KR20230104492A - Method, Apparatus and System for Converting Text in Image Adaptively - Google Patents

Abstract

An adaptive image-to-text conversion method, apparatus and system are disclosed. The adaptive image text conversion method displays an editing screen of a web page including an image based on a web-based web page editor, extracts an image editing file corresponding to the image from a database, and edits the extracted image editing file. Display on one side of the screen, extract text expressed in the second language corresponding to text expressed in the first language from the database, analyze the image to generate second text attribute information according to the analysis result, and Text expressed in two languages may be converted based on one of first text property information and second text property information. Therefore, the text of the image included in the web page can be easily converted into a web-based text without cumbersome process.

Description

Method, Apparatus and System for Converting Text in Image Adaptively}

The present invention relates to an adaptive image-to-text conversion method, apparatus, and system, and more particularly, to an adaptive image-to-text conversion method capable of easily converting text of an image included in a web page on a web basis without cumbersome processes; It is about devices and systems.

In general, in order to register a product in an Internet shopping mall, a considerable number of processes are required. After taking a picture of the product, upload it, design the detailed screen, enter the title, product introduction and information, and select the representative image of the product to be printed on the product list.

In particular, the task of introducing a product and inputting information when registering a product is a very important part when building an online shopping mall. Therefore, inputting product-related information on a product sales web page requires a person familiar with the product to input it individually, which can be a very cumbersome task. As the number of types of products increases, such troubles become more severe and excessive human and time resources may be consumed.

Therefore, there is an urgent need for a technology to prevent consumption of human and time resources by automating product-related information on a product sales web page of a shopping mall and accurately and quickly input product information.

In addition, since it is possible to promote business expansion through global sales due to the nature of online shopping, it is necessary to translate product sales web pages into multiple languages smartly for global sales of products.

However, when text is included in an image, real-time translation through a general web text translator is almost impossible. Therefore, conventionally, in order to translate text included in an image, an editor manually extracts text, converts a file, uploads the converted image, and the like. Therefore, in order to translate the text included in the image, a considerable amount of time is consumed and an editor's work is put in. In particular, excessive time and human resources are consumed in case of mass conversion.

Registered Patent No. 10-1666611

An object of the present invention is to provide an adaptive image-to-text conversion method, apparatus, and system capable of easily converting the text of an image included in a web page on a web basis without a cumbersome process.

In order to achieve this object, the present invention provides an adaptive image-to-text conversion method in one aspect. The adaptive image-to-text conversion method is performed by a conversion device operating on the basis of, and displaying an editing screen of a web page including an image based on a web-based web page editor - the image is a first image text including; extracting an image editing file corresponding to the image from a database, wherein the image editing file includes text expressed in a first language corresponding to the first image text and first text attribute information; displaying the extracted image editing file on one side of the editing screen; extracting text expressed in a second language corresponding to the text expressed in the first language from the database; analyzing the image and generating second text attribute information according to an analysis result; and converting the text expressed in the second language based on any one of the first text property information and the second text property information.

The adaptive image text conversion method may include inserting the converted text expressed in the second language into the image editing file; Generating a converted image based on an image editing file into which text expressed in the second language is inserted, wherein the converted image includes second image text corresponding to text expressed in the second language- ; and updating the created image on the web.

The generating of the second text attribute information may include setting a related text attribute value based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image. can

Setting the associated text property value may include extracting emotion information corresponding to the image; and setting a font value associated with the extracted emotion information.

Setting the associated text property value may include extracting complexity information corresponding to the image; and setting at least one of a text weight value and a font value associated with the extracted complexity information.

Setting the associated text property value may include extracting brightness information corresponding to the image; and setting a text color value associated with the extracted brightness information.

The setting of the associated text property value may include extracting complementary color information corresponding to the image; and setting an emphasis color value associated with the extracted complementary color information.

The converting may include: applying first text property information and second text property information to the text expressed in the second language, comparing them, and displaying a selectable user interface on the editing screen; and converting text expressed in the second language according to text property information selected based on the user interface.

On the other hand, in order to achieve the above object of the present invention, the present invention provides an adaptive image-to-text conversion device in another aspect. The image text conversion device displays an editing screen of a web page including an image based on a web-based web page editor - the image includes a first image text - and an image editing file corresponding to the image. Extract from the database—the image editing file includes text expressed in a first language corresponding to the first image text and first text property information—and display the extracted image editing file on one side of the editing screen Web page editor to display; and extracting text expressed in a second language corresponding to the text expressed in the first language from the database, analyzing the image, generating second text property information according to an analysis result, and expressing the text in the second language. and a conversion unit for converting the converted text based on any one of the first text property information and the second text property information.

The conversion unit inserts the converted text expressed in the second language into the image editing file, and generates a converted image based on the image editing file into which the text expressed in the second language is inserted - the conversion The generated image includes second image text corresponding to the text expressed in the second language, and the created image can be updated on the web.

The conversion unit may set a related text property value based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image. The conversion unit may extract emotion information corresponding to the image and set a font value associated with the extracted emotion information.

The conversion unit may extract complexity information corresponding to the image and set at least one of a text thickness value and a font value associated with the extracted complexity information. The conversion unit may extract brightness information corresponding to the image and set a text color value associated with the extracted brightness information. The conversion unit may extract complementary color information corresponding to the image and set an emphasis color value associated with the extracted complementary color information.

The converting unit applies first text property information and second text property information to the text expressed in the second language, compares them, and displays a user interface for selection, and the text property selected based on the user interface. Text expressed in the second language may be converted according to the information.

On the other hand, in order to achieve the above object of the present invention, the present invention provides an adaptive image data conversion system in another aspect. The adaptive image data conversion system may include a database; And based on a web-based web page editor, displaying an editing screen of a web page including an image-the image includes a first image text-, extracting an image editing file corresponding to the image from the database- The image editing file includes text expressed in a first language corresponding to the first image text and first text property information-, displays the extracted image editing file on one side of the editing screen, and Text expressed in a second language corresponding to text expressed in the first language is extracted from the database, the image is analyzed, and second text attribute information is generated according to an analysis result, and the text expressed in the second language , Image text conversion unit for converting based on any one of the first text property information and the second text property information.

The image text conversion unit inserts the converted text expressed in the second language into the image editing file, and generates a converted image based on the image editing file into which the text expressed in the second language is inserted- The converted image may include second image text corresponding to the text expressed in the second language, and the generated image may be updated on the web.

The image-to-text conversion unit may set a related text attribute value based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image.

The image text conversion unit may extract emotion information corresponding to the image and set a font value associated with the extracted emotion information. The image-to-text conversion unit may extract complexity information corresponding to the image and set at least one of a text thickness value and a font value associated with the extracted complexity information. The image text conversion unit may extract brightness information corresponding to the image and set a text color value associated with the extracted brightness information. The image text conversion unit may extract complementary color information corresponding to the image and set an emphasis color value associated with the extracted complementary color information.

The image-to-text conversion unit applies first text property information and the second text property information to the text expressed in the second language, respectively, compares them, displays a user interface capable of selection, and is selected based on the user interface. Text expressed in the second language may be converted according to text property information.

On the other hand, in order to achieve the object of the present invention described above, in another aspect, the present invention is a step of displaying an editing screen of a web page including an image based on a web-based web page editor on a computer - the image is the first 1 contains image text-; extracting an image editing file corresponding to the image from a database, wherein the image editing file includes text expressed in a first language corresponding to the first image text and first text attribute information; displaying the extracted image editing file on one side of the editing screen; extracting text expressed in a second language corresponding to the text expressed in the first language from the database; analyzing the image and generating second text attribute information according to an analysis result; A program stored in a medium is provided to execute the step of converting the text expressed in the second language based on any one of the first text property information and the second text property information.

According to the present invention, the text of an image included in a web page can be easily converted on a web basis in terms of translation and adaptive properties without a cumbersome process.

1 is a block diagram showing the configuration of a system for realizing a product information generating method according to a preferred embodiment of the present invention.
FIG. 2 is a flowchart for explaining the operation flow of the server 100 shown in FIG. 1, and shows the flow of a product information generating method according to a preferred embodiment of the present invention.
3 is an exemplary diagram showing a category hierarchical organization that exemplarily represents category keywords.
4 is an exemplary diagram exemplarily illustrating a state of attribute hierarchical organization that exemplarily represents attribute keywords.
Fig. 5 is a block diagram for explaining the configuration of a system for realizing a method of displaying product information in multiple languages.
FIG. 6 is an operation of the server shown in FIG. 5 and a flow for explaining the flow of a method of displaying product information in multiple languages.
7 is an exemplary diagram illustrating a keyword set managed in a database by way of example.
8 is an exemplary view illustrating a sentence template set managed in a database by way of example.
9 is an exemplary view illustrating that sentences about a product are generated in multiple languages based on a keyword set and a sentence template set.
10 shows the configuration of a system for realizing web-based image-text translation according to a preferred embodiment of the present invention.
FIG. 11 is a flowchart for explaining the detailed operation flow of the server shown in FIG. 10, and shows a web-based image-text translation process.
12 is an exemplary diagram illustrating an image text translation request icon by way of example.
13 is a flowchart illustrating an image text translation process performed by an image text translation unit.
Fig. 14 is a block diagram showing the configuration of a system for realizing a web page translation method according to a preferred embodiment of the present invention.
FIG. 15 is a flowchart for explaining the operation flow of the server shown in FIG. 14 and shows the flow of a web page translation method.
16 is an exemplary view illustrating a web page translation request icon by way of example.
17 is a block diagram showing the configuration of a system for performing image-to-text conversion according to a preferred embodiment of the present invention.
FIG. 18 is a flowchart for explaining the operation flow of the image-to-text conversion system shown in FIG. 17 .
19 is a flowchart for explaining the detailed operation flow of step S63 performed by the server.
Fig. 20 is a block diagram showing the configuration of a system for realizing an image conversion method according to a preferred embodiment of the present invention.
FIG. 21 is a flowchart for explaining an operation flow of the image conversion system shown in FIG. 20 .
22 is a flowchart for explaining the detailed operation flow of step S83 performed by the server.
23 illustratively shows an original image in which a Korean sentence is inserted.
FIG. 24 illustratively shows a converted image obtained by converting Korean sentences into English from the image shown in FIG. 23 by the image generator.
25 is a block diagram showing the configuration of a system for performing adaptive image-to-text conversion according to a preferred embodiment of the present invention.
FIG. 26 is a flowchart for explaining an operation flow of the adaptive image-to-text conversion system shown in FIG. 25 .
27 is a flowchart for explaining the detailed operation flow of step S103 performed by the server.
28 is a block diagram showing the configuration of a system for realizing an adaptive image conversion method according to a preferred embodiment of the present invention.
FIG. 29 is a flowchart for explaining an operation flow of the adaptive image conversion system shown in FIG. 28 .
30 is a flowchart for explaining the detailed operation flow of step S123 performed by the server.
31 exemplarily shows an original image in which a Korean sentence is inserted.
FIG. 32 illustratively shows a converted image obtained by adaptively converting Korean sentences into English from the image shown in FIG. 31 by the image generator.

Since the present invention can make various changes and have various embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the present invention to specific embodiments, and should be construed as including all modifications, equivalents, or substitutes included in the spirit and technical scope of the present invention.

Terms such as first and second may be used to describe various components, but the components should not be limited by the terms. These terms are only used for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The terms and/or include any combination of a plurality of related recited items or any of a plurality of related recited items.

It is understood that when a component is referred to as being 'connected' or 'connected' to another component, it may be directly connected or connected to the other component, but other components may exist in the middle. It should be. On the other hand, when a component is referred to as being 'directly connected' or 'directly connected' to another component, it should be understood that no other component exists in the middle.

Terms used in this application are only used to describe specific embodiments, and are not intended to limit the present invention. Expressions in the singular number include plural expressions unless the context clearly indicates otherwise. In this application, terms such as 'include' or 'have' are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in the present application, they should not be interpreted in an ideal or excessively formal meaning. don't

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present invention will be described in more detail. In order to facilitate overall understanding in the description of the present invention, the same reference numerals are used for the same components in the drawings, and redundant descriptions of the same components are omitted.

1 is a block diagram showing the configuration of a system for realizing a product information generation method according to a preferred embodiment of the present invention.

As shown in FIG. 1 , the server 100 may interwork with the database 20 and the shopping mall 10 through a communication network. The server 100 extracts a plurality of keywords and sentence templates related to products sold in the shopping mall 10, and dynamically creates various types of product description sentences based on the extracted keywords, thereby providing a website for selling products in the online shopping mall 10. Product-related information can be easily displayed on the page. Such a server may include a keyword extractor 110, a sentence template extractor 120, a sentence generator 130, an uploader 140, and the like. Each unit (1120 ~ 140) can interlock with each other.

FIG. 2 is a flowchart for explaining the operation flow of the server 100 shown in FIG. 1, and shows the flow of a product information generating method according to a preferred embodiment of the present invention.

Referring to FIGS. 1 and 2 , first, the keyword extraction unit 110 of the server 100 extracts a plurality of keywords corresponding to a plurality of keyword items corresponding to the product based on analyzing the image of the product. It can be extracted (step: S1).

The plurality of keyword items may include a category keyword indicating a category to which the product belongs, an attribute keyword indicating a characteristic attribute of the product, and a mood keyword indicating an emotional attribute of the product.

The category keyword is a keyword representing a category in which a product is included. 3 is an exemplary diagram showing a category hierarchical organization that exemplarily represents category keywords. As shown in FIG. 3 , for example, the category keyword condensedly indicates which category a product belongs to, such as a knit top, a skirt, a blouse, and the like. In the present invention, women's clothing, for example, is classified into a total of 132 keywords through such category hierarchical systematization.

The attribute keyword is a keyword representing a characteristic attribute of a corresponding product. 4 is an exemplary diagram exemplarily illustrating a state of attribute hierarchical organization that exemplarily represents attribute keywords. As shown in FIG. 4, the attribute keyword represents the characteristics of the product, such as logo, rippon, buckle decoration, partial lace, and solid bristle. In the present invention, through such attribute hierarchical systematization, for example, fashion-related attributes are classified into a total of 494 keywords. The emotional keyword is an identifier representing a characteristic emotional part of a corresponding product, such as moody, sad, pleasant, and the like.

Next, the sentence template extraction unit 120 of the server 100 may extract at least one sentence template corresponding to the plurality of keywords from the linked database 20 (step: S2).

The at least one sentence template may include a basic sentence template corresponding to the plurality of keywords and at least one popular sentence template. The popular sentence template may be extracted from big data and stored in a database in association with the basic sentence template.

The popular sentence template is a sentence containing the plurality of keywords, based on at least one of the most frequently used sentence, the sentence used to describe the product with the most sales, and the sentence used to describe the product with the highest number of hits It can be a generated template.

The sentence template extractor 120 may insert the plurality of keywords into the basic sentence template and at least one popular sentence template and display a user interface requesting selection of one of the keywords on the user terminal.

Subsequently, when at least one sentence template is determined, the sentence generation unit 130 of the server 100 inserts the plurality of keywords into the at least one sentence template to generate at least one product description sentence corresponding to the product. It can be created (step: S3). Then, the upload unit 140 of the server 100 may upload the at least one product description sentence to the product sales web page (step: S4).

In this way, product information may be generated by extracting keywords and sentence templates corresponding to the product through image analysis of the product.

In this way, in the generation of sentences related to products according to the present invention, the optimized expressions are made into sentence templates in advance, and the server 100 extracts a plurality of keywords and sentence templates related to products sold in the shopping mall 10. Information related to the product can be easily displayed on the web page for selling the product of the online shopping mall 10 by dynamically generating various types of product description sentences based on the product description.

On the other hand, the generation of sentences related to such products may be associated with multiple languages. That is, by pre-associating an expression optimized for a word sentence with a database and applying it in real time according to word or sentence replacement, the accuracy of multilingual expression is increased and information is displayed quickly. This is performed based on sentence templates, and is a completely different concept from real-time translation.

5 is a block diagram for explaining the configuration of a system for realizing a method of displaying product information in multiple languages, and FIG. 6 is an operation of the server shown in FIG. 5 to explain the flow of the method for displaying product information in multiple languages. to flow.

Referring to FIGS. 5 and 6 , first, the keyword set extractor 210 of the server 200 extracts a plurality of keywords corresponding to a plurality of keyword items corresponding to the product based on analyzing the image of the product. A set can be extracted (step: S11). The plurality of keyword sets may include a first keyword expressed in the first language and a second keyword expressed in a second language different from the first language. That is, the keyword set may mean a set of multilingual expressions for a specific word.

7 is an exemplary view showing keyword sets managed in the database 22. As shown in FIG. 7, it can be seen that a plurality of languages having the same meaning are managed as a set to correspond to a specific attribute keyword. .

For example, extracting a plurality of keyword sets of the product (step: S11) extracts a plurality of first keywords corresponding to a plurality of keyword items corresponding to the product based on a first language, , a plurality of second keywords associated with the plurality of extracted first keywords may be extracted from the database 22 .

As mentioned above, the plurality of keyword items may include a category keyword indicating a category to which the product belongs, an attribute keyword indicating a characteristic attribute of the product, and a mood keyword indicating an emotional attribute of the product.

Meanwhile, the sentence template set extractor 220 of the server 200 may extract at least one sentence template set corresponding to the plurality of keyword sets from the database 22 (step: S12). Here, the at least one sentence template set may include a first sentence template expressed in the first language and a second sentence template expressed in the second language. That is, the sentence template set may mean a set of multilingual expressions for a specific sentence template.

8 is an exemplary view showing a sentence template set managed in the database 22. As shown in FIG. 8, it can be seen that a plurality of languages having the same meaning are managed as a set to correspond to a specific sentence. . Each sentence may be an expression optimized for locals who speak the corresponding language.

Next, the sentence generation unit 230 of the server 200 selects keywords and sentence templates corresponding to the language selected by the language selection signal from the keyword set and the sentence template set based on the language selection signal. It can (step: S13).

The language selection signal may be a signal for selecting a language desired by the user. For example, the server 200 may display a user interface including a language selection unit through which a user may select a desired language, and the language selection unit may generate a language selection signal corresponding to the language selected by the user. .

The sentence generating unit 230 inserts the selected keyword into the selected sentence template to generate a product description sentence of the product expressed in the selected language (step: S14), and converts the generated product description sentence to the product. can be uploaded to the sales webpage of (step: S15).

9 is an exemplary view showing that sentences about products are generated in multiple languages based on keyword sets and sentence template sets. As shown in FIG. 9, sentences about products based on keyword sets and sentence template sets Sentences can be created and managed in multiple languages. It can be seen that a plurality of sentence sets having the same meaning are managed to correspond to a specific sentence. Each sentence may be an expression optimized for locals who speak the corresponding language by template.

Meanwhile, the at least one sentence template set may include a basic sentence template set corresponding to the plurality of keyword sets and at least one popular sentence template set. The popular sentence template set includes sentence templates expressed in multiple languages, and the sentence templates expressed in each language are used for the sentence most frequently used in the region where the language is used and for the description of the product with the highest sales volume in the region. The template may be generated based on a locally used expression using the language, based on at least one of a sentence used to describe a product having the highest number of hits in the region.

Hereinafter, embodiments related to web-based translation in which text included in an image can be easily translated on a web-based basis without a cumbersome process will be described.

10 shows the configuration of a system for realizing web-based image-text translation according to a preferred embodiment of the present invention.

As shown in FIG. 10 , the server 300 is linked with the shopping mall 14 and the database 24 and may include a web page editor 310 and an image text translation unit 320 . The server 300 can easily translate image text included in images included in various web pages for sale in the shopping mall 14 .

FIG. 11 is a flowchart for explaining the detailed operation flow of the server shown in FIG. 10, and shows a web-based image-text translation process.

Referring to FIGS. 10 to 11 , the server 300 may display an editing screen of a web page including an image based on the web page editor 310 operating in real time on the web (step: S31). The image may include image text expressed in the first language. Next, an image text translation request icon for requesting translation of the image text included in the image may be displayed (step: S32).

When the image text translation request icon is selected, the image text translation unit 320 may modify the image text included in the image into a second language, which is a target language, and update the modified image on the web in real time (step: S33)

12 is an exemplary diagram illustrating an image text translation request icon by way of example. Referring to FIG. 12 , an image may be displayed in an editing screen of a web page, and an image text translation request icon lC1 may be displayed in the image, for example. Preferably, the image text translation request icon is provided near or within the image so as to intuitively recognize that the image is an icon for requesting translation of the corresponding image.

13 is a flowchart illustrating an image text translation process performed by an image text translation unit.

As shown in FIG. 13 , the image text translation unit 320 may extract an image editing file corresponding to the image from a database in response to selecting the image text translation request icon (step: S41 ).

Subsequently, the image text translation unit 320 displays the extracted image editing file using a web-based image editor (step: S42), and on the other hand, may select a second language that is a target language to be translated (step: S43).

The image text translation unit 320 may extract text expressed in the second language corresponding to the text expressed in the first language from the database (step: S44). Translating the text expressed in the first language into the text expressed in the second language may be performed by the multilingual translation process described above with reference to FIGS. 5 to 9 .

The image text translation unit 320 modifies the image editing file based on inserting the extracted text expressed in the second language into the image editing file (step: S45), and converts the modified image editing file to Based on this, it is possible to generate an image in a form inserted into a web page (step; S46). And, the converted image can be inserted into the web page.

Meanwhile, the image text translation unit 320 may store the modified image editing file in the database. In addition, the image text translation unit 320 extracts a plurality of texts expressed in the second language corresponding to the text expressed in the first language from the database, and displays the extracted texts expressed in the plurality of second languages. and display a user interface for selecting one of the displayed texts. Text displayed in a second language selected through the user interface may be inserted into the image file.

Also, the image text translation unit 320 may re-edit the extracted text expressed in the second language so as to be suitable for the insertion position of the image editing file.

Fig. 14 is a block diagram showing the configuration of a system for realizing a web page translation method according to a preferred embodiment of the present invention.

As shown in FIG. 14, the server 400 may interwork with the shopping mall 16 and the database 26 through a communication network, and may include a web page editor 410 and a web page translator 420. The server 400 can easily translate image text included in images included in various web pages for sale in the shopping mall 16 .

FIG. 15 is a flowchart for explaining the operation flow of the server shown in FIG. 14 and shows the flow of a web page translation method.

Referring to FIGS. 15 to 14 , the web page editor 410 may display a web page translation request icon on the web page editing screen of the web-based web page editor (step: S51). The web page may include first text and images expressed in a first language, and the image may include second text expressed in the first language.

16 is an exemplary view illustrating a web page translation request icon by way of example. Referring to FIG. 16 , an image is displayed in a web page editing screen, and an image web page translation request icon 1C2 may be displayed on one side of the screen.

When the web page translation unit 420 selects the image text translation request icon, the first text included in the web page and the second text included in the image are translated into a second language, which is the target language, in conjunction with a database. By modifying, the web page can be updated on the web (step: S53).

In response to selecting the web page translation request icon, the web page translation unit 420 extracts an image editing file corresponding to the image from a database and displays the extracted image editing file using a web-based image editor. select a second language, which is a target language to be translated, and extract first text and second text expressed in the second language corresponding to the first text and the second text expressed in the first language from the database; , Based on inserting the extracted second text expressed in the second language into the image editing file, modifying the image editing file, and converting the modified image editing file into an image in a format inserted into a web page and insert the converted image into the web page.

The web page translation unit may store the modified image editing file in the database. , The web page translation unit extracts a plurality of first texts and second texts expressed in the second language corresponding to the first text and the second text expressed in the first language, respectively, from the database, and extracts the plurality of extracted texts. Displays a first text and a second text expressed in a second language and displays a user interface capable of selecting any one of the displayed text, and the first text expressed in the second language selected through the user interface is displayed on the web A second text inserted into a page and displayed in a second language selected through the user interface may be inserted into the image file. The web page translation unit may re-edit the extracted second text expressed in the second language to be suitable for an insertion position of the image editing file.

17 is a block diagram showing the configuration of a system for performing image-to-text conversion according to a preferred embodiment of the present invention.

As shown in FIG. 17 , the server 500 is linked with the shopping mall 50 and the database 60 and may include a web page editor 510 and a conversion unit 520 . The server 500 can easily translate image text included in images included in various web pages for sale in the shopping mall 50 .

FIG. 18 is a flowchart for explaining the operation flow of the image-to-text conversion system shown in FIG. 17 .

Referring to FIGS. 17 to 18 , the web page editor 510 is a web-based application that operates in real time on the web and can display a web page editing screen (step: S61). The web page editor 510 may display a web page translation request icon (eg, in a form similar to that shown in FIG. 12 or 16) on the web page editing screen (step: S62).

The web page is, for example, a web page for selling products, and may include various images such as product images in the web page. The image may include a first image text expressed in a first language. The first image text is displayed as text in the first language in the image, but since it is a part of the image, it cannot be translated in the same way as a general text correction. The present invention can automatically convert untranslatable image text on the web.

When the image text translation request icon is selected, the server 500 interworks with the database 60 to convert the first image text included in the image included in the web page into a second language expressed in a second language, which is the target language. Image text may be converted into image text, and an image including the converted second image text may be updated on the web (step: S63).

19 is a flowchart for explaining the detailed operation flow of step S63 performed by the server 500.

As shown in FIG. 19 , when the image text translation request icon is selected, the web page editor 510 may extract an image editing file corresponding to the image from the database 60 (step: S71).

The image editing file may include text expressed in a first language corresponding to the first image text and text attribute information of the text expressed in the first language. For example, the image editing file may be a Photoshop editing file or the like. The text property information may include information related to the size, font, color, and highlighting effect of the text expressed in the first language, such as setting values. The emphasis effect may be information related to bold, italic, underline, shading, inverse, and emphasis color, for example, a setting value.

The web page editor 510 may display the extracted image editing file on one side of the editing screen (step: S72). The conversion unit 520 of the server 500 may extract text expressed in a second language corresponding to the text expressed in the first language from the database (step: S73). For example, assuming that the first language is Korean and the second language is English, the conversion unit 520 may extract an English sentence corresponding to a sentence expressed in Korean from the database 60 .

Translating the text expressed in the first language into the text expressed in the second language may be performed by the multilingual translation process described above with reference to FIGS. 5 to 9 .

The conversion unit 520 may convert the extracted text expressed in the second language based on the text attribute information (step: S74). For example, the conversion unit 520 applies text attribute values according to text attribute information to the extracted text expressed in the second language.

The conversion unit 520 may insert the converted text expressed in the second language based on the text property information into the image editing file (step: S75). The conversion unit 520 may generate a converted image based on the image editing file in which the text expressed in the second language is inserted (step: S76). The converted image may include second image text corresponding to text expressed in the second language. The conversion unit 520 may update the created image on the web (step: S77).

The conversion unit 520 may re-edit the text expressed in the second language to be suitable for a position to be inserted in the image editing file, based on the image editing file in which the text expressed in the second language is inserted. For example, since the length or the number of words is different for each language even if the meaning is the same, line breaks and sizes of the text are edited so that the text can be properly inserted into the area.

Meanwhile, a plurality of texts expressed in the second language may be extracted. In this case, the conversion unit 520 may display a plurality of extracted texts expressed in the second language and display a user interface for selecting one of the displayed texts. The conversion unit 520 may use text expressed in the second language selected through the user interface.

The user interface converts the extracted texts expressed in the plurality of second languages into a sequence of at least one of a frequently expressed sentence, a sentence used to describe a product with a high sales volume, and a sentence used to describe a product with the highest number of hits. Sort or search by .

As previously mentioned in the description with reference to FIG. 9 , sentences about products may be generated in multiple languages based on keyword sets and sentence template sets and managed by a database. It can be seen that a plurality of sentence sets having the same meaning are managed to correspond to a specific sentence. Each sentence may be an expression optimized for locals who speak the corresponding language by template.

The at least one sentence template set may include a basic sentence template set corresponding to the plurality of keyword sets and at least one popular sentence template set. The popular sentence template set includes sentence templates expressed in multiple languages, and the sentence templates expressed in each language are used for the sentence most frequently used in the region where the language is used and for the description of the product with the highest sales volume in the region. The template may be generated based on a locally used expression using the language, based on at least one of a sentence used to describe a product having the highest number of hits in the region.

Fig. 20 is a block diagram showing the configuration of a system for realizing an image conversion method according to a preferred embodiment of the present invention.

As shown in FIG. 20 , the server 600 is linked with the shopping mall 52 and the database 62 and may include an extraction unit 610 , a translation unit 620 and an image generation unit 630 . The server 600 can easily translate image text included in images included in various web pages for sale in the shopping mall 52 .

FIG. 21 is a flowchart for explaining an operation flow of the image conversion system shown in FIG. 20 .

As shown in FIGS. 20 to 21 , the server 600 may display a web page editing screen using a web page editor operating in real time on the web (step: S81). The web page editor may display a web page translation request icon (for example, in a form similar to that shown in FIG. 12 or 16) on the web page editing screen (step: S82).

The web page is, for example, a web page for selling products, and may include various images such as product images in the web page. The image may include a first image text expressed in a first language. The first image text is displayed as text in the first language in the image, but since it is a part of the image, it cannot be translated in the same way as a general text correction. The present invention can automatically convert untranslatable image text on the web.

When the image text translation request icon is selected, the server 600 interworks with the database 60 to convert the first image text included in the image included in the web page into the target language while maintaining its properties. The second image text expressed in the second language may be converted into second image text, and an image including the converted second image text may be updated on the web (step: S83).

22 is a flowchart for explaining the detailed operation flow of step S83 performed by the server 600.

As shown in FIG. 22 , when the image text translation request icon is selected, the extraction unit 610 analyzes the first text image in the image and extracts the text expressed in the first language (step: S91). For example, the extractor 610 may extract the text expressed in the first language by recognizing the text expressed in the first language from the image based on Optical Character Recognition (OCR). there is.

The extractor 610 may extract text attribute information corresponding to the text expressed in the first language from the first text image (step: S92). The text property information may include information related to the size, font, color, and highlighting effect of the text expressed in the first language, such as setting values. The emphasis effect may be information related to bold, italic, underline, shading, inverse, and emphasis color, for example, a setting value.

Next, the translation unit 620 may extract the text expressed in the second language corresponding to the extracted text expressed in the first language from the database 62 (step: S93). For example, assuming that the first language is Korean and the second language is English, the translation unit 620 may extract an English sentence corresponding to a sentence expressed in Korean from the database 62 .

Translating the text expressed in the first language into the text expressed in the second language may be performed by the multilingual translation process described above with reference to FIGS. 5 to 9 .

The image generator 630 generates a second text image, which is an image in which the text property information is applied to the text expressed in the second language, based on the text expressed in the second language and the text property information. Can (Step: S94)

For example, the image generator 630 applies text attribute values according to text attribute information to text expressed in the second language and converts the text expressed in the second language to which the text attribute value is applied to an image layer. A second text image may be generated.

Subsequently, the image generator 630 may generate an image including the second text image using the generated second text image (step: S95). For example, the image generator 630 may remove the first text image from the image and insert the second text image into a region from which the first text image is removed.

Here, the image generator 630 may convert the second text image to match the size or shape of the area from which the first text image is removed. That is, based on the image editing file into which the text expressed in the second language is inserted, the text expressed in the second language is re-edited so that the text expressed in the second language is suitable for the insertion position of the image editing file.

For example, even if a sentence has the same meaning, the length or number of words may vary depending on the language, so the image generator 630 appropriately inserts the second text image into the image through re-editing.

FIG. 23 exemplarily shows an original image in which a Korean sentence is inserted, and FIG. 24 exemplarily shows a converted image obtained by converting a Korean sentence into English from the image shown in FIG. 23 by the image generator 630. indicates

As shown in FIG. 23, the server 600 displays an image in which a clear gothic sentence of brown color expressed in Korean “Extended Sleeve Extended T-Shirt” is inserted, and is expressed in English “Extended Sleeve Cotton T-Shirt” by the server 600. It can be converted into a colored gothic sentence. That is, according to the present invention, it can be seen that the text in the image is automatically translated on the web and even the text properties are maintained. The image generating unit 630 may update the created image on the web (step: S96).

Meanwhile, the number of texts expressed in the second language extracted from the database 62 may be plural. In this case, the translation unit 620 may display a plurality of extracted texts expressed in the second language and display a user interface for selecting one of the displayed texts. The image generator 630 may use text expressed in the second language selected through the user interface.

The user interface converts the extracted texts expressed in the plurality of second languages into a sequence of at least one of a frequently expressed sentence, a sentence used to describe a product with a high sales volume, and a sentence used to describe a product with the highest number of hits. Sort or search by .

As previously mentioned in the description with reference to FIG. 9 , sentences about products may be generated in multiple languages based on keyword sets and sentence template sets and managed by a database. It can be seen that a plurality of sentence sets having the same meaning are managed to correspond to a specific sentence. Each sentence may be an expression optimized for locals who speak the corresponding language by template.

The at least one sentence template set may include a basic sentence template set corresponding to the plurality of keyword sets and at least one popular sentence template set. The popular sentence template set includes sentence templates expressed in multiple languages, and the sentence templates expressed in each language are used for the sentence most frequently used in the region where the language is used and for the description of the product with the highest sales volume in the region. The template may be generated based on a locally used expression using the language, based on at least one of a sentence used to describe a product having the highest number of hits in the region.

25 is a block diagram showing the configuration of a system for performing adaptive image-to-text conversion according to a preferred embodiment of the present invention.

As shown in FIG. 25 , the server 700 is linked with the shopping mall 54 and the database 64 and may include a web page editor 710 and a conversion unit 720 . The server 700 can easily and adaptively translate image text included in images included in various web pages for sale in the shopping mall 54 .

FIG. 26 is a flowchart for explaining an operation flow of the adaptive image-to-text conversion system shown in FIG. 25 .

Referring to FIGS. 25 and 26 , the web page editor 710 is a web-based application that operates in real time on the web and can display a web page editing screen (step: S101). The web page editor 710 may display a web page translation request icon (for example, in a form similar to that shown in FIG. 12 or 16) on the web page editing screen (step: S102).

The web page is, for example, a web page for selling products, and may include various images such as product images in the web page. The image may include a first image text expressed in a first language. The first image text is displayed as text in the first language in the image, but since it is a part of the image, it cannot be translated in the same way as a general text correction. The present invention can automatically convert untranslatable image text on the web.

When the image text translation request icon is selected, the server 700 interworks with the database 64 to adaptively convert the first image text included in the image included in the web page into the target language according to image analysis. Second image text expressed in two languages may be converted into second image text, and an image including the converted second image text may be updated on the web (step: S103).

27 is a flowchart for explaining the detailed operation flow of step S103 performed by the server 700.

As shown in FIG. 27 , when the image text translation request icon is selected, the web page editor 710 may extract an image editing file corresponding to the image from the database 64 (step: S111).

The image editing file may include text expressed in a first language corresponding to the first image text and first text attribute information of the text expressed in the first language. For example, the image editing file may be a Photoshop editing file or the like. The text property information may include information related to the size, font, color, and highlighting effect of the text expressed in the first language, for example, setting values. The emphasis effect may be information related to bold, italic, underline, shading, inverse, and emphasis color, for example, a setting value.

The web page editor 710 may display the extracted image editing file on one side of the editing screen (step: S112). The conversion unit 720 of the server 700 may extract text expressed in a second language corresponding to the text expressed in the first language from the database (step: S113). For example, assuming that the first language is Korean and the second language is English, the conversion unit 720 may extract an English sentence corresponding to a sentence expressed in Korean from the database 64 .

Translating the text expressed in the first language into the text expressed in the second language may be performed by the multilingual translation process described above with reference to FIGS. 5 to 9 .

The conversion unit 720 may analyze the image and generate second text attribute information according to the analysis result (step: S114). That is, by analyzing the image, the properties of the translated sentence to be inserted into the image after translation are also converted to suit the image.

Specifically, the conversion unit 720 may set a related text attribute value based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image by analyzing the image.

For example, the converter 720 may extract emotion information corresponding to the image and set a font value associated with the extracted emotion information. The conversion unit 720 may extract emotion information by analyzing a facial expression of a model included in an image, and may set a corresponding predetermined text attribute value.

Here, emotion information can be divided into emotion factors such as anger, contempt, disgust, fear, happiness, neutral expression, sadness, and surprise. there is. Each image may have a score for at least some of these eight emotion factors. Scores can be given from 0 to 1.0. An emotion information extraction algorithm for extracting such an emotion factor may be evolved by machine learning by matching an object or color in an image with a corresponding emotion score value. For example, it can be learned by generating (object-related information, emotion scores for each emotion factor) as training data and inputting them to a deep learning model. The device may store in advance a text attribute value corresponding to the emotion score obtained in this way.

The conversion unit 720 may extract complexity information corresponding to the image and set at least one of a text thickness value and a font value associated with the extracted complexity information.

For example, since the higher the complexity of the image, the more readable text should be used. Therefore, bold text may be applied when the complexity is high, and thin text may be applied when the complexity is low. That is, the complexity is divided into high/middle/low, and matching is applied so that "Bold" is applied in case of "high", "Regular font" is applied in case of "medium", and "thin font" is applied in case of "low". . On the other hand, complexity can be used to determine the font of text. For example, it may be controlled to use a font with high readability when the complexity of the image is high, and a font with high readability when the complexity of the image is low. Readability represents the degree of how easily a text can be read, and legibility may represent the degree of conspicuousness in the process of identifying and recognizing the shape of a font.

The converter 720 may extract brightness information corresponding to an image and set a text color value associated with the extracted brightness information. For example, the conversion unit 720 may extract brightness information of an image and match it with a color attribute value of text. Brightness represents the degree of lightness and darkness of a color. This brightness can affect the color of text. Accordingly, the conversion unit 720 may determine the color of the text based on the text color attribute value table corresponding to the previously stored image brightness value. Since the image is in a bright state when the brightness is high, a dark color corresponding to the color of the text may be suitable. If the brightness is low, the image is dark, so it is preferable to use a conspicuous color (eg, red) for the text.

The converter 720 may extract complementary color information corresponding to the image and set an emphasis color value associated with the extracted complementary color information. Complementary color information of an image may be extracted and matched to a text accent color attribute value. The conversion unit 720 can obtain complementary color information based on data of primary color code values (information including Hex codes, RGB values, and percentages) from the image analysis unit, and emphasizes on the banner screen by using the complementary color information. It can be used for the element you want to do or the element that interacts with it. Complementary colors are calculated through │ (H value of primary color + 180ㅀ) - 360ㅀ│ based on primary color data. The color of the highlighted area of the text can be determined using this complementary color information. In particular, if the background color of the corresponding image is the main color, the highlighted part of the text is made white or black so that it can be emphasized. The highlighted part of the text can be extracted through morphological analysis and semantic analysis.

The conversion unit 720 may convert the text expressed in the second language based on any one of the first text property information and the second text property information (step: S115). For example, the conversion unit 720 applies first text property information and second text property information to the text expressed in the second language, respectively, compares them, and displays a user interface capable of selection, and the user The text expressed in the second language may be converted according to text property information selected based on the interface. For example, the conversion unit 720 compares text expressed in the second language with images when first text property information is applied and images when second text property information is applied.

The conversion unit 720 inserts the text expressed in the second language converted based on the first or second text property information into the image editing file, and inserts the text expressed in the second language into the image editing file. A converted image may be generated based on (step: S106). The converted image may include second image text corresponding to text expressed in the second language. The conversion unit 720 may update the created image on the web (step: S117).

Meanwhile, a plurality of texts expressed in the second language may be extracted. In this case, the conversion unit 720 may display a plurality of extracted texts expressed in the second language and display a user interface for selecting one of the displayed texts. The conversion unit 720 may use text expressed in the second language selected through the user interface.

The user interface converts the extracted texts expressed in the plurality of second languages into a sequence of at least one of a frequently expressed sentence, a sentence used to describe a product with a high sales volume, and a sentence used to describe a product with the highest number of hits. Sort or search by .

As previously mentioned in the description with reference to FIG. 9 , sentences about products may be generated in multiple languages based on keyword sets and sentence template sets and managed by a database. It can be seen that a plurality of sentence sets having the same meaning are managed to correspond to a specific sentence. Each sentence may be an expression optimized for locals who speak the corresponding language by template.

The at least one sentence template set may include a basic sentence template set corresponding to the plurality of keyword sets and at least one popular sentence template set. The popular sentence template set includes sentence templates expressed in multiple languages, and the sentence templates expressed in each language are used for the sentence most frequently used in the region where the language is used and for the description of the product with the highest sales volume in the region. The template may be generated based on a locally used expression using the language, based on at least one of a sentence used to describe a product having the highest number of hits in the region.

28 is a block diagram showing the configuration of a system for realizing an adaptive image conversion method according to a preferred embodiment of the present invention.

As shown in FIG. 28 , the server 800 is linked with the shopping mall 56 and the database 66 and may include an extraction unit 810 , a translation unit 820 and an image generation unit 830 . The server 800 may easily and adaptively translate image text included in images included in various web pages for sale in the shopping mall 56 .

FIG. 29 is a flowchart for explaining an operation flow of the adaptive image conversion system shown in FIG. 28 .

As shown in FIGS. 28 to 29 , the server 800 may display a web page editing screen using a web page editor operating in real time on the web (step: S121). The web page editor may display a web page translation request icon (eg, in a form similar to that shown in FIG. 12 or 16) on the web page editing screen (step: S122).

The web page is, for example, a web page for selling products, and may include various images such as product images in the web page. The image may include a first image text expressed in a first language. The first image text is displayed as text in the first language in the image, but since it is a part of the image, it cannot be translated in the same way as a general text correction. The present invention can adaptively and automatically convert untranslatable image text on the web.

When the image text translation request icon is selected, the server 800 interworks with the database 66 to adaptively convert the first image text included in the image included in the web page into the target language according to image analysis. Second image text expressed in two languages may be converted into second image text, and an image including the converted second image text may be updated on the web (step: S123).

30 is a flowchart for explaining the detailed operation flow of step S123 performed by the server 800.

As shown in FIG. 30 , when the image text translation request icon is selected, the extraction unit 810 analyzes the first text image in the image and extracts the text expressed in the first language (step: S131). For example, the extractor 810 may extract the text expressed in the first language by recognizing the text expressed in the first language from the image based on Optical Character Recognition (OCR). there is.

The extractor 810 may extract first text attribute information corresponding to the text expressed in the first language from the first text image (step: S132). The first text property information may include information related to the size, font, color, and highlighting effect of the text expressed in the first language, such as setting values. The emphasis effect may be information related to bold, italic, underline, shading, inverse, and emphasis color, for example, a setting value.

Next, the translation unit 820 may extract the text expressed in the second language corresponding to the extracted text expressed in the first language from the database 66 (step: S133). For example, assuming that the first language is Korean and the second language is English, the translation unit 820 may extract an English sentence corresponding to a sentence expressed in Korean from the database 66 .

Translating the text expressed in the first language into the text expressed in the second language may be performed by the multilingual translation process described above with reference to FIGS. 5 to 9 .

The image generating unit 830 may analyze the image and generate second text attribute information according to the analysis result (step: S134). That is, by analyzing the image, the properties of the translated sentence to be inserted into the image after translation are also converted to suit the image.

Specifically, the image generator 830 may set a related text attribute value based on analyzing the image and extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image.

For example, the image generator 830 may extract emotion information corresponding to the image and set a font value associated with the extracted emotion information. The image generator 830 may extract emotion information by analyzing a facial expression of a model included in the image, and may set a corresponding predetermined text attribute value.

Here, emotion information can be divided into emotion factors such as anger, contempt, disgust, fear, happiness, neutral expression, sadness, and surprise. there is. Each image may have a score for at least some of these eight emotion factors. Scores can be given from 0 to 1.0. An emotion information extraction algorithm for extracting such an emotion factor may be evolved by machine learning by matching an object or color in an image with a corresponding emotion score value. For example, it can be learned by generating (object-related information, emotion scores for each emotion factor) as training data and inputting them to a deep learning model. The device may store in advance a text attribute value corresponding to the emotion score obtained in this way.

The image generator 830 may extract complexity information corresponding to the image and set at least one of a text thickness value and a font value associated with the extracted complexity information.

For example, since the higher the complexity of the image, the more readable text should be used. Therefore, bold text may be applied when the complexity is high, and thin text may be applied when the complexity is low. That is, the complexity is divided into high/middle/low, and matching is applied so that "Bold" is applied in case of "high", "Regular font" is applied in case of "medium", and "thin font" is applied in case of "low". . On the other hand, complexity can be used to determine the font of text. For example, it may be controlled to use a font with high readability when the complexity of the image is high, and a font with high readability when the complexity of the image is low. Readability represents the degree of how easily a text can be read, and legibility may represent the degree of conspicuousness in the process of identifying and recognizing the shape of a font.

The image generator 830 may extract brightness information corresponding to the image and set a text color value associated with the extracted brightness information. For example, the image generating unit 830 may extract brightness information of an image and match it with a color attribute value of text. Brightness represents the degree of brightness and darkness of a color. This brightness can affect the color of text. Accordingly, the image generator 830 may determine the color of the text based on the text color attribute value table corresponding to the previously stored image brightness value. Since the image is in a bright state when the brightness is high, a dark color corresponding to the color of the text may be suitable. If the brightness is low, the image is dark, so it is preferable to use a conspicuous color (eg, red) for the text.

The image generating unit 830 may extract complementary color information corresponding to the image and set an emphasis color value associated with the extracted complementary color information. Complementary color information of an image may be extracted and matched to a text accent color attribute value. The conversion unit 720 can obtain complementary color information based on data of primary color code values (information including Hex codes, RGB values, and percentages) from the image analysis unit, and emphasizes on the banner screen by using the complementary color information. It can be used for the element you want to do or the element that interacts with it. Complementary colors are calculated through │ (H value of primary color + 180ㅀ) - 360ㅀ│ based on primary color data. The color of the highlighted area of the text can be determined using this complementary color information. In particular, if the background color of the corresponding image is the main color, the highlighted part of the text is made white or black so that it can be emphasized. The highlighted part of the text can be extracted through morphological analysis and semantic analysis.

The image generator 830 may convert the text expressed in the second language based on any one of the first text property information and the second text property information (step: S135). For example, the image generator 830 applies first text property information and second text property information to the text expressed in the second language, respectively, compares them, and displays a user interface capable of selection, and the user The text expressed in the second language may be converted according to text property information selected based on the interface. For example, the image generator 830 compares text expressed in the second language with images when first text property information is applied and images when second text property information is applied.

The image generator 830 generates an image in which first or second text attribute information is applied to text expressed in the second language based on the text expressed in the second language and the first or second text attribute information. A second text image may be generated (step: S136).

For example, the image generator 830 applies text property values according to first or second text property information to text expressed in the second language, and converts text expressed in the second language to which the text property values are applied. By converting to an image layer, a second text image can be created.

Subsequently, the image generator 830 may generate an image including the second text image using the generated second text image (step: S137). For example, the image generator 830 may remove the first text image from the image and insert the second text image into a region from which the first text image is removed.

Here, the image generator 830 may convert the second text image to match the size or shape of the area from which the first text image is removed. That is, based on the image editing file into which the text expressed in the second language is inserted, the text expressed in the second language is re-edited so that the text expressed in the second language is suitable for the insertion position of the image editing file.

For example, even if a sentence has the same meaning, the length or number of words may vary depending on the language, so the image generator 830 appropriately inserts the second text image into the image through re-editing.

FIG. 31 exemplarily shows an original image in which Korean sentences are inserted, and FIG. 32 exemplifies a converted image obtained by adaptively converting Korean sentences into English from the image shown in FIG. 31 by the image generator 830. is shown negatively.

As shown in FIG. 31, an image in which a clear gothic sentence of green color expressed in Korean, “Extended Sleeve Extended T-Shirt,” is inserted, is displayed by the server 800 in navy blue expressed in English, “Extended Sleeve Cotton T-Shirt.” By being converted into a couture style sentence of color, it can be converted into a more sophisticated image text as well as translation. That is, according to the present invention, it can be seen that text in an image is automatically translated on the web and even text properties are converted to fit the image. The image generating unit 830 may update the created image on the web.

Meanwhile, the number of texts expressed in the second language extracted from the database 66 may be plural. In this case, the translation unit 820 may display a plurality of extracted texts expressed in the second language and display a user interface for selecting one of the displayed texts. The image generator 830 may use text expressed in a second language selected through the user interface.

The user interface converts the extracted texts expressed in the plurality of second languages into a sequence of at least one of a frequently expressed sentence, a sentence used to describe a product with a high sales volume, and a sentence used to describe a product with the highest number of hits. Sort or search by .

As previously mentioned in the description with reference to FIG. 9 , sentences about products may be generated in multiple languages based on keyword sets and sentence template sets and managed by a database. It can be seen that a plurality of sentence sets having the same meaning are managed to correspond to a specific sentence. Each sentence may be an expression optimized for locals who speak the corresponding language by template.

The at least one sentence template set may include a basic sentence template set corresponding to the plurality of keyword sets and at least one popular sentence template set. The popular sentence template set includes sentence templates expressed in multiple languages, and the sentence templates expressed in each language are used for the sentence most frequently used in the region where the language is used and for the description of the product with the highest sales volume in the region. The template may be generated based on a locally used expression using the language, based on at least one of a sentence used to describe a product having the highest number of hits in the region.

Although the present invention described above has been described based on a series of functional blocks, it is not limited by the above-described embodiments and the accompanying drawings, and various substitutions, modifications, and changes are made within the scope of the technical spirit of the present invention. That this is possible will be apparent to those skilled in the art.

Combinations of the above-described embodiments are not limited to the above-described embodiments, and may be provided in various forms as well as the above-described embodiments according to implementation and/or needs.

In the foregoing embodiments, the methods are described on the basis of a flow chart as a series of steps or blocks, but the present invention is not limited to the order of steps, and some steps may occur in a different order or concurrently with other steps as described above. there is. In addition, those skilled in the art will understand that the steps shown in the flow chart are not exclusive, that other steps may be included, or that one or more steps of the flow chart may be deleted without affecting the scope of the present invention. You will understand.

The foregoing embodiment includes examples of various aspects. It is not possible to describe all possible combinations to represent the various aspects, but those skilled in the art will recognize that other combinations are possible. Accordingly, it is intended that the present invention cover all other substitutions, modifications and variations falling within the scope of the following claims.

Although the present invention has been described by exemplifying preferred embodiments, those skilled in the art will variously modify and change the present invention within the scope not departing from the technical details and scope of the present invention described in the claims below. You will understand that this can be done. Accordingly, changes to future embodiments of the present invention will not deviate from the technology of the present invention.

Claims (25)

It is performed by a web-based conversion device,
Displaying an editing screen of a web page including an image based on a web-based web page editor - the image includes a first image text;
extracting an image editing file corresponding to the image from a database, wherein the image editing file includes text expressed in a first language corresponding to the first image text and first text attribute information;
displaying the extracted image editing file on one side of the editing screen;
extracting text expressed in a second language corresponding to the text expressed in the first language from the database;
analyzing the image and generating second text attribute information according to an analysis result;
and converting the text expressed in the second language based on any one of the first text property information and the second text property information.
The method of claim 1 , further comprising: inserting the converted text expressed in the second language into the image editing file;
Generating a converted image based on an image editing file into which text expressed in the second language is inserted, wherein the converted image includes second image text corresponding to text expressed in the second language- ; and
Image text conversion method characterized in that it further comprises the step of updating the generated image on the web.
The method of claim 1, wherein generating the second text attribute information comprises:
and setting a related text attribute value based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image.
4. The method of claim 3, wherein setting the associated text property value comprises:
extracting emotion information corresponding to the image; and
Image text conversion method comprising the step of setting a font value associated with the extracted emotion information.
4. The method of claim 3, wherein setting the associated text property value comprises:
extracting complexity information corresponding to the image; and
and setting at least one of a text weight value and a font value associated with the extracted complexity information.
4. The method of claim 3, wherein setting the associated text property value comprises:
extracting brightness information corresponding to the image; and
Image text conversion method comprising the step of setting a text color value associated with the extracted brightness information.
4. The method of claim 3, wherein setting the associated text property value comprises:
extracting complementary color information corresponding to the image; and
An image-to-text conversion method comprising the step of setting an accent color value associated with the extracted complementary color information.
The method of claim 1, wherein the converting step,
displaying, on the editing screen, a user interface capable of comparing and selecting first text property information and second text property information, respectively, to the text expressed in the second language; and
and converting text expressed in the second language according to text property information selected based on the user interface.
Based on a web-based web page editor, displaying an editing screen of a web page including an image—the image includes a first image text—extracting an image editing file corresponding to the image from a database—the image The edit file includes text expressed in a first language corresponding to the first image text and first text attribute information-, a web page editor displaying the extracted image edit file on one side of the edit screen; and
Text expressed in a second language corresponding to the text expressed in the first language is extracted from the database, the image is analyzed to generate second text attribute information according to an analysis result, and text expressed in the second language and a conversion unit for converting text based on any one of the first text property information and the second text property information.
The method of claim 9, wherein the conversion unit,
Inserting the converted text expressed in the second language into the image editing file, generating a converted image based on the image editing file into which the text expressed in the second language is inserted - the converted image is the An image-to-text conversion device comprising a second image text corresponding to text expressed in a second language, and updating the generated image on the web.
The method of claim 9, wherein the conversion unit,
Based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image, a related text property value is set.
The method of claim 11, wherein the conversion unit,
Image-to-text conversion device characterized in that for extracting emotion information corresponding to the image, and setting a font value associated with the extracted emotion information.
The method of claim 11, wherein the conversion unit,
Extracting complexity information corresponding to the image;
Image-to-text conversion device, characterized in that for setting at least one of a text thickness value and a font value associated with the extracted complexity information.
The method of claim 11, wherein the conversion unit,
Extract brightness information corresponding to the image;
An image-to-text conversion device characterized in that for setting a text color value associated with the extracted brightness information.
The method of claim 11, wherein the conversion unit,
An image-to-text conversion device characterized in that for extracting complementary color information corresponding to the image and setting an accent color value associated with the extracted complementary color information.
The method of claim 9, wherein the conversion unit,
The first text property information and the second text property information are respectively applied to the text expressed in the second language, compared, and a user interface capable of selection is displayed, and the text property information selected based on the user interface is displayed. An image-to-text conversion device characterized in that for converting text expressed in a second language.
database; and
Displaying an editing screen of a web page including an image based on a web-based web page editor—the image includes a first image text—extracting an image editing file corresponding to the image from the database—the The image editing file includes text expressed in a first language corresponding to the first image text and first text property information-, the extracted image editing file is displayed on one side of the editing screen, and the first The text expressed in the second language corresponding to the text expressed in the language is extracted from the database, the image is analyzed to generate second text attribute information according to the analysis result, the text expressed in the second language, Image-to-text conversion system characterized in that it comprises an image-to-text conversion unit for converting based on any one of the first text property information and the second text property information.
The method of claim 17, wherein the image text conversion unit,
Inserting the converted text expressed in the second language into the image editing file, generating a converted image based on the image editing file into which the text expressed in the second language is inserted - the converted image is the An image-to-text conversion system comprising a second image text corresponding to text expressed in a second language, and updating the created image on the web.
The method of claim 17, wherein the image text conversion unit,
Based on extracting at least one of emotion information, complexity information, brightness information, and complementary color information corresponding to the image, a related text property value is set.
The method of claim 19, wherein the image text conversion unit,
Image-to-text conversion system, characterized in that for extracting emotion information corresponding to the image, and setting a font value associated with the extracted emotion information.
The method of claim 19, wherein the image text conversion unit,
Extracting complexity information corresponding to the image;
Image text conversion system, characterized in that for setting at least one of a text thickness value and a font value associated with the extracted complexity information.
The method of claim 19, wherein the image text conversion unit,
Extract brightness information corresponding to the image;
Image text conversion system, characterized in that for setting a text color value associated with the extracted brightness information.
The method of claim 19, wherein the image text conversion unit,
An image-to-text conversion device characterized in that for extracting complementary color information corresponding to the image and setting an accent color value associated with the extracted complementary color information.
The method of claim 17, wherein the image text conversion unit,
The first text property information and the second text property information are respectively applied to the text expressed in the second language, compared, and a user interface capable of selection is displayed, and the text property information selected based on the user interface is displayed. An image text conversion system characterized in that for converting text expressed in a second language.
on the computer,
Displaying an editing screen of a web page including an image based on a web-based web page editor, wherein the image includes a first image text;
extracting an image editing file corresponding to the image from a database, wherein the image editing file includes text expressed in a first language corresponding to the first image text and first text property information;
displaying the extracted image editing file on one side of the editing screen;
extracting text expressed in a second language corresponding to the text expressed in the first language from the database;
analyzing the image and generating second text attribute information according to an analysis result;
A program stored in a medium to execute a step of converting the text expressed in the second language based on any one of the first text property information and the second text property information.

Images