JP2019057255A - Customer ornament matching system - Google Patents

Abstract

To effectively present a matching result between a customer and an ornament and thereby allow the ornament to be recommended and selected more precisely.SOLUTION: A customer ornament matching system includes search means that refers to three-stages or more of association between a previously acquired reference image, and/or character string, and/or classification result related to a customer, and a reference image, and/or character string, and/or classification result related to an ornament, and searches for one or more customer ornament matching results on the basis of three-stages or more of association between the image, and/or character string, and/or classification result related to the customer acquired by voice input or manual input or automatic input and/or the reference image, and/or character string, and/or classification result related to the customer corresponding to the classification result, and the reference image, and/or character string, and/or classification result related to the ornament.SELECTED DRAWING: Figure 3

Description

The present invention relates to a customer clothing matching system and method and a customer classification program for solving the conventional problems in customer classification.

Nowadays, hundreds of thousands of clothing items are sold in department stores, specialty stores, and online shops. In addition, information related to fashion is also overflowing through magazines and blogs, etc. It is very difficult for a person to choose clothes suitable for him in a limited time.

Under such circumstances, consumers can tell their preferences etc. on the store clerk or online to assist in the selection. For example, in Patent Document 1, a consumer who wants to purchase a fashion product himself The technology related to the fashion product sales system that can receive the recommendation of the fashion product based on the desired conditions is disclosed, and in Patent Document 2, the consumer's social networking service and usage history are used to The technology related to the coordination support device that analyzes the fashion trend and presents the coordination based on the trend is disclosed. Also, in Patent Document 3, the clothing items challenged by the consumer are heard and reflected in the recommended content. Contents to be made are written.

However, none of the methods has solved the problem, and currently more than 80% of people may feel uncomfortable with choosing clothing items, and 70% of them often feel uncomfortable. Although more than 50% of people purchase clothing after an unpleasant experience, they often feel dissatisfied when they look at their closets and don't wear anything. As a result, 10% of the clothing they have It is reported that it is wearing a degree. This is largely due to consumers making choices depending on their own preferences, with or without the assistance, and preferences are formed within their very limited experience. It is explained that it is because of the instinct of human beings who are longing for a different type of person and trying to get closer to that type of person.

Currently, a method that can replace the selection based on this preference is a method of recommending clothing that looks scientifically, and there are theories represented by skeletal diagnostic theory and color diagnostic theory shown in Non-Patent Document 3. Image consultants provide diagnostics and accompanying services based on these theories

JP 2004-246585 A JP2015-72639A JP 2017-72907 A

Book "Magic rules for meeting clothes that look really good" by Yoko Futagami

Whether the consumer chooses clothing according to their preference as before, or when choosing clothing that looks scientifically as described above, it is possible to make an accurate recommendation by a retail store and an accurate selection by the consumer. The important thing to do is to accurately analyze and classify customers, and match customers and clothing, but at present, it has not been done accurately. For example, Patent Document 1 and Patent Document described above In the second example, various customer attributes are merely listed, and there is a problem that an effective matching cannot always be presented and the selection does not contribute to an accurate selection.

Accordingly, the present invention has been devised in view of the above-described problems, and the purpose of the present invention is to effectively present the customer clothing matching result so that the retail store can more accurately collect the clothing. The object is to provide a customer apparel matching system and method, and a customer classification program that can be recommended and can be accurately selected by the user.

The customer clothing matching system according to the present invention includes a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. Refer to the level of relevance
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or search means for searching for one or more customer clothing item matching results based on three or more levels of association with the character string and / or the classification result.

The customer clothing matching system according to the present invention includes a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. Refer to the level of relevance
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers Search means for searching for one or more customer clothing matching results based on three or more levels of association with images and / or character strings and / or classification results.

The customer clothing product matching system according to the present invention includes a plurality of reference images and / or character strings and / or character strings and / or character strings and / or character strings and / or character strings and / or a combination of classification results. Reference images relating to a plurality of customers and / or character strings and / or a plurality of customers according to the classification results, which are obtained by voice input, manual input or automatic input with reference to three or more levels of association with the classification results And / or one or more customer clothing matching results based on a reference image and / or a string and / or classification result for each piece of clothing for a combination between character strings and / or classification results And a search means for searching for.

The customer clothing matching system according to the present invention includes a reference image and / or a character string and / or a classification related to a clothing for a combination of a plurality of reference images and / or character strings and / or classification results regarding a customer acquired in advance. Refer to multiple images and / or character strings and / or multiple clothing items according to classification results by referring to three or more levels of relevance with the results and using speech input, manual input, or automatic input. One or more customer clothing matching results based on three or more levels of relevance with reference images and / or character strings and / or classification results for each customer for a combination between images and / or character strings and / or classification results And a search means for searching for.

The customer classification method according to the present invention includes three or more stages of a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. See the relevance of
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or a search step for searching for one or more customer clothing item matching results based on three or more levels of association with the character string and / or the classification result.

The customer classification method according to the present invention includes three or more stages of a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. See the relevance of
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers And a search step for searching for one or more customer clothing matching results based on three or more levels of association with images and / or character strings and / or classification results.

The customer classification method according to the present invention includes a plurality of reference images and / or character strings and / or reference images related to each piece of clothing and / or character strings and / or classification results for a combination of a plurality of reference images and customer results obtained in advance. And / or reference images related to a plurality of customers and / or character strings and / or classification results obtained by voice input, manual input or automatic input. Alternatively, search for one or more customer clothing matching results based on a reference image and / or a string and / or classification result for each piece of clothing for the combination between the character string and / or the classification result, and / or three or more levels of relevance. And a search step.

The customer classification method according to the present invention includes a reference image and / or a character string and / or a classification result for a clothing item for a combination of a plurality of reference images and / or character strings and / or a classification result regarding a customer acquired in advance. A plurality of images and / or reference images related to a plurality of clothing items according to character strings and / or classification results Search for one or more customer clothing matching results based on a reference image and / or a string and / or classification result for each customer for a combination between / and character strings and / or classification results And a search step.

The customer classification program according to the present invention includes three or more stages of a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. See the relevance of
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or having the computer execute a search step for searching for one or more customer clothing matching results based on three or more levels of association with the character string and / or the classification result.

The customer classification program according to the present invention includes three or more stages of a reference image and / or character string and / or classification result relating to a customer acquired in advance, and a reference image and / or character string and / or classification result relating to clothing. See the relevance of
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers And a search step for searching for one or more customer clothing matching results based on three or more levels of association with images and / or character strings and / or classification results.

The customer classification program according to the present invention includes a reference image and / or a character string and / or a classification result for each piece of clothing for a combination of a plurality of reference images and / or character strings and / or classification results regarding a customer acquired in advance. And / or reference images related to a plurality of customers and / or character strings and / or classification results obtained by voice input, manual input or automatic input. Alternatively, search for one or more customer clothing matching results based on a reference image and / or a string and / or classification result for each piece of clothing for the combination between the character string and / or the classification result, and / or three or more levels of relevance. And causing the computer to execute a search step.

The customer classification program according to the present invention includes a reference image and / or a character string and / or a classification result for a clothing item for a combination of a plurality of reference images and / or character strings and / or a classification result regarding a customer acquired in advance. A plurality of images and / or reference images related to a plurality of clothing items according to character strings and / or classification results Search for one or more customer clothing matching results based on a reference image and / or a string and / or classification result for each customer for a combination between / and character strings and / or classification results And causing the computer to execute a search step.

According to the present invention having the above-described configuration, it is possible to immediately select and output a customer clothing matching result corresponding to the image and / or character string obtained by analyzing the basic information. it can. As a result, it contributes to more accurate recommendation by clothing retailers and more accurate selection by consumers.

It is a figure which shows the structural example of the customer clothing article matching system to which this invention is applied. It is a figure which shows the block structure of an electronic device. It is a figure which shows the customer clothing matching business in which the customer clothing program to which this invention is applied is used. It is a flowchart of the processing operation of the customer clothing matching program to which the present invention is applied. It is a schematic diagram which shows the relevance of a customer clothing matching result. It is a schematic diagram which shows the relevance of the combination of a reference image and / or a character string, and / or a classification result, and a customer clothing matching result.

Hereinafter, a customer clothing matching system to which the present invention is applied will be described in detail with reference to the drawings.

A customer apparel matching system 1 to which the present invention is applied is embodied by, for example, a configuration as shown in FIG. The customer clothing matching system 1 includes a mobile terminal 11 for a user who uses the system, and a server 13 connected to the mobile terminal 11 via a public communication network 12. The user in the present embodiment is a consultant (also referred to as a stylist, an image consultant, etc.) and / or a customer and / or an apparel salesperson who has been requested for diagnosis by a customer.

The public communication network 12 is an Internet network or the like that connects the mobile terminals 11 and the server 13 via a communication line. Incidentally, when the portable terminal 11 is operated in a certain narrow area, the public communication network 12 may be configured by a LAN (Local Area Network). The public communication network 12 may be a so-called optical fiber communication network. The public communication network 12 is not limited to a wired communication network, and may be realized by a wireless communication network.

The mobile terminal 11 is, for example, a mobile phone, a smartphone, a tablet terminal, a wearable terminal, a notebook personal computer (PC), or the like, and is a device that can communicate through the public communication network 12 based on at least a user operation. is there. The portable terminal 11 is a device that can be carried by the user at all times, but is not limited to this, and is a concept that includes all electronic devices such as a desktop PC. .

The portable terminal 11 does not download a customer apparel matching program, which will be described later, via the public communication network 12 when installing package software sold at a store without downloading via the public communication network 12. A device that does not perform communication may be used. In the following example, a case where a smartphone is applied to the mobile terminal 11 will be described as an example.

FIG. 2 shows a specific configuration example of the mobile terminal 11. This mobile terminal 11 includes a ROM (Read Only Memory) 22, a RAM (Random Access Memory) 23 as a work area used for storing and developing data, and a CPU (Central Processing) for controlling the entire mobile terminal 11. (Unit) 24, an operation unit 25 for inputting various control commands via operation buttons, a keyboard, an output I / F 16 for controlling display of various information, and a hard disk A storage unit 27 for storing a program for performing a search to be performed, and a data input / output unit for inputting data into the portable terminal 11 from the outside or outputting data generated in the portable terminal 11 to the outside 29 are connected to the internal bus 21, respectively. Further, a communication I / F 28 and a voice input unit 31 are connected to the internal bus 21. The output I / F 16 is connected to a display unit 26 as a monitor that actually displays information.

The ROM 22 stores a program for controlling hardware resources of the entire mobile terminal 11. The RAM 23 temporarily stores various commands for controlling the hardware resources of the entire mobile terminal 11.

The CPU 24 is a so-called central processing unit for controlling each component mounted in the mobile terminal 11 by transmitting a control signal via the internal bus 21. Further, the CPU 24 transmits various control commands via the internal bus 21 in accordance with user operations via the operation unit 25.

The operation unit 25 is embodied by a touch panel or the like, and the user inputs information related to the customer or the clothing, and an execution command for executing the customer clothing matching program is input from the user. When the execution command is input by the user, the operation unit 25 notifies the CPU 24 of this. Upon receiving this notification, the CPU 24 reads the program from the storage unit 27 and executes it.

The output I / F 16 is configured by a graphic controller that creates a display image based on control by the CPU 24. The display unit 26 connected to the output I / F 16 is realized by, for example, a liquid crystal display (LCD).

When the storage unit 27 is composed of a hard disk, based on control by the CPU 24, predetermined information is written to each address and read out as necessary. The storage unit 27 stores a customer clothing matching program for executing the present invention. This program is read by the CPU 24 and executed.

The communication I / F 28 is mounted with a line control circuit for connecting to the public communication network 12, a signal conversion circuit for performing data communication with other terminal devices, and the like. The communication I / F 28 converts various commands from the internal bus 21 and sends them to the public communication network 12 side. When data is received from the public communication network 12, the communication I / F 28 performs predetermined conversion processing on the received data. To the internal bus 21 or the CPU 24.

The data input / output unit 29 is connected to a cord or the like for USB connection with an electronic device such as a PC. Data can be input / output to / from an external device via the data input / output unit 29.

The voice input unit 31 is a device for converting voice input from a user into electronic data.

A predetermined database is constructed in the server 13. Information sent via the public communication network 12 is stored in this database. Further, the server 13 transmits the accumulated information to the mobile terminal 11 through the public communication network 12 based on a request from the mobile terminal 11.

Note that any one or more components in the mobile terminal 11 or the server 13 may be controlled by artificial intelligence. The specific application method of artificial intelligence to the present invention may be based on any one or more of all known information related to artificial intelligence.

The customer clothing matching system 1 to which the present invention is applied is executed through a customer clothing matching program installed in the mobile terminal 11. That is, the user operates the mobile terminal 11 and matches the customer and the clothing through a customer clothing matching program installed on the mobile terminal 11. The customer clothing matching program is not limited to the case where it is installed and used on the mobile terminal 11, but is stored in the server 13 and accessed from the mobile terminal 11 each time the server 13 is used. Of course, you may make it do.

In the customer clothing matching, the information about the customer or the clothing is started from the place where the user himself / herself inputs or the information is automatically inputted through another program.

Examples of this information include customer information, images, videos, customer size, neck, chest thickness, wrist shape, clavicle, hip and other body features, skin, eyes, hair Color data such as, text data that expresses hobbies and preferences, and image and / or text data that expresses psychological features such as current psychological situation, future dreams, and important values, and currently owned The data shown below and the text data on the actual use of the clothing items, and the classification results of classifying the customers described later. In the case of apparel information, images, videos, text data representing apparel types, brands, sizes, prices, fabrics, colors, usage scenes, apparel design backgrounds, good points, etc. It is the classification result which classified the clothing items. The information input in this way is hereinafter referred to as basic information.

The classification results are already known, for example, the classification results of persons in skeletal diagnostic theory, the classification results of persons in color diagnosis theory, the classification results of persons recognized by the International Image Consultant Federation (AICI), the classification of fashion types, etc. It is a concept that includes all taxonomies. As an example of the customer classification result, for example, the classification result of a person in the skeletal diagnostic theory corresponds to straight, wave, and natural. For the classification of fashion type, an example is classic, modern Manish, sporty, avant-garde, ethnic, feminine, elegant, classic.

The user manually inputs the basic information via the operation unit 25 in the mobile terminal 11 or inputs the basic information by voice via the voice input unit 31. In this input, all or part of the basic information created in the electronic device such as another mobile terminal 11 or a PC may be input from the data input / output unit 29. In addition, all or part of data created by the user in another mobile terminal 11 or an electronic device such as a PC is transmitted via the public communication network 12 to the mobile terminal 11 in which the customer clothing matching program is actually installed. It may be. The basic information transmitted or input in this way is stored in the storage unit 27.

In particular, regarding the acquisition of images and / or text data that express the psychological features, unlike conventional methods that ask for preferences, etc., it is possible to grasp potential ideas and target figures that customers themselves do not recognize. Meaningful. As a method for obtaining the image and / or text data, a known psychological test may be used. It is desirable to select a test based on necessary information and use a plurality of test results in combination. For example, a psychological test that is meaningful for understanding the customer's potential thinking about the current customer situation, the future vision to be aimed at, and the values that we want to cherish from the present to the future, is a coloring book. Test, collage test, value card test.

Here is an example of test execution and data conversion of the results. Each test is performed by displaying the test on the operation unit 25, and it is desirable that the test is performed by the customer operating the operation unit 25 because of the nature of the test. In the case of a coloring book test, a design to be colored is displayed on the operation unit 25, and the user colors using a general coloring function. In the collage test, an image prepared in advance by the user and stored in the storage unit 27 or an image available on the Internet or the like is used, and functions such as general cutting and pasting are performed. The value card test is performed using a general image transfer function by presenting an image with values written in advance and stored in the storage unit 27. Each test result is selected for the value card with the color type and ratio as a character string in the coloring test, and the image and tagged keyword as a character string in the collage test. The values and their priorities are converted into data as character strings.

Concerning the use of the data that expresses the possessed clothing and its usage results, it is possible that the consumer himself / herself is not aware of the preferences, or unlike the conventional method of obtaining only purchase history data It is meaningful to understand the actual use of clothing items. Regarding how to obtain this data, we used the feeling that consumers would like to see when they came to wear clothes that looked good, and imitated how to post photos and articles to familiar social networks. An interface is provided, and tagging of belongings can be done manually or automatically using artificial intelligence, etc., and posting to social networks actually used by consumers is also possible. Compared with the method of accumulating the usage record of, the convenience is enhanced. Measures such as giving points according to the posting results are also effective.

An example of grasping the actual use of clothing items and creating the data of the results will be given. The data of the clothing items possessed are input manually or automatically in the same manner as the basic information and stored in the storage unit 27. The input of the clothing use result is performed by a method of displaying the interface on the operation unit 25. It is desirable from the viewpoint of improving the convenience of the consumer to store the past record of the use of the apparel item for each matching, instead of inputting it for each matching. In this case, the record is stored in the storage unit 27 or the server 13. Is done.

After the basic information is input in this way, the customer clothing matching program actually executes the processing operation. The processing operation flow of this customer clothing matching program is shown in FIG. The customer clothing matching program analyzes the basic information input in step S11. (Step S12). For this analysis, any existing text mining technology, data mining technology, language analysis processing technology, image analysis processing technology, or the like may be used.

Next, this customer apparel matching program uses one or more units of all grammatical structural units such as words, morphemes, phrases, clauses, etc. for text data among the basic information to be analyzed. , Character string extraction. For example, if there is text data that says "I am worried because my bones are thick and my bones are clumsy and I am always worried about it," I am trying to hide it. ”,“ Bone ”,“ lumpy ”,“ interesting ”,“ always ”,“ let's hide ”, and the like.

Alternatively, the basic information may be input as keywords or words in pieces. For example, when “thick chest plate” is input, the character strings “thick” and “chest plate” can be extracted. In addition to the customer's characteristics being input, the basic information may be input a clothing image that the customer wants to wear. For example, when “Clothes pursuing a free lifestyle” is input, by analyzing this by text mining, “Free”, “Free”, “Life”, “Style”, “Lifestyle” Like “Pursuit”, a character string is extracted from any one or more of grammatical structural units.

Next, the customer clothing matching program shifts to step S13, and starts searching for the character string, the face image group and the classification result extracted in step S12, and the highly correlated customer clothing matching result. Before performing this search, the mobile terminal 11 performs a reference image and / or character string and / or classification result related to the customer and a reference image and / or character string and / or classification related to the clothing as shown in FIG. The association of the result is acquired in advance.

5, a11, b11, c11,... Are reference images and / or character strings and / or classification results related to clothing. On the other hand, A1, A2, A3,... Are reference images and / or character strings and / or classification results related to customers.

The portable terminal 11 acquires the association indicating the degree to which such clothing information and customer information are related to each other. As an expression method of this association, for example, reference images and / or character strings and / or classification results and / or classification results and customer reference images and / or character strings and / Alternatively, the classification results are associated with each other, and when the classification results are not connected with arrows, they may be expressed in a form not associated with each other. For example, the reference image and / or character string and / or classification result A2 related to the customer are related to the reference image and / or character string and / or the classification result c11, a21 related to the clothing, but other customers You may make it express in the form which is not related with the classification result. In other words, in this case, whether the reference image and / or character string and / or classification result for clothing is related to the reference image and / or character string and / or classification result for the customer is expressed by a true or false binary value. It will be expressed digitally.

An analog expression method may be adopted as a method for expressing other associations. For example, the reference image and / or the character string and / or the classification result B3 related to the customer are related to the reference image and / or the character string and / or the classification result a21 and b21 related to the customer at a rate of 80%, and Reference image and / or character string and / or classification result c11 related to customer is 60%, reference image and / or character string and / or classification result b11 related to customer is 45%, reference image and / or character string related to customer And / or the classification result c21 may be expressed as an association degree such as 24%, the reference image and / or character string related to the customer and / or the classification result a11 may be expressed as 0%.

In addition, when the reference image and / or character string and / or classification result relating to the customer is “company officer”, the reference image and / or character string and / or classification result relating to the apparel is “formal”. The relevance of the customer classification result may be set as high as 90%.

Prior to the analysis in step S13, the mobile terminal 11 determines the association between the reference image and / or character string and / or classification result related to clothing and the reference image and / or character string and / or classification result related to the customer. It is stored in its own storage unit 27 and the like so that it can be read at any time. At this time, when the association between the reference image and / or character string and / or classification result for clothing and the reference image and / or character string and / or classification result for the customer is not stored in its own storage unit 27 May be stored in the server 13 and executed by accessing the server 13 when step S13 is actually executed.

After shifting to step S13, the customer clothing matching program obtains the customer clothing matching result 1 or 2 from the image and / or character string and / or classification result related to the customer or clothing extracted from the basic information analyzed in step S12. The selection operation is performed as described above.

Regarding the selection of the customer clothing matching result from the image and / or character string and / or classification result extracted from this basic information, the reference image and / or character string and / or You may make it refer to the correlation of a classification result, the image for a reference regarding a customer, and / or a character string, and / or a classification result. For example, when the character string analyzed from the basic information is “thick chest plate”, the image and / or the character string and / or the reference image and / or the character string and / or the character string and / or the character string and / or the classification result for the reference customer. Or, when referring to the association with the classification result, it is the “deep V-neck shirt” that is highly associated with the “thick chest plate”. Therefore, a “deep V-neck shirt” that is highly related to the analyzed character string “thick chest plate” is searched as a customer clothing matching result.

Also, if the character string analyzed from the basic information is “bank officer”, there is no identical reference character string, but if “company officer” exists as a reference character string, it will be referred to It is also possible to search for “formal” as the result of clothing matching and present it. In this way, even if the image and / or character string and / or classification result from the basic information and the reference image and / or character string and / or classification result do not completely match, the “officer” as in this example If there is a partial match, the customer clothing matching result that is highly related to the reference image and / or character string and / or classification result may be searched.

Furthermore, the character string analyzed from the basic information is the “Board of Directors”, and even if there is no reference character string that completely or partially matches this, There are similarities in semantic concepts. The similarity of the semantic concept may be set in advance on the system side, and a customer clothing matching result that is highly related to the reference image and / or the character string and / or the classification result having a high similarity may be searched. .

In other words, the image and / or character string and / or classification result analyzed from this basic information partially matches the reference image and / or character string and / or classification result with the image and / or character string and / or classification result. Or reference images and / or characters that have some correspondence to images and / or character strings and / or classification results from basic information as well as the degree of similarity of semantic concepts If the result is a column and / or classification result, a customer clothing matching result that is highly related to the reference image and / or character string and / or classification result may be searched.

The above-described work is performed not only on one character string analyzed from the basic information but also on a plurality of other images and / or character strings and / or classification results. As a result, not only one customer clothing matching result but also a plurality of customer clothing matching results may be selected.

The method of discriminating the customer clothing matching result that is more related to the analyzed image and / or character string and / or classification result is not limited to the method described above. For example, the degree of association may be set in advance between the reference image and / or character string and / or classification result related to clothing and the reference image and / or character string and / or classification result related to the customer. In such a case, 84% of the “deep V-neck shirt” and 25% of the “round neck shirt” with respect to the “thick chest plate” as the reference image and / or character string and / or classification result for the customer, “Turtleneck sweater” is 10%, “Formal shirt” is 77%, “Polo shirt” is 43%, etc. Reference images for clothing and / or character strings and / or classification results and reference images for customers An association degree is set between the character string and / or the classification result. When the reference character string “thick chest plate” is selected according to the image and / or character string analyzed from the basic information and / or the classification result, they are linked by the% of the above-mentioned relevance. You may make it judge. A customer classification result having an association degree equal to or higher than a predetermined threshold may be selected.

Note that the data on the association between the reference image and / or character string and / or classification result for the clothing item and the reference image and / or character string and / or classification result for the customer is downloaded on the mobile terminal 11 side. It may be stored in the server 13 at all times, and may be read from the server 13 or downloaded to the portable terminal 11 as necessary.

That is, this customer clothing matching program immediately selects a customer clothing matching result corresponding to the image and / or character string and / or classification result obtained by analyzing the basic information, and uses this. Can be output. The output of the customer clothing matching result may be displayed on the display unit 26 via the output I / F 16 or may be transmitted to the outside via the data input / output unit 29. The output of the customer clothing matching result may be stored in the server 13 from the public communication network 12 via the communication I / F 28, or may be transmitted to another portable terminal 11.

The user can recommend clothing to the consumer based on the customer clothing matching result output in this way, and the consumer can make a more accurate selection.

To make the recommendations more accurate, accurate classification of customers and apparel is effective. In the present embodiment, this customer and apparel classification is performed by experienced personnel, and furthermore, the customer and apparel category types must match or the category types do not match. The classification results are converted so that they can be easily matched by the correspondence table created by the same experienced personnel. Moreover, in deriving the customer clothing matching result, it is possible to place importance on the classification result among the images and / or character strings and / or classification results, and to reduce the importance of other images and character strings.

The classification of the customer and the clothing is preferably performed automatically using artificial intelligence or the like, not manually as described above.

Furthermore, according to the present invention, a search for specific clothing coordinates is performed based on the image and / or character string and / or classification result extracted from the basic information or the proposed customer clothing matching result. You may make it display. Or you may make it search for a specific coordination based on both the image extracted from basic information and / or a character string, and / or a classification result, and the proposed customer clothing matching result, and may display this. In any case, these operations are performed through artificial intelligence.

In such a case, refer to the result of matching the customer's clothing items, and select the clothing item group that has the same or similar classification of the clothing items (skirts, tops, etc.) of the corresponding clothing item group. It is. In performing this search, the portable terminal 11 acquires in advance the relevance between the reference image and / or the character string and / or the classification result and the coordination result regarding the clothing item. And a coordination result is searched using the method similar to the above-mentioned clothing goods classification. Artificial intelligence can also be used for this.

Next, another embodiment of the customer clothing matching system 1 to which the present invention is applied will be described.

In the example of FIG. 5, reference images and / or character strings and / or classification results relating to one or more customers are associated with reference images and / or character strings and / or classification results related to one piece of clothing. In the example of FIG. 6, reference images and / or characters for each customer for a combination between a plurality of clothing items and / or character strings and / or classification results are used. Refers to associations with columns and / or classification results.

In FIG. 6, as reference character strings, “V-neck shirt”, “straight cut jeans”, “skin shoes”, “skin heels”, “fine knit scarf”, “tweed jacket”, “ Round neck shirts ", etc., and customer classification results A1, A2,... Are assigned to these combinations. A solid line indicates the association, and a dotted line indicates the combination. For example, “V-neck shirt” and “straight-cut jeans” are connected to each other by dotted lines, and solid lines extend from there to A2 and C1. This indicates that customer classification results having a high association with the combination of “V-neck shirt” and “straight cut jeans” are A2 and C1. In other words, if the reference character string is only “V-neck shirt”, customers who match the V-neck shirt are matched, whereas “V-neck shirt” and “straight-cut jeans” are combined. , Customers that match this combination are matched.

FIG. 5 illustrates the association of the reference image and / or the character string and / or the classification result with respect to one or more customers with respect to the combination between the reference image and / or the character string and / or the classification result with respect to each clothing item. By acquiring in advance in the same manner as above, it is possible to refer to this when selecting the customer clothing matching result from the image and / or character string and / or classification result actually analyzed from the basic information. . In particular, it is possible to further narrow down the intended customer clothing matching result by combining them with each other than the reference image and / or the character string and / or the classification result alone, and to obtain a more accurate customer clothing matching result. It is also possible to select and propose effectively.

When referring to the relevance of one or more reference images and / or character strings and / or classification results to a combination between the reference images and / or character strings and / or classification results for each piece of clothing Extracts a plurality of images and / or character strings and / or classification results from the basic information. Then, the customer clothing items that are highly related to the combination of the plurality of reference images and / or character strings and / or classification results according to the extracted plurality of images and / or character strings and / or classification results. A matching result will be proposed. In such a case as well, as described above, the image and / or character string and / or classification result analyzed from the basic information may partially match the reference image and / or character string and / or classification result with the image and / or character. Reference images and / or character strings that have some correspondence to images and / or character strings and / or classification results from basic information, as well as whether or not they match completely, the degree of similarity of semantic concepts And if it is a classification result, you may make it search the customer's clothing matching result highly relevant with the reference image and / or character string and / or classification result.

At this time, the combination between the reference character strings may be any combination of noun-verb, noun-adjective, adjective-verb, and noun-noun. When extracting a character string from basic information, a plurality of character strings composed of one or more combinations of noun-verb, noun-adjective, adjective-verb, and noun-noun as words or phrases are extracted, The customer clothing matching result may be searched while referring to the same part of speech combination in the reference character string.

In addition, when the server 13 is controlled by artificial intelligence, the reference image and / or character string and / or classification result relating to clothing and the reference image and / or character string and / or classification result relating to the customer are associated. You may make it learn sex. That is, the relationship between the reference image and / or character string and / or classification result for clothing as shown in FIG. 5 and the reference image and / or character string and / or classification result for the customer, as shown in FIG. The relationship between the reference image and / or character string and / or classification result for various clothing items and the relationship between the reference image and / or character string and / or classification result for the customer is updated as needed by learning with artificial intelligence. Go. The artificial intelligence learns based on information provided from the outside to the server 13 via the public communication network 12. The information provided from the outside to the server 13 may be transmitted from, for example, the mobile terminal 11 using this system.

As a result of using the customer clothing matching program to which the present invention is applied by the mobile terminal 11, a reference image and / or character string and / or classification result related to the clothing selected by the user and a reference image related to the customer and / or Or the relationship between the character string and / or the classification result, or the reference image and / or the character string and / or the classification result and the reference image and / or the character string and / or the classification result for the customer not selected by the user The past history information such as the relationship is recorded on the mobile terminal 11 side. Then, the past history information is transmitted from the portable terminal 11 to the server 13. The artificial intelligence that controls the server 13 is configured to refer to the reference image and / or the character string and / or the classification result and the reference image and / or the character string related to the customer based on the past history information transmitted. Update the association with the classification result. That is, the relationship between the reference image and / or character string and / or classification result related to the clothing item selected by the user and the reference image and / or character string and / or classification result related to the customer is highly updated, Further, the relationship between the reference image and / or character string and / or classification result related to the clothing item not selected by the user and the reference image and / or character string and / or classification result related to the customer is updated with a lower relevance. For this update, a neural network or the like may be used.

By collecting such past history information from all the mobile terminals 11 using this system to the server 13, reference images related to clothing and / or character strings and / or classification results and reference images related to customers And / or associations with strings and / or classification results can be learned from time to time based on large amounts of data, providing customer clothing matching results that are easier for users to accept and provide the quickest solution. It becomes easy.

Learning of the association between the reference image and / or character string and / or classification result related to clothing and the reference image and / or character string and / or classification result related to the customer in the server 13 is implemented on the mobile terminal 11 side. Of course, you may make it make artificial intelligence perform.

Incidentally, external data for learning by this artificial intelligence includes not only the past history information of the mobile terminal 11 using this system described above, but also all Internet information that can be acquired via the public communication network, You may use big data.

For example, if two character strings “V-neck shirt” and “straight-cut jeans” are searched on the Internet under the AND condition, a home page appears. The image and / or character string information described in the homepage is acquired, image analysis is performed, and text mining is performed to find out what customer clothing matching result corresponds to using artificial intelligence.

In addition, according to the present invention, the correspondence relationship between the image and / or character string and / or classification result related to the customer or clothing analyzed from the basic information and the respective reference image and / or character string and / or classification result May be learned by artificial intelligence. This learning by artificial intelligence may be executed on either the mobile terminal 11 side or the server 13 side.

The correspondence between images and / or character strings and / or classification results relating to customers or clothing analyzed from the basic information and the respective reference images and / or character strings and / or classification results is the relationship between images and / or characters. The degree of similarity of semantic concepts may be learned by artificial intelligence from information acquired from the Internet or from a document read from the outside, in addition to whether or not they partially match or completely match. For example, if you have a high probability of filing both within a single homepage and within a single document, “Bone” and “Rug” are newly referred to as having similar semantic concepts. It may be registered as a word corresponding to the character string “bone thick”.

Further, according to the present invention, a process for prompting the user to input basic information may be introduced. The input of the basic information is promoted according to the image and / or character string extracted by mining the basic information and / or the classification result, or the search result of the customer clothing matching result in step S13. Good.

For example, if the character strings extracted from the basic information are only “bone” and “lumpy”, the customer clothing matching result may not be narrowed down unless further information is obtained and specified. In such a case, input of further basic information may be prompted.

Further, when “bone thick”, “clavicle”, “hand”, and “bone” can be extracted from the basic information, it can be read that the customer is thick and not only the clavicle but also the hand has a feeling of bone. In particular, when narrowing down a highly related customer classification result for a combination of a plurality of reference character strings as shown in FIG. In such a case, if an appropriate customer clothing matching result is not obtained as a result of mining the basic information, input of further basic information may be prompted in order to specify the result. The presence / absence of images and / or character strings corresponding to customer clothing matching results is pre-registered for each group such as “color”, “silhouette”, “style”, etc. In addition, it may be determined whether the basic information includes an image and / or character string that matches all or part of the registered character string. The classification result may be essential information.

Further, as a result of the customer clothing matching in step S13, if the customer does not select the clothing from the matching results, or if there is an objection to the matching result, input of more detailed information may be prompted. .

In the present invention, these determinations may be executed via artificial intelligence. This input promotion may be realized by displaying through the display unit 26 or may be performed by voice through a voice output unit (not shown).

DESCRIPTION OF SYMBOLS 1 Customer clothing matching system 11 Mobile terminal 12 Public communication network 13 Server 21 Internal bus 22 ROM
23 RAM
24 CPU
25 Operation unit 26 Display unit 27 Storage unit 28 Communication IF
29 Data input / output unit 31 Voice input unit

Claims (26)

Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or a search means for searching for one or more customer clothing matching results based on three or more levels of association with the character string and / or the classification result,
Customer clothing matching system. Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers Search means for searching for one or more customer clothing matching results based on three or more levels of association with images and / or character strings and / or classification results;
Customer clothing matching system. The relationship between the reference image and / or the character string and / or the classification result for each piece of clothing for the combination between the plurality of reference images and / or the character string and / or the classification result regarding the customer acquired in advance is three or more steps. Browse
For a combination of multiple images and / or character strings and / or classification images and / or reference images and / or character strings and / or classification results for a plurality of customers according to the results obtained by voice input or manual input or automatic input And a search means for searching for one or more customer clothing matching results based on three or more levels of relevance with reference images and / or character strings and / or classification results for each clothing item. Apparel matching system. Refer to three or more levels of association with reference images and / or character strings and / or classification results related to clothing for a combination of a plurality of reference images and / or character strings and / or classification results related to customers acquired in advance. And
Between a plurality of images and / or character strings and / or classification images and / or character strings and / or classification results according to a plurality of clothing and / or character strings and / or classification results obtained by voice input or manual input or automatic input Search means for searching for one or more customer clothing matching results based on three or more levels of relevance with reference images and / or character strings and / or classification results for each customer for the combination Customer clothing matching system. 5. The customer clothing matching system according to claim 1, wherein the mining means and the search means are provided in a portable information terminal. The customer clothing matching system according to any one of claims 1 to 4, wherein the mining means and / or the search means are controlled by artificial intelligence. A server capable of wireless communication with the portable information terminal via a communication network, wherein the server records the relevance to be referred to and transmits it to the portable information terminal; The customer clothing matching system according to claim 5. The server is controlled by artificial intelligence, and based on information provided from the outside, images and / or character strings and / or classification results for customers, images and / or character strings and / or classification results for clothing The customer clothing matching system according to claim 7, wherein learning is performed so that the relationship is reflected in the association degree to be referred to. The server is controlled by artificial intelligence, and the customer-related image and / or character string and / or classification result obtained by the voice input or manual input or automatic input, the reference image and / or character string and / or the customer 9. The customer clothing matching system according to claim 7 or 8, wherein a correspondence relationship with the classification result is learned. The server is controlled by artificial intelligence, and the image and / or character string and / or classification result related to the clothing acquired by the voice input or manual input or automatic input, the reference image and / or the character string related to the clothing, and 9. The customer clothing matching system according to claim 7 or 8, wherein a correspondence relationship with the classification result is learned. The search means is controlled by artificial intelligence, and the relationship between the reference image and / or character string and / or classification result related to the customer selected by the user and the reference image and / or character string and / or classification result related to the clothing item 7. The customer clothing matching system according to claim 1, wherein the past history information is provided from outside and learning is performed so that the past history information is reflected in the relevance to be referred to. The search means is controlled by artificial intelligence, and a plurality of images and / or character strings and / or classification results related to a customer acquired by voice input or manual input or automatic input, and a reference image and / or character string related to a customer and 7. The customer apparel matching system according to claim 1, wherein the correspondence relationship with the classification result is learned. The search means is controlled by artificial intelligence, and a plurality of images and / or character strings and / or classification results relating to clothing items acquired by voice input, manual input or automatic input, and reference images and / or characters relating to clothing items. The customer clothing matching system according to any one of claims 1 to 6, wherein a correspondence relationship with a column and / or a classification result is learned. The search means is controlled by artificial intelligence, and a plurality of images and / or character strings and / or classification results related to a customer acquired by voice input, manual input, or automatic input, and / or a reference image related to the searched clothing item 5. The customer clothing matching system according to claim 1, wherein coordinates of the clothing are searched based on a character string and / or a classification result and displayed. A plurality of images and / or character strings and / or classification results relating to customers acquired by voice input, manual input or automatic input, and / or reference images and / or character strings and / or classification results relating to the searched clothing items The customer clothing matching system according to any one of claims 1 to 14, further comprising an input promoting unit that prompts the user to input information related to the customer. 16. The customer clothing matching system according to claim 15, wherein the input promoting means is controlled by artificial intelligence. The customer according to any one of claims 1 to 16, wherein the search means uses a customer's desire at a deep psychological level and / or potential desire and / or values as customer information. Apparel matching system. The above search means has an interface imitating the article posting means to social network sites etc. in order to grasp the customer's use status of clothing items, in addition to tagging means for possession items and actually one or more social items The customer clothing matching system according to any one of claims 1 to 17, further comprising means for posting to a network. Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or a search step of searching for one or more customer clothing matching results by the customer clothing matching system based on three or more levels of association with the character string and / or the classification result,
A customer clothing matching method characterized by: Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers A search step of searching for one or more customer clothing matching results by a customer clothing matching system based on three or more levels of association with images and / or character strings and / or classification results;
A customer clothing matching method characterized by: The relationship between the reference image and / or the character string and / or the classification result for each piece of clothing for the combination between the plurality of reference images and / or the character string and / or the classification result regarding the customer acquired in advance is three or more steps. Browse
For a combination of multiple images and / or character strings and / or classification images and / or reference images and / or character strings and / or classification results for a plurality of customers according to the results obtained by voice input or manual input or automatic input A search step of searching for one or more customer clothing matching results by the customer clothing matching system based on three or more levels of relevance with reference images and / or character strings and / or classification results for each clothing about,
A customer clothing matching method characterized by: Refer to three or more levels of association with reference images and / or character strings and / or classification results related to clothing for a combination of a plurality of reference images and / or character strings and / or classification results related to customers acquired in advance. And
Between a plurality of images and / or character strings and / or classification images and / or character strings and / or classification results according to a plurality of clothing and / or character strings and / or classification results obtained by voice input or manual input or automatic input A search step of searching for one or more customer clothing matching results by a customer clothing matching system based on three or more levels of association with reference images and / or character strings and / or classification results for each customer for the combination; Having
A customer clothing matching method characterized by: Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Customer-related images and / or character strings and / or classification results obtained by voice input, manual input, or automatic input, and / or character strings and / or classification results, and reference images regarding clothing And / or causing the computer to execute a search step for searching for one or more customer clothing matching results based on three or more levels of association with the character string and / or the classification result;
A customer clothing matching program. Refer to a reference image and / or character string and / or classification result relating to a customer acquired in advance and a degree of association of three or more stages between a reference image and / or character string and / or classification result relating to clothing,
Reference images and / or character strings and / or classification results for clothing according to images and / or character strings and / or classification results obtained by voice input, manual input or automatic input, and reference for customers Causing a computer to execute a search step for searching for one or more customer clothing matching results based on three or more levels of association with images and / or character strings and / or classification results;
A customer clothing matching program. The relationship between the reference image and / or the character string and / or the classification result for each piece of clothing for the combination between the plurality of reference images and / or the character string and / or the classification result regarding the customer acquired in advance is three or more steps. Browse
For a combination of multiple images and / or character strings and / or classification images and / or reference images and / or character strings and / or classification results for a plurality of customers according to the results obtained by voice input or manual input or automatic input Causing the computer to execute a search step for searching for one or more customer clothing matching results based on three or more levels of relevance with reference images and / or character strings and / or classification results for each clothing item;
A customer clothing matching program. Refer to three or more levels of association with reference images and / or character strings and / or classification results related to clothing for a combination of a plurality of reference images and / or character strings and / or classification results related to customers acquired in advance. And
Between a plurality of images and / or character strings and / or classification images and / or character strings and / or classification results according to a plurality of clothing and / or character strings and / or classification results obtained by voice input or manual input or automatic input Causing the computer to execute a search step for searching for one or more customer clothing matching results based on three or more levels of association with reference images and / or character strings and / or classification results for each customer for the combination;
A customer clothing matching program.

Images