A personalized clothing shopping guide system based on data analysis

Abstract

The invention discloses a personalized clothing shopping guide system based on data analysis. The clothing shopping guide system comprises an image acquisition module, an image processing module, a user information input module, a model template database, a similarity processing module and a clothing matching push module. The image acquisition module is used to acquire a full-body image of a user. The image processing module is used to identify and extract information from the full-body image of the user. The user information input module is used for the user to input basic information. The model template database comprises model template basic information and clothing matching templates. The similarity processing module calculates the overall similarity between the user and the model template basic information according to the information extracted by the image processing module and the basic information input by the user information input module, and screens the model template basic information according to the overall similarity.

Description

Technical Field

[0001] The invention relates to the field of data analysis, and in particular to a personalized clothing shopping guide system based on data analysis. Background Art

[0002] In the traditional process of buying clothes, customers need to go to the clothing store in person, select the clothes they like, take off their clothes and try on the new clothes. This process takes up a lot of customers' time, and it often takes several hours to buy a piece of clothing that suits them. At present, with the development of Internet technology, more and more people choose to buy clothes online, but when buying clothes online, many people worry about what kind of clothes to wear, whether the clothes are suitable for them, and how to match the clothes. Summary of the invention

[0003] The purpose of the present invention is to provide a personalized clothing shopping guide system and method based on data analysis to solve the problems in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A personalized clothing shopping guide system based on data analysis, the clothing shopping guide system comprises an image acquisition module, an image processing module, a user information input module, a model template database, a similarity processing module and a clothing matching push module, the image acquisition module is used to acquire a full-body image of a user, the image processing module is used to identify and extract information from the full-body image of the user, the user information input module is used for the user to input basic information, the model template database comprises model template basic information and clothing matching templates, the similarity processing module calculates the overall similarity between the user and the model template basic information according to the information extracted by the image processing module and the basic information input by the user information input module, and filters the model template basic information according to the overall similarity, the clothing matching push module obtains the corresponding clothing matching template according to the filtered model template basic information, further filters the clothing matching template, and presents the filtered clothing matching template to the user.

[0006] More optimally, the image processing module includes a facial feature extraction module and a body part contour information extraction module, the facial feature extraction module is used to obtain the user's face shape, the body part contour information extraction module is used to obtain the user's shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference, the user information input module is for the user to input the user's gender, height information and weight, the similarity processing module includes a body similarity calculation module, a facial feature similarity calculation module, a body part contour similarity calculation module, an overall similarity calculation module and a similarity screening module, the body similarity calculation module calculates the body similarity according to the height information and weight, the facial feature similarity calculation module calculates the facial feature similarity according to the face shape, the body part contour similarity calculation module calculates the body part contour similarity according to the shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference, the overall similarity calculation module calculates the overall similarity according to the body similarity, the facial feature similarity and the body part contour similarity, and the similarity screening module compares the overall similarity with the similarity threshold to screen out the model template basic information whose overall similarity is greater than the similarity threshold.

[0007] More optimally, the clothing matching push module includes a clothing matching template acquisition module, a clothing matching template scoring module, a browsing timing module, a favorite value calculation module and a clothing matching template screening module. The clothing matching template acquisition module is used to obtain clothing matching templates corresponding to the basic information of the model template greater than the similarity threshold and present them to the user for viewing. The clothing matching template scoring module is used for the user to score the presented clothing matching templates. The browsing timing module times the time for the user to browse each set of clothing matching templates. The favorite value calculation module calculates the favorite value according to the score and browsing time of each set of clothing matching templates. The clothing matching template screening module screens out the top 5 sets of clothing matching templates with the favorite value scores from high to low, and presents these 5 sets of clothing matching templates to the user in order from high to low scores.

[0008] A personalized clothing shopping guide method based on data analysis, the clothing shopping guide method comprising the following steps:

[0009] Step S1: The camera shoots the user to obtain a full-body image of the user, performs facial recognition on the full-body image of the user and extracts facial feature information therefrom, and performs body part recognition on the full-body image of the user and extracts contour information of each body part therefrom;

[0010] Step S2: Input basic user information, including user gender, height information and weight. The height information includes user height, upper body length and lower body length.

[0011] Step S3: Filtering model template basic information consistent with the user's gender from a model template database pre-stored with a plurality of model template basic information, and calculating the overall similarity z1 between the user and the above model template basic information based on facial feature information, body part contour information and user basic information;

[0012] Step S4: Filter out the model template basic information whose overall similarity z1 is greater than the similarity threshold from the above model template basic information, obtain the clothing matching templates corresponding to these model template basic information and present them to the user for viewing, the user scores the presented clothing matching templates, and filter out the top 10 clothing matching templates with scores from high to low, obtain the browsing time of the user viewing these 10 clothing matching templates respectively, calculate the user's favorite value W for each clothing matching template, filter out the top 5 clothing matching templates with favorite value scores from high to low, and present these 5 clothing matching templates to the user in order from high to low scores. When the user views the favorite clothing matching template, he tends to spend more time viewing it, so the browsing time is used as a reference factor for the favorite value.

[0013] Preferably, the calculation of the overall similarity z1 between the user and the model template in step S3 includes the following steps:

[0014] Step S31: Calculate the body similarity x1=a1S0+a2S1+a3S3+a4M+a5Q,

[0015] Among them, S0 is the height similarity, S1 is the upper body length similarity, S3 is the user's lower body length similarity, M is the user's weight similarity, and Q is the similarity of the ratio of the user's upper body length to the lower body length; a1 is the weight of the height similarity, a2 is the weight of the user's upper body length similarity, a3 is the weight of the user's lower body length similarity, a4 is the weight of the user's weight similarity, and a5 is the weight of the similarity of the ratio of the user's upper body length to the lower body length; in many cases, the height is the same, but the ratio of the upper body length to the lower body length often has a large difference. In response to this situation, different clothing matching methods should be adopted. Therefore, the ratio of the user's upper body length to the lower body length is used as a reference factor for body similarity.

[0016] Step S32: Calculate the overall similarity z1=b1x1+b2x2+b3x3,

[0017] Among them, x1 is the body similarity, x2 is the facial feature similarity, x3 is the body part contour similarity, b1 is the weight of the body similarity, b2 is the weight of the facial feature similarity, and b3 is the weight of the body part contour similarity. Considering the overall similarity from three aspects, namely body similarity, facial feature similarity and body part contour similarity, makes it more targeted and reasonable to screen out the basic information of the model template.

[0018] More optimally, in step S4, the user's preference value for each clothing matching template is W=v1t / T, wherein v1 is the score of the clothing matching template, t is the browsing time of the user viewing the clothing matching template, and T is the total browsing time of the user viewing the 10 clothing matching templates.

[0019] More optimally, the facial feature information includes face shape, and the body part contour information includes shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference.

[0020] More optimally, the facial feature similarity x2 is the similarity of face shape, and the body part contour similarity x3=f1p1+f2p2+f3p3+f4p4, wherein p1 is the similarity of the ratio of shoulder width to hip circumference, p2 is the similarity of chest circumference, p3 is the similarity of waist circumference, p4 is the similarity of the ratio of thigh circumference to calf circumference, f1 is the weight of the similarity of the ratio of shoulder width to hip circumference, f2 is the weight of the similarity of chest circumference, f3 is the weight of the similarity of waist circumference, and f4 is the weight of the similarity of the ratio of thigh circumference to calf circumference.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention considers the overall similarity from three aspects: body shape similarity, facial feature similarity and body part contour similarity, thereby screening the basic information of the model template, so that the screened basic information of the model template is relatively similar to the user as a whole, and then the user's favorite clothing matching template is screened from the clothing matching template corresponding to the basic information of the model template according to the user's score and browsing time, thereby making the present invention more personalized, and does not require the user to try on and match clothes in person, saving a lot of user time. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a module schematic diagram of a personalized clothing shopping guide system based on data analysis of the present invention;

[0023] Figure 2 The figure is a flow chart of a personalized clothing shopping guide method based on data analysis of the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] See also Figure 1-2 In an embodiment of the present invention, a personalized clothing shopping guide system based on data analysis includes an image acquisition module, an image processing module, a user information input module, a model template database, a similarity processing module and a clothing matching push module. The image acquisition module is used to acquire a full-body image of a user, the image processing module is used to identify and extract information from the full-body image of the user, the user information input module allows the user to input basic information, the model template database includes model template basic information and clothing matching templates, the similarity processing module calculates the overall similarity between the user and the model template basic information based on the information extracted by the image processing module and the basic information input by the user information input module, and filters the model template basic information based on the overall similarity, the clothing matching push module obtains the corresponding clothing matching template based on the filtered model template basic information, further filters the clothing matching template, and presents the filtered clothing matching template to the user.

[0026] The image processing module includes a facial feature extraction module and a body part contour information extraction module, the facial feature extraction module is used to obtain the user's face shape, the body part contour information extraction module is used to obtain the user's shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference, the user information input module is for the user to input the user's gender, height information and weight, the similarity processing module includes a body similarity calculation module, a facial feature similarity calculation module, a body part contour similarity calculation module, an overall similarity calculation module and a similarity screening module, the body similarity calculation module calculates the body similarity according to the height information and weight, the facial feature similarity calculation module calculates the facial feature similarity according to the face shape, the body part contour similarity calculation module calculates the body part contour similarity according to the shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference, the overall similarity calculation module calculates the overall similarity according to the body similarity, the facial feature similarity and the body part contour similarity, and the similarity screening module compares the overall similarity with the similarity threshold to screen out the model template basic information whose overall similarity is greater than the similarity threshold.

[0027] The clothing matching push module includes a clothing matching template acquisition module, a clothing matching template scoring module, a browsing timing module, a favorite value calculation module and a clothing matching template screening module. The clothing matching template acquisition module is used to obtain clothing matching templates corresponding to model template basic information greater than a similarity threshold and present them to users for viewing. The clothing matching template scoring module is used for users to score the presented clothing matching templates. The browsing timing module times the time for users to browse each set of clothing matching templates. The favorite value calculation module calculates the favorite value according to the score and browsing time of each set of clothing matching templates. The clothing matching template screening module screens out the top 5 sets of clothing matching templates with the favorite value scores from high to low, and presents these 5 sets of clothing matching templates to users in order from high to low scores.

[0028] A personalized clothing shopping guide method based on data analysis, the clothing shopping guide method comprising the following steps:

[0029] Step S1: a camera shoots a user to obtain a full-body image of the user, performs facial recognition on the full-body image of the user and extracts facial feature information therefrom, performs body part recognition on the full-body image of the user and extracts contour information of each body part therefrom, wherein the facial feature information includes face shape, and the body part contour information includes shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference;

[0030] Step S2: Input basic user information, including user gender, height information and weight. The height information includes user height, upper body length and lower body length.

[0031] Step S3: Filter out basic model template information that is consistent with the user's gender from a model template database that pre-stores multiple basic model template information, and calculate the overall similarity z1 between the user and the basic model template information based on facial feature information, body part contour information, and user basic information:

[0032] Calculating the overall similarity z1 between the user and the model template includes the following steps:

[0033] Step S31: Calculate body similarity

[0034] x1=0.3S0+0.1S1+0.1S3+0.2M+0.3Q,

[0035] Among them, S0 is the height similarity, S1 is the upper body length similarity, S3 is the user's lower body length similarity, M is the user's weight similarity, and Q is the similarity of the ratio of the user's upper body length to the lower body length;

[0036] Step S32: Calculate the overall similarity z1=0.45x1+0.2x2+0.35x3,

[0037] Among them, x1 is the body shape similarity, x2 is the facial feature similarity, and x3 is the body part contour similarity.

[0038] The facial feature similarity x2 is the similarity of face shape, and the body part contour similarity x3=0.25p1+0.25p2+0.25p3+0.25p4, wherein p1 is the similarity of the ratio of shoulder width to hip circumference, p2 is the similarity of chest circumference, p3 is the similarity of waist circumference, and p4 is the similarity of the ratio of thigh circumference to calf circumference;

[0039] Step S4: Filter out the model template basic information whose overall similarity z1 is greater than the similarity threshold from the above model template basic information, obtain the clothing matching templates corresponding to these model template basic information and present them to the user for viewing, the user scores the presented clothing matching templates, and selects the top 10 clothing matching templates with scores from high to low, respectively obtains the browsing time of the user viewing these 10 clothing matching templates, and calculates the user's favorite value W for each clothing matching template.

[0040] The user's preference value for each clothing matching template is W=v1t / T, where v1 is the score of the clothing matching template, t is the browsing time of the user viewing the clothing matching template, and T is the total browsing time of the user viewing these 10 clothing matching templates.

[0041] The top five clothing matching templates with the most favorable scores from high to low are selected, and the five clothing matching templates are presented to the user in order of the scores from high to low.

[0042] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A personalized clothing shopping guide system based on data analysis, characterized by: The clothing shopping guide system includes an image acquisition module, an image processing module, a user information input module, a model template database, a similarity processing module and a clothing matching push module, wherein the image acquisition module is used to acquire a full-body image of a user, the image processing module is used to identify and extract information from the full-body image of a user, the user information input module allows the user to input basic information, the model template database includes model template basic information and clothing matching templates, the similarity processing module calculates the overall similarity between the user and the model template basic information based on the information extracted by the image processing module and the basic information input by the user information input module, and filters the model template basic information based on the overall similarity, the clothing matching push module obtains the corresponding clothing matching template based on the filtered model template basic information, further filters the clothing matching template, and presents the filtered clothing matching template to the user; The shopping guide method of the personalized clothing shopping guide system comprises the following steps: Step S1: The camera shoots the user to obtain a full-body image of the user, performs facial recognition on the full-body image of the user and extracts facial feature information therefrom, and performs body part recognition on the full-body image of the user and extracts contour information of each body part therefrom; Step S2: Input basic user information, including user gender, height information and weight. The height information includes user height, upper body length and lower body length. Step S3: Filtering model template basic information consistent with the user's gender from a model template database pre-stored with a plurality of model template basic information, and calculating the overall similarity z1 between the user and the above model template basic information based on facial feature information, body part contour information and user basic information; Step S4: Filter out the model template basic information whose overall similarity z1 is greater than the similarity threshold from the above model template basic information, obtain the clothing matching templates corresponding to these model template basic information and present them to the user for viewing, the user scores the presented clothing matching templates, and filter out the top 10 clothing matching templates with scores from high to low, respectively obtain the browsing time of the user viewing these 10 clothing matching templates, calculate the user's favorite value W for each clothing matching template, filter out the top 5 clothing matching templates with favorite value scores from high to low, and present these 5 clothing matching templates to the user in descending order of scores; The calculation of the overall similarity z1 between the user and the model template in step S3 includes the following steps: Step S31: Calculate the body similarity x1=a1S0+a2S1+a3S3+a4M+a5Q, Among them, S0 is the height similarity, S1 is the upper body length similarity, S3 is the lower body length similarity, M is the user weight similarity, Q is the similarity of the ratio of the upper body length to the lower body length of the user; a1 is the weight of the height similarity, a2 is the weight of the upper body length similarity, a3 is the weight of the lower body length similarity, a4 is the weight of the weight similarity, and a5 is the weight of the ratio of the upper body length to the lower body length of the user; Step S32: Calculate the overall similarity z1=b1x1+b2x2+b3x3, Among them, x1 is the body similarity, x2 is the facial feature similarity, x3 is the body part contour similarity, b1 is the weight of body similarity, b2 is the weight of facial feature similarity, and b3 is the weight of body part contour similarity; In step S4, the user's preference value for each clothing matching template is W=v1t / T, where v1 is the score of the clothing matching template, t is the browsing time of the user viewing the clothing matching template, and T is the total browsing time of the user viewing the 10 clothing matching templates; The facial feature information includes face shape, and the body part contour information includes shoulder width, hip circumference, chest circumference, waist circumference, thigh circumference and calf circumference; The facial feature similarity x2 is the similarity of face shape, and the body part contour similarity x3=f1p1+f2p2+f3p3+f4p4, wherein p1 is the similarity of the ratio of shoulder width to hip circumference, p2 is the similarity of chest circumference, p3 is the similarity of waist circumference, and p4 is the similarity of the ratio of thigh circumference to calf circumference. f1 is the weight of the similarity of the ratio of shoulder width to hip circumference, f2 is the weight of the similarity of chest circumference, f3 is the weight of the similarity of waist circumference, and f4 is the weight of the similarity of the ratio of thigh circumference to calf circumference. The clothing matching push module includes a clothing matching template acquisition module, a clothing matching template scoring module, a browsing timing module, a favorite value calculation module and a clothing matching template screening module. The clothing matching template acquisition module is used to obtain clothing matching templates corresponding to model template basic information greater than a similarity threshold and present them to users for viewing. The clothing matching template scoring module is used for users to score the presented clothing matching templates. The browsing timing module times the time for users to browse each set of clothing matching templates. The favorite value calculation module calculates the favorite value according to the score and browsing time of each set of clothing matching templates. The clothing matching template screening module screens out the top 5 sets of clothing matching templates with the favorite value scores from high to low, and presents these 5 sets of clothing matching templates to users in order from high to low scores.

Images