KR20090053109A - Apparatus for creating semantic metadata of image and method thereof, and system for offering semantic metadata of image using that and method thereof - Google Patents

Abstract

The present invention relates to an apparatus for generating image semantic metadata, a method thereof, a system for providing image semantic metadata using the same, and a computer readable recording medium having recorded thereon a program for realizing the method. Generates semantic metadata for the image based on the description information of the user and information selected by the user who is the target of image search (selection information), and an article-based image semantic metadata service based on the generated image semantic metadata An apparatus for generating image semantic metadata and a method thereof, an image semantic metadata providing system using the same, and a computer readable recording medium having recorded thereon a program for realizing the method and the method are provided.

To this end, the present invention provides an image semantic metadata generating apparatus, comprising: image metadata generating means for generating image metadata using peripheral description information of an image; And image semantic metadata generating means for recommending the image and the corresponding image metadata to the search user, and receiving the feedback of the search user's selection information to generate the semantic metadata for the image.

Image, Semantic Web, Ontology, Image Metadata, Image Semantic Metadata, Creation, Search, Description Information, Article Based

Description

Apparatus for creating semantic metadata of image and method Thereof, and system for offering semantic metadata of image using that and method Thereof}

The present invention provides an apparatus and method for generating image semantic metadata for generating and providing semantic metadata for an image included in an article (web article, newspaper article) and the like, and a system and method for providing image semantic metadata using the same The present invention relates to a computer-readable recording medium that records a program for realizing each method. More specifically, the title and image of the article, which is information of the surrounding description of the image included in the article (web article, newspaper article), etc. Photo) Acquire metadata about an image based on the main vocabulary (meaning unit) of description and article contents, and recommend image and metadata about the image to the user at the same time and generate image semantic metadata according to the feedback result. Apparatus and method for generating image semantic metadata for providing the same Yonghan image relates to semantic metadata providing system and a computer-readable recording medium recording a program for realizing the procedure as described in the respective methods.

In the following embodiment, the image (Image) is an example included in the image included in the article (News), such as a web article or newspaper article, for example, but the present invention is not limited thereto.

The image metadata description of a conventional article only allows repository information of images and meaningless image file names (e.g. http://imgnews.naver.com/image/014/2007/10/09/00000921117267.jpg ) or user Since there is only tag information and a classification method of a provider, it should be regarded as a simple keyword search rather than a desired image search.

More specifically, FIG. 1 is a diagram illustrating an exemplary embodiment of an existing system that photographs a conventional image, registers it with a web article, and stores the image.

Referring to FIG. 1, a conventional article management system collects articles (web articles) which are web documents from a web article collector 110 through a mutual contract with a news content provider, and temporarily store them in a web document repository 120. Save as.

Since the web article temporarily stored in the web document storage 120 is composed of an image (including multimedia data) and text, after being classified into image data and text data by the text classifier 130, the image database 150 is respectively. And are stored and managed in the text database 160. In this case, the image data and the text data classified by the text classifier 130 are separately stored in the image database 150 and the text database 160 respectively, or the image is stored through the image / text metadata storage unit 140. The database 150 and the text database 160 are stored separately.

At this time, the text data and the image data are stored in different places (directory, ie, text database and image database) even though they are from one web article, so that the relevant location information is stored in the image / text metadata storage unit 140. Stored as metadata (140). In other words, the metadata stored in the image / text metadata storage 140 is an article number, an image / text storage location, and the like.

As described above, in general, web articles are classified into image data and text data, and article numbers are managed by a link, or by entering classification criteria and tags, such as categories, based on a service provider or a general user. Managing image metadata can't provide users with more satisfying search results.

In addition, the conventional article classification-based search system performs keyword-based search term processing, and search users are provided with a huge amount of search results because they search through input of only 1-2 words. Not only is it difficult to find the desired information, but there is also a need for another method of searching or additional actions of the user in order to obtain semantically related documents or photos.

Therefore, by improving existing search service and search technology for web articles or newspaper articles including variously produced images, search users can efficiently and conveniently retrieve the information they want while minimizing the cognitive load. There is an urgent need for a method of generating and providing semantic metadata for searching.

That is, the prior art as described above has a problem in that input time, cognitive burden, and psychological burden for image search and management are increased due to an increase in the frequency of searching for images and a large amount of images, and also desired by a search user. Not only is it difficult to find the information, but there is a problem that requires another method of searching or additional actions of the user in order to obtain semantically related contents (documents or photographs). It is a task.

Accordingly, the present invention provides an apparatus and method for generating image semantic metadata for generating semantic metadata for an image based on description information of the image and selection information of a search user, rather than features of the image itself. It is an object of the present invention to provide a computer-readable recording medium having recorded thereon a program for realizing the method.

The present invention also provides an image semantic metadata providing system and method for providing image semantic metadata generated based on the description information of the image and the selection information of the search user rather than the feature of the image itself. And a computer readable recording medium having recorded thereon a program for realizing the above method.

That is, the present invention generates semantic metadata for an image based on the description information of the article to which the image belongs and the information (selection information) selected by a user who is an image search target and generates the semantic metadata of the image. Based on this, we want to provide an article-based image semantic metadata service.

The objects of the present invention are not limited to the above-mentioned objects, and other objects and advantages of the present invention which are not mentioned above can be understood by the following description, and will be more clearly understood by the embodiments of the present invention. Also, it will be readily appreciated that the objects and advantages of the present invention may be realized by the means and combinations thereof indicated in the claims.

According to an aspect of the present invention, there is provided an apparatus for generating image semantic metadata, comprising: image metadata generating means for generating image metadata using peripheral description information of an image; And image semantic metadata generating means for recommending the image and the corresponding image metadata to the search user, and receiving the feedback of the search user's selection information to generate the semantic metadata for the image.

According to an aspect of the present invention, there is provided a system for providing image semantic metadata, comprising: image metadata generating means for generating image metadata using peripheral description information of an image; Image semantic metadata generating means for recommending an image and corresponding image metadata to a search user, and receiving and receiving feedback of the search user's selection information to generate and store semantic metadata for the image; And searching for and providing an image corresponding to the keyword input from the search user and corresponding image metadata, and receiving feedback of the search user's selection information to update the stored image semantic metadata. Means;

On the other hand, the method of the present invention for achieving the above object, in the image semantic metadata generating method, the first image metadata generating step of generating image metadata using the description information of the image; And generating a first image semantic metadata for recommending the image and the corresponding image metadata to the search user, and receiving the feedback of the search user's selection information to generate the semantic metadata for the corresponding image.

In addition, another method of the present invention for achieving the above object is, Image semantic metadata providing method, Image metadata generation step of generating image metadata using the description information of the image; An image semantic metadata generation step of recommending an image and corresponding image metadata to a search user, and generating and storing semantic metadata of the corresponding image by receiving feedback of the selection information of the search user; A search and update step of searching for and providing an image corresponding to a keyword input from the search user and corresponding image metadata, and receiving the feedback of the search user's selection information to update the stored image semantic metadata; It includes.

On the other hand, the present invention, an image semantic metadata generating device having a processor, the image metadata generating function for generating image metadata using the description information of the image (Description); And a program for recommending an image and the corresponding image metadata to the search user, and receiving a feedback of the search user's selection information to generate an image semantic metadata generating function for generating the semantic metadata for the image. It provides a recording medium that can be read by.

The present invention also provides an image semantic metadata providing system having a processor, comprising: an image metadata generating function for generating image metadata using description information of an image; An image semantic metadata generation function for recommending an image and corresponding image metadata to a search user and generating and storing semantic metadata for the corresponding image by receiving feedback of the search user's selection information; And a search and update function for searching for and providing an image corresponding to a keyword input from the search user and corresponding image metadata, and receiving the feedback of the search user's selection information to update the stored image semantic metadata. A computer readable recording medium having recorded thereon a program for realizing the present invention is provided.

As described above, the present invention generates semantic metadata for an image based on description information of the image and selection information of a search user, rather than features of the image itself, and generates the image semantic metadata. Using the image semantic metadata service can be provided.

In addition, the present invention minimizes the user's input time, cognitive burden, and psychological burden corresponding to the cost of searching and managing due to the numerous images generated according to the increase in the frequency of searching for images and the spread of digital cameras. In addition to maximizing the efficiency and convenience of the company's profits, there is an excellent effect of additionally easily acquiring semantically related contents (documents or photos).

In addition, the semantic metadata for the image generated in accordance with the present invention is suitable to be applied to processing such as searching and classification.

The above objects, features, and advantages will become more apparent from the detailed description given hereinafter with reference to the accompanying drawings, and accordingly, those skilled in the art to which the present invention pertains may share the technical idea of the present invention. It will be easy to implement. In addition, in describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted.

First, the semantic web technology will be described to help understand the present invention.

The World Wide Web (WWW), proposed by Tim Berners-Lee in 1989, expresses information in a simple markup language called HyperText Markup Language (HTML) to share information in the Internet space. By doing so, it has revolutionized Internet-based information sharing and delivery. As a result, a huge information space was established that interconnected various information resources, and the Internet rapidly spread to the real world, thereby innovating the information society.

However, the existing web technologies face many problems as the amount of web information increases. In particular, since the existing web only allows access to information by keywords, it generates a flood of information by producing a myriad of unnecessary information for simple keyword search. In addition, the existing web is a human-oriented information processing technology that the computer does not provide enough functions to effectively extract, interpret, and process the necessary information on its own.

Therefore, the semantic web is a technology that can extend the existing web to realize semantic interoperability based on well-defined meanings that computers can understand and to build an effective cooperative system between humans and computers. It appeared.

The Semantic Web is a semantic-based search that allows not only people but machines (computers) to understand the information that exists on the web, and finds only the results that meet the user's needs, and facilitates collaboration between people and machines or machines and machines. By doing so, it refers to a web that can be used automatically for people.

In other words, the semantic web is a next-generation web technology that enables a computer to understand, automate, integrate, and reuse the meaning of information resources.

1) Ontology

Ontology is a formal specification system for shared conceptualization and provides semantic information of domain vocabulary. Ontology is a kind of knowledge expression, and the computer can understand the concept represented by the ontology and process the knowledge. In order to deal with inferences, the ontology's axiom and rule system are needed.

2) semantically annotated web

An ontology annotated web, a knowledge base. The Semantic Web can build a huge knowledge base that semantically integrates the distributed information resources of the Internet. In a narrow sense, it may be possible to build a knowledge base of information resources of a company or institution.

3) agent

It is an intelligent agent that collects, retrieves and infers information resources on behalf of a person (user) and exchanges information with other agents. Intelligent agents are the core of semantic web-based application systems.

The semantic web realizes semantic interoperability by using ontology and agent technology, and thus, the semantic web can leap from the information-based web to the knowledge-based semantic web.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2 is a block diagram of an apparatus for generating semantic metadata of an article-based image and a system for providing semantic metadata of an article-based image using the same according to an exemplary embodiment of the present invention.

As shown in FIG. 2, the apparatus for generating semantic metadata of an article-based image according to an exemplary embodiment of the present invention may include image meta for generating initial image metadata by using description information of the image. Image semantic metadata for recommending an image and image metadata associated with the search user to the data generation server 20 and a search user, and generating semantic metadata for the image by receiving feedback of the search user's selection information. And a generation server 30.

Next, the components of the image metadata generation server 20 will be described in more detail as follows.

First, the web article collecting unit 210 collects a web article (article in a web document) from an external news content provider through a mutual contract or the like.

In this case, since various web articles are provided from different news content providers, the web articles are temporarily stored in the web document storage 220 in consideration of news content providers or article categories. In this case, the web article temporarily stored in the web document storage 220 includes image data, text data, multimedia data, and the like.

In addition, the image metadata generator 230 generates initial image metadata by using the description information of the image temporarily stored in the web document storage 220 and the image semantic metadata generation server together with the corresponding image. Transfer to the image metadata recommendation unit 310 of (30). The image metadata generator 230 will be described later in detail with reference to FIG. 3.

Here, the image metadata generator 230 may be implemented by using a parser as an example. In this case, the parser compares the web article with the terminology dictionary 250 or the ontology 240 for each field of the area to which the web article belongs, or compares the image with the terminology dictionary 250 and the ontology 240 for each field. Generate metadata. For example, the parser may include the entire entertainment page article (including images) temporarily stored in the web document repository 220, including the person DB (including the celebrity name, gender, age, etc.) in the terminology dictionary 250 for each field. Use to compare. The matching vocabulary is regarded as initial image metadata in the comparison process using the field-specific terminology dictionary 250. In addition, when there is a pre-established entertainment-related ontology, the vocabulary obtained by matching the ontology 240 and the web article is also regarded as the initial image metadata.

As a result, the image metadata server 20 generates various initial image metadata that may be an alternative. The initial image metadata generated as described above is transferred to the image metadata recommendation unit 310 of the image semantic metadata generation server 30 together with the corresponding image and used to generate final semantic metadata for the image.

Next, the components of the image semantic metadata generation server 30 will be described in more detail as follows.

First, the initial image metadata received from the image metadata server 20 is used to provide an image search result according to a keyword inputted by the search user on a screen where the search user searches an article based on an image. That is, the image metadata recommendation unit 310 of the image semantic metadata generation server 30 determines whether or not the initial image metadata matches the keyword input by the search user, and if the image is matched, the image and the image metadata related thereto. I recommend (provide) as a search result at the same time. In this case, the image metadata recommendation unit 310 searches the image semantic metadata database 340 by the keyword inputted by the search user, and searches the image semantic metadata database 340 and the image metadata according to the searched image. You can also recommend it as a search result.

As described above, the image metadata recommendation unit 310 searches for an image corresponding to a keyword input from the search user and image metadata related thereto, and recommends the image metadata to the search user.

Accordingly, when a search user selects (clicks) a specific image from the image and image metadata recommended by the image metadata recommendation unit 310 or additionally selects specific image metadata, the selection information accordingly generates image semantic metadata. It is fed back to the selection information receiver 320 of the server 30.

Then, the selection information receiver 320 receives the image and the image metadata along with the selection information according to the search result selection of the search user and transmits the image and the image metadata to the image semantic metadata generator 330.

Then, the image semantic metadata generating unit 330 generates image semantic metadata for the corresponding image based on the selection information and the image metadata transmitted through the selection information receiving unit 320 and the image semantic metadata database 340. ).

That is, the image semantic metadata generating unit 330 adjusts the ranking of the corresponding image metadata based on the selection information received through the selection information receiving unit 320 to generate image semantic metadata for the corresponding image. Stored in the semantic metadata database 340.

In more detail, the image semantic metadata generator 330 will be described in detail. For example, the image semantic metadata generator 330 may refer to selection information received through the selection information receiver 320. After identifying which image has been selected or additionally which image metadata has been selected, the image metadata ranking is adjusted by weighting the image metadata associated with the selected image, and additionally, various image metadata associated with the image. The image metadata ranking is adjusted by giving additional weight to additionally selected image metadata. That is, the image semantic metadata generating unit 330 checks the selection information and, when a specific image is selected, adjusts the image metadata ranking by weighting the entire image metadata of the specific image and variously related to the specific image. If a specific image metadata is additionally selected from the alternatives of the image metadata, the weight of the image metadata is additionally adjusted to adjust the image metadata ranking so as to be representative image metadata of the image.

As described above, if a process of recommending an image and image metadata related thereto and changing the ranking of the image metadata by receiving feedback of selection information according to the selection of the search user according to the search user is performed, As a result, reliable image semantic metadata for the image is generated. In addition, the image semantic metadata generated as described above is circulated in a manner that is first provided when recommending the image metadata as described above.

Next, a process of searching for image semantic metadata using the image semantic metadata database 340 constructed as described above will be described.

First, as described above, when a search user inputs a keyword to be searched into the image metadata recommendation unit 310, the image metadata recommendation unit 310 searches the image semantic metadata database 340 using the corresponding keyword. If the searched image is recommended (provided) as a search result along with the searched image semantic metadata, or if a search user inputs a keyword to be searched by a known search system (not shown), the corresponding search system is applied. The image semantic metadata database 340 is searched by a keyword, and the image semantic metadata corresponding to the searched image is provided as a search result.

Accordingly, when a search user selects (clicks) a specific image from the image and image semantic metadata received from the image metadata recommending unit 310 or a known search system, or additionally selects specific image semantic metadata, the selection is made accordingly. The information is fed back to the selection information receiver 320 of the image semantic metadata generation server 30.

Then, the selection information receiver 320 receives the image and the image semantic metadata along with the selection information according to the search result selection of the search user, and transmits the image and the image semantic metadata to the image semantic metadata generator 330.

Then, the image semantic metadata generator 330 generates a new image semantic metadata for the image by adjusting the ranking of the image semantic metadata based on the selection information received through the selection information receiver 320. To update the image semantic metadata database 340. This retrieval and update process continues to cycle.

3 is a detailed configuration diagram of the image metadata generating unit of FIG. 2 according to an exemplary embodiment of the present invention. The image metadata generating unit 230 (parsing parser) generates initial image metadata based on an article. The detailed configuration is shown.

First, the first metadata extractor 301 extracts image metadata from a title of a corresponding article based on a classification category of a web article temporarily stored in the web document storage 220.

Here, the first metadata extractor 301 extracts the image metadata by comparing the article title term with the terminology dictionary 250 or / and ontology 240 for each category of the category to which the article belongs, or the photo description term. The image metadata is extracted by comparing the terminology terminology dictionary 250 and / or ontology 240 of the category of the article to which the article belongs. Of course, the first metadata extractor 301 may extract the image metadata using the article title term and then extract the image metadata using the photo description term. The priority of the image metadata derived in this way has priority over the image metadata derived from the article title.

The second metadata extractor 302 extracts image metadata from the content of the article based on the classification category of the web article.

Here, the second metadata extractor 302 extracts a semantic unit by separating article contents of a web article using a known N * GRAM indexing technique, and frequently appears in other articles among the extracted semantic units terms. Therefore, after removing weak terms or terms unnecessary for comparison, the terms of the semantic units remaining after the removal are removed from the terminology dictionary 250 or / or ontology 240 of the category to which the article belongs, and known LSI (Latent Semantic). Indexing) to compare and extract the metadata of the image.

The image metadata generator 303 integrates the image metadata extracted by the first and second metadata extractors 301 and 302 and generates a mutual relationship between the extracted image metadata.

Here, as an example, the relationship between the image metadata, the highest priority of the image metadata extracted from the article title, and then the highest priority of the image metadata extracted from the photo description, and finally the article The image metadata extracted by comparing the frequency and term in the content itself has the lowest priority.

If the same or similar image metadata is derived from the metadata priorities for the image, it is integrated into representative terms based on the domain ontology, and the terms of the category, title based image metadata, description based image metadata, and article based Configure the image metadata to form a parent / child relationship. For example, image metadata including hierarchies in the entertainment section-> entertainer name-> entertainer rumor, etc. is regarded as initial image metadata.

4 is a flowchart illustrating a method of generating semantic metadata of an article-based image and a method of providing semantic metadata of an article-based image using the same according to an exemplary embodiment of the present invention. Only the operation summary will be briefly described.

First, the image metadata generation server 20 generates initial image metadata by using the description information of the image (401). This will be described later in detail with reference to FIG. 5.

Subsequently, the image semantic metadata generation server 30 recommends the image and the image metadata related thereto to the search user, receives feedback of the search user's selection information, and generates semantic metadata for the corresponding image ( 402). This will be described later in detail with reference to FIG. 6.

Next, a well-known search system and the image semantic metadata generating server 30 search for and provide an image corresponding to a keyword input from a search user and image metadata related thereto, and feed back selection information of the search user. In operation 403, the semantic metadata of the corresponding image is updated. Since it has been described above in detail with reference to Figure 2 will not be described herein any further.

FIG. 5 is a detailed flowchart of the image metadata generating process of FIG. 4 according to an exemplary embodiment of the present invention. Since the specific embodiment is the same as described above, only an operation summary thereof will be briefly described.

First, in operation 510, the first metadata extractor 301 extracts image metadata from a title of an article based on a classification category of the article.

That is, the first metadata extractor 301 extracts image metadata by comparing the article title term with the terminology dictionary 250 or the ontology 240 for each category of the category in which the article belongs. In operation 522, the first metadata extractor 301 extracts image metadata by comparing the photographic description term with the terminology dictionary 250 or / or ontology 240 for each category of the category to which the article belongs.

Thereafter, the second metadata extractor 302 extracts image metadata from the content of the article based on the classification category of the article (520).

That is, the second metadata extractor 302 extracts the semantic unit by separating the article content of the article (521), and appears frequently in other articles among the terms of the extracted semantic unit, which is unnecessary for terms having weak features or comparisons. After the term is removed (522), the metadata of the image is extracted by comparing the term of the remaining semantic unit with the terminology dictionary 250 or / and ontology 240 of the category to which the article belongs (523).

Thereafter, the image metadata generator 303 integrates the extracted image metadata using the article title and the article content and generates a mutual relationship between the image metadata (530).

That is, when the image metadata generator 303 derives the same or similar image metadata in addition to the priority of the metadata for the image, the image metadata generator 303 integrates the representative term based on the domain ontology (531), and based on the term and title of the category. The image metadata, the description-based image metadata, and the article-based image metadata are generated to form a parent / child relationship (532).

FIG. 6 is a detailed flowchart illustrating the process of generating the image semantic metadata of FIG. 4 according to an exemplary embodiment of the present invention. Since the specific embodiment is the same as described above, only the operation summary will be briefly described.

First, the image metadata recommendation unit 310 searches for an image corresponding to a keyword input from a search user and image metadata related thereto and recommends the image metadata to the corresponding search user (610). In this case, when a search user inputs a specific keyword, the search user checks whether the matching image metadata is included and provides all matching image metadata to the search user along with the corresponding image. Here, the image metadata recommendation unit 310 provides the hierarchical image metadata as well as the most frequent term image metadata (the least term image metadata in consideration of the specific term frequency in the article content) with the image.

Accordingly, when the search user selects (clicks) a specific image from the image and image metadata recommended by the image metadata recommender 310 or additionally selects an appropriate specific image metadata, the selection information according to the image semantic metadata is selected. It is fed back to the selection information receiver 320 of the generation server 30.

That is, the search user who searches for the image metadata and the image may select not only the image closest to the searched term but also among the listed image metadata. This selection process is regarded as semantic metadata for a specific image determined by the search user, and the selected values (selection information) are fed back to the selection information receiver 320 as an evaluation result along with the image and the image metadata. .

Then, the selection information receiver 320 receives the corresponding image and the image metadata along with the selection information according to the search result selection of the search user (620).

Thereafter, the image semantic metadata generator 330 generates image semantic metadata for the corresponding image based on the received selection information and the image metadata and stores the image semantic metadata in the image semantic metadata database 340 (630).

That is, the image semantic metadata generator 330 adjusts the ranking of the corresponding image metadata based on the received selection information to generate image semantic metadata for the corresponding image to the image semantic metadata database 340. Save it.

In the present invention, as described above, by automating the generation of semantic metadata for an image, it is expected to add value in real life than a tag input method which is a method directly input by an existing user.

On the other hand, the method of the present invention as described above can be written in a computer program. And the code and code segments constituting the program can be easily inferred by a computer programmer in the art. In addition, the written program is stored in a computer-readable recording medium (information storage medium), and read and executed by a computer to implement the method of the present invention. The recording medium may include any type of computer readable recording medium.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the technical spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited by the drawings.

The present invention may be used for a metadata recommendation service for an article-based image, an image semantic metadata-based search system, an image classification system, or a browsing search system.

1 is an embodiment configuration diagram of an existing system for recording a conventional image and registering with a web article and storing the image;

2 is a block diagram of a device for generating semantic metadata of an article-based image and a system for providing semantic metadata of an article-based image using the same according to an embodiment of the present invention;

3 is a detailed configuration diagram of the image metadata generator of FIG. 2 according to an embodiment of the present invention;

4 is a flowchart illustrating a method of generating semantic metadata of an article-based image and a method of providing semantic metadata of an article-based image using the same according to an embodiment of the present invention;

5 is a detailed flowchart illustrating an image metadata generation process of FIG. 4 according to an embodiment of the present invention;

6 is a detailed flowchart illustrating a process of generating image semantic metadata of FIG. 4 according to an embodiment of the present invention.

* Explanation of symbols for the main parts of the drawings

20: image metadata generation server

30: Image Semantic Metadata Generation Server

210: Web article collector 220: Web document repository

230: image metadata generation unit 240: ontology

250: terminology dictionary for each field

310: image metadata recommendation unit 320: selection information receiver

330: image semantic metadata generator

340: Image Semantic Metadata Database

301: first metadata extractor 302: second metadata extractor

303: Image Metadata Generator

Claims (29)

In the image semantic metadata generating device, Image metadata generating means for generating image metadata using peripheral description information of the image; And Image semantic metadata generating means for recommending an image and corresponding image metadata to a search user, and generating semantic metadata for the corresponding image by receiving feedback of the search user's selection information; Image semantic metadata generating device comprising a. The method of claim 1, The image metadata generating means, An image semantic metadata generating device, comprising: generating initial image metadata by using peripheral description information of an article-based image and transferring the initial image metadata to the image semantic metadata generating means together with the corresponding image; The method of claim 2, The image metadata generating means, A first metadata extractor for extracting image metadata from the title of the article; A second metadata extractor for extracting image metadata from the content of the article; And An image metadata generator for integrating image metadata extracted by the first and second metadata extractors and generating a mutual relationship Image semantic metadata generating device comprising a. The method of claim 3, wherein The first metadata extractor, An image semantic metadata generating device extracting image metadata using an article title term or extracting image metadata using a photo description term. The method of claim 3, wherein The first metadata extractor, An image semantic metadata generating device, comprising: extracting image metadata using article title terms and extracting image metadata using photo description terms. The method of claim 3, wherein The second metadata extractor, An image semantic metadata generating device, comprising: extracting a semantic unit by separating article contents, removing a stopword from the terms of the extracted semantic unit, and extracting image metadata using the remaining semantic unit terms . The method of claim 3, wherein The image metadata generator, And image metadata extracted by the first and second metadata extractors as representative terms, and generate a mutual relationship such that the extracted image metadata forms a hierarchical structure. The method of claim 7, wherein The image metadata generator, Set the priority of the image metadata extracted from the article title to the highest priority, set the priority of the image metadata extracted from the photo description to the next priority, and prioritize the image metadata extracted from the article content to the next priority. And image semantic metadata generating device. The method according to any one of claims 1 to 8, The image semantic metadata generating means, An image metadata recommending unit for searching for and recommending an image corresponding to a keyword input from the search user and corresponding image metadata; A selection information receiver configured to receive selection information according to the search result selection of the search user, corresponding image and image metadata; An image semantic metadata generating unit for generating image semantic metadata for the corresponding image based on the selection information and the image metadata received through the selection information receiving unit; And Storage means for storing the image semantic metadata generated by the image semantic metadata generator Image semantic metadata generating device comprising a. The method of claim 9, The image metadata recommendation unit, An image inputted by a search user, the image metadata received from the image metadata generating means and the image metadata stored in the storage means are searched to recommend matching image metadata and the corresponding image Semantic metadata generation device. The method of claim 9, The image semantic metadata generator, And image semantic metadata for the image by adjusting the ranking of the image metadata based on the selection information received through the selection information receiver. The method of claim 11, wherein The image semantic metadata generator, After identifying which image is selected based on the selection information received through the selection information receiving unit, image semantic meta is adjusted by assigning a weight to the image metadata for the selected image. Data generation device. The method of claim 11, wherein The image semantic metadata generator, After checking which image and which image metadata are selected based on the selection information received through the selection information receiving unit, weighting is applied to the image metadata for the selected image to adjust the image metadata ranking, and And image weight ranking is adjusted by additionally weighting the selected image metadata. In the image semantic metadata providing system, Image metadata generating means for generating image metadata using peripheral description information of the image; Image semantic metadata generating means for recommending an image and corresponding image metadata to a search user and generating and storing semantic metadata for the corresponding image by receiving feedback of the selection information of the search user; And A search and update for searching and providing an image corresponding to the keyword input from the search user and the corresponding image metadata, and receiving the feedback of the search user's selection information to update the stored image semantic metadata Way Image semantic metadata providing system comprising a. The method of claim 14, The search and update means, Means for searching for and providing an image corresponding to a keyword input from the search user and corresponding image semantic metadata; Means for receiving selection information according to the search result selection of the search user and corresponding image and image semantic metadata; And Means for generating image semantic metadata for the corresponding image based on the received selection information and the image semantic metadata to update previously stored image semantic metadata Image semantic metadata providing system comprising a. The method according to claim 14 or 15, The image metadata generating means, A first metadata extractor for extracting image metadata from the title of the article; A second metadata extractor for extracting image metadata from the content of the article; And An image metadata generator for integrating image metadata extracted by the first and second metadata extractors and generating a mutual relationship Image semantic metadata providing system comprising a. In the image semantic metadata generation method, A first image metadata generating step of generating image metadata using peripheral description information of the image; And A first image semantic metadata generation step of recommending an image and corresponding image metadata to a search user and generating semantic metadata for the corresponding image by receiving feedback of the search user's selection information; Image semantic metadata generation method comprising a. The method of claim 17, The first image metadata generating step, A first metadata extraction step of extracting image metadata from a title of the article; A second metadata extracting step of extracting image metadata from the content of the article; And A second image metadata generating step of integrating the image metadata extracted in the first and second metadata extracting steps and generating a mutual relationship; Image semantic metadata generation method comprising a. The method of claim 18, The first metadata extraction step, An image semantic metadata generation method comprising extracting image metadata using an article title term or extracting image metadata using a photo description term. The method of claim 18, The first metadata extraction step, Extracting image metadata using article title terms; And Extracting Image Metadata Using Photo Descriptive Terms Image semantic metadata generation method comprising a. The method of claim 18, The second metadata extraction step, Extracting a semantic unit by separating article contents; Removing stop words from terms of the extracted semantic units; And Extracting image metadata by using the terms of the removed semantic unit Image semantic metadata generation method comprising a. The method of claim 18, The second image metadata generating step, And integrating the image metadata extracted in the first and second metadata extraction steps into representative terms, and generating a mutual relationship such that the extracted image metadata has a hierarchical structure. The method of claim 22, The second image metadata generating step, Set the priority of the image metadata extracted from the article title to the highest priority, set the priority of the image metadata extracted from the photo description to the next priority, and prioritize the image metadata extracted from the article content to the next priority. And image semantic metadata generation method. The method according to any one of claims 17 to 23, The generating of the first image semantic metadata includes: Searching for and recommending an image corresponding to a keyword input from the search user and corresponding image metadata; Receiving feedback on selection information according to the search result selection of the search user, corresponding image and image metadata; A second image semantic metadata generating step of generating image semantic metadata for the corresponding image based on the feedback selection information and the image metadata; And Storing the generated image semantic metadata Image semantic metadata generation method comprising a. The method of claim 24, The second image semantic metadata generation step, And generating image semantic metadata for the corresponding image by adjusting a ranking of the corresponding image metadata based on the feedback selection information. In the method of providing image semantic metadata, An image metadata generation step of generating image metadata using peripheral description information of the image; An image semantic metadata generation step of recommending an image and corresponding image metadata to a search user, and generating and storing semantic metadata of the corresponding image by receiving feedback of the selection information of the search user; And A search and update step of searching for and providing an image corresponding to a keyword input from the search user and corresponding image metadata, and updating the stored image semantic metadata by receiving feedback of the search user's selection information; Image semantic metadata providing method comprising a. The method of claim 26, The searching and updating step, Searching for and providing an image corresponding to a keyword input from the search user and corresponding image semantic metadata; Receiving feedback of selection information according to a search result selection of the search user, corresponding image, and image semantic metadata; And Updating image pre-stored image semantic metadata by generating image semantic metadata for the corresponding image based on the feedback selection information and the image semantic metadata Image semantic metadata providing method comprising a. In the image semantic metadata generating device having a processor, An image metadata generation function for generating image metadata using peripheral description information of the image; And Image semantic metadata generation function that recommends an image and corresponding image metadata to a search user and generates semantic metadata for the corresponding image by receiving feedback of the search user's selection information. A computer-readable recording medium having recorded thereon a program for realizing this. In the image semantic metadata providing system having a processor, An image metadata generation function for generating image metadata using peripheral description information of the image; An image semantic metadata generation function for recommending an image and corresponding image metadata to a search user and generating and storing semantic metadata for the corresponding image by receiving feedback of the search user's selection information; And A search and update function for searching and providing an image corresponding to a keyword input from the search user and corresponding image metadata, and updating the stored image semantic metadata by receiving feedback of the search user's selection information A computer-readable recording medium having recorded thereon a program for realizing this.

Images