preset style feature database;

Obtain the length ratio values of all the style feature connecting lines, and obtain the clothing size data according to the scale bar.
A clothing size data identification device, comprising: a preprocessing module, an interception module, an extraction module, a judgment module, a splitting module, and an acquisition module;

The pre-processing module is configured to pre-process the target image to determine the overall clothing outline in the target image;

The intercepting module is configured to intercept the smallest screenshot of the overall outline of the clothing in the target image according to the overall outline of the clothing;

The extraction module is used to extract the style feature points in the minimum screenshot;

The judging module is configured to judge whether the style feature point in the minimum screenshot includes a detachable feature;

The splitting module is configured to perform splitting processing on the smallest screenshot according to the splittable feature when the splittable feature is included in the smallest screenshot to obtain a clothing splitting unit as an image of a region to be identified;

The acquisition module is configured to identify the image of the area to be identified, and acquire clothing size data in the image of the area to be identified.
A user terminal, comprising a memory and a processor, the memory is configured to store a clothing size data identification program, and the processor runs the clothing size data identification program to cause the user terminal to execute the method of claim 1 The method for identifying clothing size data according to any one of -6.
A computer-readable storage medium, characterized in that the computer-readable storage medium stores a clothing size data identification program, and when the clothing size data identification program is executed by a processor, it realizes any one of claims 1-6. Item of clothing size data identification method.