JP2011108242A - Method and system for dynamically extracting and providing most suitable image according to user's request, and computer-readable recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and a system for dynamically extracting and providing a most suitable image according to an inquiry of a user. <P>SOLUTION: The method for dynamically extracting and providing the most suitable image according to an inquiry of a user includes steps of: receiving an inquiry for an image to be searched by the user from a terminal device of the user; searching an image group related to the received inquiry; dynamically determining perfect matching of one or more images in the searched image group with the received inquiry, and extracting, when perfect matching is confirmed, the matched image; and providing the extracted image to the terminal device of the user. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

The present invention relates to a method, a system, and a computer-readable recording medium for dynamically extracting and providing a most suitable image in response to an inquiry from a user. More specifically, when providing an image that matches the requirements entered by the user, calculate the relationship with each image group, select the group, and calculate the relevance to the user requirements. The present invention relates to a method, a system, and a computer-readable recording medium for extracting an optimal image for each group from images corresponding to each group and providing them to a user.

Recently, as the Internet has rapidly spread, the Internet has become a must in modern life. Most users who use the Internet connect to the Internet through a browser and then search for information through an Internet portal site. When the user inputs a search term in the search window of the Internet portal, the Internet portal extracts various information corresponding to the search term from the database and provides it to the user as a search result. The extracted information is displayed to the user in categories such as dictionary, knowledge information, blog, online community, specialized material, site, book, news, web page, image and video.

Korean Registered Patent No. 0706389 Korean Published Patent No. 2009-048812 US Pat. No. 7,099,860

Of these, in the case of images, there are many similar images, and users prefer search results for a variety of images. Therefore, through duplication judgment, only the representative image of the extracted images is left, and other images are stored. It is preferable to remove and communicate to the user. However, if all images are processed in real time when such an image search operation is requested, the cost of the disk and the network increases and the search time also increases, thereby obtaining satisfactory results. I can't.

Therefore, after the images are grouped and stored together, when the user actually inputs a search term, there is a method of quickly extracting the corresponding images from the grouped images. However, there is a problem that it is difficult to group images together because the resolution and size are generally different even when they are the same image.

As an example of a solution to such a problem, after extracting a thumbnail (thumbnail; a reduced image of a certain size and resolution) for each image, a signature value is obtained from such a thumbnail and obtained. A method of comparing the obtained signature values and determining whether the images to be compared are the same image can be considered.

However, even if this method is used, only one specific image is selected as a representative image for each image group, and the search term and title entered by the user are related to other images in the group. If the image is not related to the representative image, an unrelated representative image is presented to the user. Also, when comparing images, the signature values extracted from the thumbnails of the images are compared, so if the images have different patterns even if the images are different, the signature values will match and it will be mistaken as the same image. The case where the judgment is made cannot be excluded. For this reason, in this case, there is a problem that an image completely different from the image searched for by the user may be displayed.

One of the objects of the present invention is to solve these problems.

Another object of the present invention is to dynamically extract and provide a most suitable image dynamically according to a user's inquiry in a preset image group.

Another object of the present invention is to arrange images in the order of reliability within a group, compress image information within the group, and search space for extracting and providing an image that most closely matches the user's query. And minimizing search time.

A characteristic configuration of the present invention for achieving the object of the present invention as described above and achieving a specific effect of the present invention described later is as follows.

According to an embodiment of the present invention, in a method for providing an image most suitable for an inquiry about a user's image search on a communication network, an inquiry about an image that the user wants to search is made from the user's terminal device. Receiving, searching for an image group associated with the received query, and dynamically determining whether a perfect match between one or more images in the searched image group and the received query is possible. A method is provided that includes extracting a matched image when complete matching is confirmed, and providing the extracted image to the user terminal.

According to another embodiment of the present invention, in a system for providing an image most suitable for an inquiry about a user's image search on a communication network, an inquiry about an image that the user intends to search from a user terminal device. A means for retrieving an image group associated with the received query, and one or more images in the retrieved image group and a perfect match of the received query. Image search having means for dynamically determining whether or not perfect matching is confirmed and extracting a matched image when the perfect matching is confirmed, and means for providing the extracted image to the terminal device of the user A system including the unit is provided.

According to the present invention, it is possible to provide a user who searches for an image without duplicating an image extracted from the most accurate source in response to the user's inquiry.

Further, according to the present invention, the most suitable image can be dynamically extracted and provided from a preset image group in response to a user's inquiry.

Further, according to the present invention, it is possible to reduce the possibility that an image that is not related to an inquiry is provided to the user as a representative image by having the same signature value even though they are different images.

Furthermore, according to the present invention, images are arranged and compressed in the order of reliability within one group, so that the image search memory and the search speed can be reduced.

1 is a diagram schematically illustrating an overall configuration of a system for dynamically extracting and providing a most suitable image based on a user query according to an embodiment of the present invention; 3 is a detailed diagram illustrating an internal configuration of an image server according to an exemplary embodiment of the present invention. 4 is a flowchart illustrating a series of processes for selecting a representative image in an image grouping unit including a representative image selection unit according to an embodiment of the present invention. A series of processes in which an image search unit according to an embodiment of the present invention dynamically searches for a representative image that most closely matches an inquiry received from a user terminal device and provides the search result to the user terminal device. It is a flowchart shown exemplarily. 5 is a diagram illustrating a web page including an image search result provided to a terminal device of a user according to the related art and the present invention. 5 is a diagram illustrating a web page including an image search result provided to a terminal device of a user according to the related art and the present invention.

The present invention will now be described more fully hereinafter with reference to the accompanying drawings, which illustrate, by way of illustration, specific embodiments for carrying out the invention. In addition, these embodiments are described in detail so that those skilled in the art can fully practice the present invention. While various embodiments of the invention are different, it should be understood that they need not be mutually exclusive. For example, the specific shapes, structures, and characteristics described below can be embodied in other embodiments without departing from the spirit and scope of the present invention. It should also be understood that the position or arrangement of the individual components within each disclosed embodiment can be altered without departing from the spirit and scope of the present invention. Accordingly, the detailed description set forth below should not be construed in a limiting sense. The technical scope of the present invention is limited only by the appended claims, to the full extent equivalent to that claimed. Also, in the drawings, like reference numerals indicate identical or similar functions across various aspects.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily practice the present invention. .

[Preferred embodiment of the present invention]
Overall System Configuration FIG. 1 is a diagram schematically showing an overall configuration of a system for dynamically extracting and providing a most suitable image based on a user's inquiry according to an embodiment of the present invention.

As shown in FIG. 1, an overall system according to an embodiment of the present invention includes a communication network 100, an image server that groups and stores images and dynamically extracts an image most suitable for a user's inquiry. 200, a plurality of web servers 400 that search and provide content (including images extracted from the image server 200) in response to user inquiries received from the user terminal device 300 and the user terminal device 300. Can be configured.

First, the communication network 100 can be configured regardless of the communication mode such as wired and wireless. For example, it can be composed of various communication networks such as a local area network (LAN), a metropolitan area network (MAN), and a wide area network (WAN). Preferably, the network of the present invention can provide a known World Wide Web (WWW).

According to an embodiment of the present invention, the image server 200 can search for an image suitable for a query input by a user, and can quickly search for images corresponding to the search target thumbnails having a certain size and resolution. It is possible to provide a function of grouping images having the same signature value by comparing the signature values extracted from the thumbnails after the conversion.

In addition, according to an embodiment of the present invention, the image server 200 may include a title, a file name, and a source of an image (referred to as a source when the corresponding image is searched). Sources include, but are not limited to, news, blogs, online communities, dictionaries, etc.), information such as resolution, image registration date, etc., extracted and processed into description information. It is possible to provide a function of assigning a specific weight value to each piece of information and calculating and storing the reliability of each image. In this case, the image with the highest reliability can be designated as the representative image of the corresponding image group.

Further, according to an embodiment of the present invention, the image server 200 receives the inquiry about the user's image search from the user's terminal device 300 or the web server 400 via the communication network 100, and the description information and / or Alternatively, it is possible to provide a function of searching for and providing the most suitable image by determining whether or not a match with the user's inquiry can be made using the reliability. In this case, it is possible to search for the most suitable image through various methods such as the resolution preferred by the user and the source preferred by the user, instead of whether or not the user's inquiry matches.

On the other hand, a user terminal device 300 according to an embodiment of the present invention is a digital device including a function that enables a user to connect to the web server 400 via the communication network 100 to enable communication. For example, it may be a digital device equipped with memory means, equipped with a microprocessor, and equipped with computing power, such as a personal computer (eg, desktop computer, notebook computer, etc.), workstation, PDA, web pad, mobile phone, etc. Thus, it can be used as the user terminal device 300 according to the present invention. The user's terminal device 300 may further include a web browser (not shown) program that allows the user to request content from the web server 400 and receive the content.

According to an embodiment of the present invention, the web server 400 performs communication with the image server 200 and the user terminal device 300 via the communication network 100. For example, when the web server 400 is an Internet search portal site management server, the web server 400 includes a web content search engine (not shown), stores the query keyword input by the user, and then stores the query keyword in the query keyword. The corresponding contents can be searched and the search result can be provided so that the user can browse. The content provided as a search result can be various information related to news, knowledge, games, communities, and the like. Of course, the web content search engine may be physically included in another computing device or recording medium instead of the web server 400 as necessary. FIG. 1 shows that the image server 200 and the web server 400 are separately configured. However, if the image search service is included in the search service for content according to the needs of those skilled in the art embodying the present invention, the image server 200 may be included in the web content search engine, or the image server 200 may be The web server 400 may be included in the configuration.

Image Server Configuration Hereinafter, an internal configuration of the image server 200 that provides important functions for implementing the present invention and functions of each component will be described.

FIG. 2 is a detailed diagram illustrating an internal configuration of the image server 200 according to an embodiment of the present invention. Referring to FIG. 2, the image server 200 according to an embodiment of the present invention may include an image grouping unit 210, an image search unit 220, a communication unit 230, a control unit 240 and a database 250. Also, FIG. 2 shows that the image grouping unit 210 can be configured by an image processing unit 212, an image comparison unit 214, a representative image selection unit 216, and the like. According to one embodiment of the present invention, the image grouping unit 210, the image search unit 220, the communication unit 230, the control unit 240, and the database 250 are at least partly a web server 400, a web content search engine (not shown), It may be a program module that communicates with at least one of the user terminal devices 300. Such a program module may be included in the image server 200 in the form of an operating system, an application program module, and other program modules, and can be physically stored on various known storage devices. Such a program module can also be stored in a remote storage device that can communicate with the image server 200. On the other hand, such a program module includes a routine, a subroutine, a program, an object, a component, a data structure, and the like for executing a specific task and executing a specific abstract data type described later according to the present invention. Not.

First, the image grouping unit 210 according to an embodiment of the present invention can provide a function of generating thumbnails, signatures, and description data for all images that can be searched by the user. Further, it is possible to further provide a grouping function for analyzing the generated data and selecting a representative image by grouping the same image into one group. Hereinafter, the internal configuration of the image grouping unit 210 and the function of each component will be described with reference to FIG.

More specifically, the image grouping unit 210 of the present invention can include an image processing unit 212, an image comparison unit 214, and a representative image selection unit 216.

The image processing unit 212 according to an embodiment of the present invention searches for images that can be searched by the user from their respective sources, generates thumbnails and signatures for the corresponding images, and generates the generated data in a database. The function to store in 250 is provided.

The image comparison unit 214 according to an embodiment of the present invention compares the signature values of the images stored in the database 250, collects images having the same signature value into respective groups, and stores related information in the database 250. Execute grouping function. Further, the image comparison unit 214 can provide a function of reading the description data of each image and storing it in the database 250. In addition, the image comparison unit 214 can provide a function of calculating the reliability using the description data of each image.

The representative image selection unit 216 according to an exemplary embodiment of the present invention searches for an image most likely to be searched by the user using the description data and / or reliability generated by the image comparison unit 214. After selecting as a representative image of the image group, a function of storing this in the database 250 is provided.

However, the image processing unit 212, the image comparison unit 214, and the representative image selection unit 216 exemplified above do not provide functions by being limited to the above-described examples. For example, the image processing unit 212 provides the functions provided by the image comparison unit 214. Those who have ordinary knowledge in the field to which the present invention belongs, such as a part can be provided by the image processing unit 212 or the representative image selection unit 216, and vice versa, can be designed in various ways. Can do.

Next, when the image search unit 220 according to an embodiment of the present invention receives an inquiry about a user's image search, the image search unit 220 searches for an image group stored in the database 250 and relates to the user's inquiry. It is possible to provide a function of dynamically retrieving and providing the most suitable image from the image group and the corresponding image group.

In addition, the communication unit 230 according to an embodiment of the present invention has a function of enabling the image server 200 to communicate with devices such as the user terminal device 300 and the web server 400 via the communication network 100. provide.

In addition, the controller 240 according to an embodiment of the present invention provides a function of controlling the data flow among the image grouping unit 210, the image search unit 220, the communication unit 230, and the database 250.

Hereinafter, a specific process of selecting a representative image through the representative image selection unit 216 according to an embodiment of the present invention will be described.

FIG. 3 is a flowchart illustrating a series of processes for selecting a representative image by the image grouping unit 210 including the representative image selection unit 216 according to an embodiment of the present invention.

Referring to FIG. 3, in one embodiment of the present invention, the image comparison unit 214 compares signature values stored in the database 250 and collects images having the same signature value into one group (S310). As described above, the signature value is generated based on the thumbnail generated by the image processing unit 212 in one embodiment of the present invention.

Next, the image comparison unit 214 generates description data from the images grouped into one group (S320). Here, the description data may include, but is not limited to, an image title, a file name, an original source of the image, a corresponding URL (Uniform Resource Locator), a resolution, a registration date, a creator, and the like. In addition, a part or all of the description data of each image extracted for comparison with a user search query is stored in the database 250 as image group description data, excluding duplicate parts. be able to.

Next, the image comparison unit 214 calculates the reliability of the images in the group using the generated description data (S330). For example, when the source is highly reliable, such as news or a dictionary, the reliability can be calculated by assigning a high weight value according to the source. Also, when the reliability of an image is low, such as an individual blog or an online community, the reliability can be calculated by assigning a low weight value according to the origin. In addition, when each of the morphemes extracted by parsing the image title and file name in the description data, or an expression including the morpheme, causes duplicates in the description data, the image title of the corresponding image. The reliability can also be calculated by giving a higher weight to the file name. There is also a method of assigning a high weight value according to the resolution of an image having a high resolution, or assigning a high weight value according to the image registration date of the latest image most recently registered based on the image registration date. Can be used. Therefore, the final reliability can be indicated by a total value of values obtained by assigning weight values to the description data of each image.

However, the calculation method is only an example. On the other hand, the reliability can be calculated by combining the various conditions described above or by using other various methods. Further, when the description data for a plurality of images is the same, it can be determined that the same image has been searched redundantly, and the image determined to be redundant can be removed from the group.

Based on the reliability calculated by the above method, the representative image selection unit 216 can arrange images in the order of reliability in each image group, and can select the image with the highest reliability as the representative image. (S340). In addition, for example, when an image in an image group and description data for the image are arranged one-dimensionally like a list, weighting is performed so that the reliability of adjacent images on the list is the same or almost similar, for example. A value may be assigned, and the difference between adjacent description data may be increased. That is, for example, on the list, the weights can be assigned so that the reliability of the image having the description data having the largest difference from the image having the previous description data is the same or almost similar. As a result, the images can be arranged so that various inquiries can be processed within a short time by making the storage positions of the images having completely different description data close to each other when the images are arranged thereafter. This is because if the description data is different, the images corresponding to the description data are often different even if the reliability is the same or almost similar, so different images can be arranged on the list so as to be adjacent to each other. It is. Each image group in which an image is arranged can compress image information in the group in order to execute a process of dynamically determining and providing a representative image in response to a user inquiry in real time. In this case, if processing can be performed in the main memory, the amount of memory used can be reduced, and the search time can be shortened.

As described above, FIG. 3 is a flowchart showing a representative image selection process according to an embodiment of the present invention. According to the necessity or selection of a person having ordinary knowledge in the technical field to which the present invention belongs, the generation of the description in step S320 is extracted in advance by the image processing unit 212 and stored in the database 250 before the execution of step S310. It can be substituted for the step of extracting description data. If the image server finds a new image by performing a search, a representative image selection procedure is performed for the new image, and the signature value of the new image is obtained in step S310 and has already been generated. If an image group has the same signature value, it is assigned to that image group. If there is no image group having the same signature value, the image group is transformed into a new image group generation stage. It can also be implemented as an embodiment. The description data item includes a description data item for the image to be frequently searched (by user's inquiry or selection by clicking), and when the reliability is calculated in step S330. Various modifications and changes are possible, such as an implementation method in which a weight value is assigned to and the reliability is recalculated at a fixed period.

Hereinafter, a method for dynamically determining and providing a representative image according to a user inquiry will be described in detail.

FIG. 4 shows that the image search unit 220 according to an embodiment of the present invention dynamically searches for a representative image that most closely matches the query received from the user's terminal device 300, and uses the search result as the user's search result. 4 is a flowchart exemplifying a series of processes provided to a terminal device 300. Prior to FIG. 4, the image search unit 220 parses the query if the query includes a plurality of words after receiving a query related to the user's image search from the user terminal device 300 or through the web server 400. To analyze and transform into meaningful morphemes, or combined forms of morphemes. Such parsing is performed in advance through the web server 400, and the image search unit may perform only a function of receiving an inquiry regarding the image search in the parsed form. Hereinafter, a form that has undergone transformation through parsing, or an inquiry itself is referred to as an “inquiry about image search”. Further, the inquiry about the image search is compared with the description data of each image group stored in the database 250 to determine whether or not the image group is related to the inquiry about the image search of the user. In this case, the representative image search is dynamically started within the corresponding image group.

First, the image search unit 220 determines whether each element of the description data of the representative image selected in advance and the inquiry related to the image search completely match each other. If the representative image is completely matched, the process moves to step S460 (S410).

If it is determined in step S410 that complete matching is not performed, it is determined whether or not each element of the description data and the query related to image search are completely matched in the order of images arranged in the order of reliability. If there is a perfect matching image, the process moves to step S460 (S420).

If there is no image having a description data element that perfectly matches the query related to the image search through the steps S410 and S420, whether or not each element of the query related to the image search and the description data of the representative image partially matches. to decide. In the case of partial matching, the process moves to step S460 (S430). Here, partial matching refers to whether or not each element (image title, file name, etc.) of the description data of the representative image includes a part of the meaningful morpheme in the inquiry about the parsed image search. Can mean.

If partial matching with the representative image is not performed in step S430, it is determined whether partial matching is possible in the order of reliability for the remaining images in the image group. If there is an image to be partially matched, the process proceeds to step S460 (S440), and if it does not exist, a representative image set to the existing is provided (S450). Alternatively, in step S450, an image may not be provided, or content indicating that there is no matching image may be provided.

In step S460, an image that is matched through steps S410, S420, S430, or S440 is provided. The image provision in steps S450 and S460 may mean that the image information is provided to the user's terminal device 300 via the web server 400 indirectly or directly via the communication network 100. The image information may basically include a URL of an image to be matched, a thumbnail, and source information, but is not necessarily limited thereto.

In FIG. 4, in order to dynamically determine an image, an example of a method of comparing an inquiry about a user's image search with an image title is used. However, the method is not necessarily limited to such a method, and is based on the method that compares the resolution that the user likes most with the resolution of the image or the source that the user likes (for example, if it likes news, it is extracted from the news) Various modifications and variations can be made by a person who has ordinary knowledge in the field to which the present invention belongs, such as a method of searching for a target image) or a combination of such methods.

FIG. 4 illustrates an embodiment in which representative images are dynamically extracted and provided within one image group. However, when there are a large number of image groups related to an inquiry related to image search, the above procedure may be repeated for a large number of groups to provide the user with all or part of the large number of extracted images. it can.

5A and 5B exemplarily illustrate web pages including image search results provided to a user terminal device 300 according to the related art and the present invention.

Referring to FIG. 5a, it can be seen that the three images on the left are images grouped into the same image group having the same signature, and the middle image is set as the representative image. According to the prior art, once the representative image is determined in this way, the same representative image is provided regardless of the inquiry related to the user's image search. An image is provided to the user (upper right corner). On the other hand, according to an embodiment of the present invention, it can be seen that even the same image is dynamically extracted and provided to the user having an image having the most suitable source for an inquiry regarding the user's image search ( Bottom right)

Also, referring to FIG. 5b, the three images on the left side are combined into one image group in order to recognize the same image by having the same signature value even though they are different images. It can be seen that the lowermost image is set as the representative image. According to the prior art, if different images are actually grouped into the same image group due to the same signature value, a completely different image may be provided to the user by an inquiry regarding the image search of the user Is (upper right corner). On the other hand, according to an embodiment of the present invention, even when such completely different images are included in the same image group, an image most suitable for a user's inquiry about image search can be provided. I understand (bottom right side).

Embodiments according to the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the present invention, or those known and usable by those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, flexible disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and floppy disks (Flopical disks). Hardware devices specially configured to store and execute program instructions such as magneto-optical media such as disk and ROM, RAM, flash memory, etc. Is included. Examples of program instructions include not only machine language code such as that produced by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above can be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

In the present invention as described above, specific matters such as specific components and the like, and limited embodiments and drawings have been described. However, this is only for the purpose of facilitating a more general understanding of the present invention, and the present invention is not limited to the above-described embodiments, but has ordinary knowledge in the field to which the present invention belongs. If so, various modifications and variations are possible from such description.

Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, but includes not only the claims appended to the application, but also all equivalent or equivalent modifications to the claims. Should belong to the category of the idea of the present invention.

100: communication network 200: image server 210: image grouping unit 212: image processing unit 214: image comparison unit 216: representative image selection unit 220: image search unit 230: communication unit 240: control unit 250: database

Claims (16)

In the method of providing the most suitable image for inquiries regarding user image search,
A first step of receiving an inquiry about an image to be searched by the user from a user terminal device by an image search unit;
A second step of searching for an image group related to the received inquiry by the image search unit;
The image search unit dynamically determines whether or not complete matching between one or more images in the searched image group and the received inquiry is possible, and when the complete matching is confirmed, the matched image And a fourth step of providing the extracted image to the user terminal device by the image search unit;
Including methods. The method of claim 1, wherein the first step further comprises transforming the received query into a meaningful morpheme or a combined form of morphemes through parsing. The method according to claim 1, wherein in the perfect matching, at least one of elements of description data included in the image and the entire received query is matched. The third step determines whether or not partial matching between one or more images in the searched image group and the received query is possible when there is no image to be completely matched, and the partial matching is confirmed. The method of claim 1, further comprising extracting a matched image. The one or more images in the group are arranged in descending order of reliability of the image, and whether or not the complete matching or partial matching is performed is made in the order in which the one or more images are arranged. 5. The method according to any one of 4 to 4. 6. The method according to claim 5, wherein the reliability is calculated as a total value of numerical values obtained by assigning weight values to each element of description data included in an image. The method of claim 1, wherein when a plurality of image groups related to the received query are searched, the third to fourth steps are performed on the plurality of image groups. The method of claim 1, wherein the fourth step further comprises a step of extracting a representative image that has been set and transmitting the extracted representative image to the user's terminal device when the extracted image does not exist. 9. The method of claim 8, wherein the preset representative image is an image having the highest reliability in the image group. The image group is
Combining images with the same signature value into one group;
Generating description data from each image grouped into the one image group;
2. The method of claim 1, wherein the image data is generated through calculating the reliability of the image using the generated description data; and arranging the images in the order of reliability in the image group. The method described. A computer-readable recording medium having recorded thereon a computer program for executing the method according to any one of claims 1 to 10. In a system that provides the most suitable image for inquiries regarding user image search on a communication network,
Means for receiving an inquiry about an image that the user intends to search from a terminal device of the user;
Means for retrieving an image group associated with the received query;
Means for dynamically determining whether or not perfect matching between one or more images in the retrieved image group and the received query is possible, and extracting a matched image when complete matching is confirmed And a system including an image search unit having means for providing the extracted image to the terminal device of the user. The system of claim 12, wherein the image inquiry receiving means further comprises means for transforming the received inquiry into meaningful morphemes or combined morphemes through parsing. The matching image extracting unit determines whether or not partial matching is possible between one or more images in the searched image group and the received inquiry when there is no image to be completely matched, and the partial matching is confirmed. The system of claim 12, further comprising means for extracting a matched image when done. The one or more images in the group are arranged in descending order of reliability of the image, and whether or not the complete matching or partial matching is performed is made in the order in which the one or more images are arranged. 15. The system according to any one of 14 to 14. 16. The system according to claim 15, wherein the reliability is calculated as a total value of numerical values obtained by assigning weight values to each element of description data included in an image.