JP2002082969A - Automatically indexing robot system and processing method using the system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an automatically indexing robot system capable of automatically indexing a text type material such as a text or word processor data, an image and a representative screen of video, and each material obtained by developing graphic data with an image and to provide a processing method using this system. SOLUTION: This automatically indexing robot system is constituted of a server 10 for storing information such as an XML document, a WPS material, image scan, a moving video material, and photograph picking-up and an index word or an image index and a robot PC 20 for retrieving the information stored in the server 10 by using an index word retrieving machine or a character recognizing machine or for extracting vector image data. Then, processing using this automatic indexing robot system is performed by successively executing a first process 100 for automatically indexing a character resource type with the material stored in the server 10, a second process 20 for automatically indexing the scanned origin image with the material stored in the server 10, and a third process 300 for automatically indexing the photographic image with the material stored in the server 10.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an index (Indeed) formed in a structure that allows a search engine to easily search for a document when searching for a website on the Internet.
More specifically, the present invention relates to an automatic indexing robot system capable of automatically indexing text-type materials such as texts and word processor data, representative screens of images and videos, and materials obtained by developing graphic data into images. It is related to the processing method performed.

[0002]

2. Description of the Related Art Conventionally, information retrieval has only been accepted as the task of organizing and using information in a library or searching for an index collection created by a group of experts. In this era, important information was mostly created in documents, and the method of asking a document person was mainly used to utilize the constructed information.

However, as the performance of computers has been improved and the use of computers has been generalized, the content of information has been transferred from documents to computers, and searches have been performed by computers instead of experts. Was.

In particular, the advent of search computers has played a leading role in the information revolution, and even if the amount of information has increased exponentially, information can be searched more quickly and accurately than in the past. It has become. That is, various and enormous amounts of information that could not be accessed before can be easily accessed.

[0005] Therefore, the performance of an information retrieval system depends on accuracy and speed. In particular, the role of an index word as a search word is very important to improve the accuracy of retrieval.

[0006] The Korean index targets characters used in Korean documents, and the main character is Hangul.
In addition, since English characters, Chinese characters, numbers, and other symbolic characters are used, it is better to perform indexing work on foreign language characters. Morphological analysis and extract index words through an automatic indexing system.

Also, in the process of extracting index words, for words that are used irregularly, the original form is restored and converted to the basic form.
The extracted index words should be subjected to unnecessary word inspection and compound noun expansion processing at the registration stage.

[0008] On the other hand, in the information search, in short, various techniques have been used to improve the accuracy of such search in a process of searching for and providing desired information by a user. In particular, the inverse file technique adapted to a large amount of information that is not standardized is widely used, and the inverse file technique basically depends on the index term. I do.

Here, the method of generating the index word is as follows.
The method is divided into a manual index manually and an automatic indexing method based on the morphological analysis principle. The automatic indexing method analyzes a morpheme using a computer and extracts an index word using the result. For example, in the case of a Korean automatic indexing system, an index word is extracted from a document produced based on Korean. First, a morpheme is analyzed using a morphological analysis system, and the result is used as an index word. Unnecessary unnecessary words are removed and an index word is finally extracted.

[0010] Here, Korean includes Hangul,
It includes English characters, numbers, kanji, Japanese characters, Greek characters, Russian characters and other symbolic characters, and especially in the case of Korean, from the viewpoint of language classification, it has characteristics as a sticky word, so words Is difficult to use immediately as an index term, and a separate processing step should be performed. That is,
Performing a morphological analysis process to analyze morphemes adapted to Korean characteristics plays the most important role in automatic indexing.

In the analysis of Korean morphemes, it is possible to analyze not only Hangul characters, but also English characters, Japanese characters, Russian characters, etc., and it is also possible to partially analyze Chinese characters.

Here, the index is useful when searching for information, and means an index arranged according to a specific principle or purpose. That is, it can be said that it is a connection medium that connects the searcher and the information. Thus, the index basically has a sorting function (distinguishing from others) and a pointing function (pointing to the original information).

There are various types of indexes, but they are roughly classified into a subject index and a non-subject index. The subject index selects an element representing a subject from an information source as an index, and the non-subject index selects an element unrelated to the subject as an index.

More specifically, first, in the case of the non-subject index, since it is a mere collection of data such as dates, names and titles which are not related to the subject, there is no particular difficulty in the indexing operation. It is easy to accumulate index information using a database using a computer, and search is very easy.

On the other hand, in the case of the subject index, an element which is likely to be a subject should be searched for from an information source, so that the work is intelligent and time-consuming. For this reason, people who have received specialized education or have specialized knowledge in the relevant field should perform the work, but it is virtually impossible in the recent situation where a huge amount of literature is being produced. is there.

Therefore, instead of subject indexing by humans,
2. Description of the Related Art A text search system that uses a computer to select a word in a text as a subject word and uses it as a search keyword has attracted attention.

Therefore, among the elements constituting the search engine,
The indexer is the most important part of the search system. This is because the index file creation speed, the index file size, and the search speed depend on the file structure in which the noun list and the meta information extracted by the morphological analysis are stored. Therefore, the structure of the indexer is also a part that is not generally disclosed.

In general, the structure of an index file uses an "inverted file technique". In particular, many products use an inverted file using a B-tree. It is organized into a list linked in both directions. In addition, there are various index file structures using such a modification of the file structure.

[0019]

However, in such a conventional indexing method, after performing scanning, it is transmitted to a PC capable of character recognition one by one, or a subject using a workstation dedicated to image pattern analysis. Extraction and reacquisition must be performed, which is disadvantageous in that it takes time.

The present invention has been made in view of such conventional problems, and has been developed in consideration of text-type materials such as text and word processor data, representative screens of images and videos, and various materials obtained by developing graphic data into images. It is an object of the present invention to provide an automatic indexing robot system capable of automatically indexing a robot and a processing method using the same.

[0021]

In order to achieve the above object, in an automatic indexing robot system according to the present invention, information such as an XML document, a WPS material, an image scan, a moving image material, and a photograph, and an index word or A server that stores an image index, and a robot PC that searches information stored in the server using an index word search or a character recognizer or extracts vector image data. It is characterized by the following.

In order to achieve the above object, in the processing method using the automatic indexing robot system according to the present invention, a first step of automatically indexing a character data type based on the data stored in the server; And a third step of automatically indexing the original image scanned by the material stored in the server and a third step of automatically indexing the photographic image by the material stored in the server.

The first step includes a first step of transmitting character data type data from the server, and a second step of extracting an index term character string from the character data type data transmitted in the first step. And transmitting the index words extracted as character strings in the second step to the server in a third step.

In the second step, a first step of receiving the transmitted original image data from the server is performed, and a character recognizer is operated to extract a character string from the image data transmitted in the first step. A second step of extracting an index word character string by the character recognizer operated in the second step, and a second step of transmitting the index word or character string extracted in the third step to the server. 4 steps,
When the character string extracted in the fourth step is transmitted, a fifth step of extracting an index word from the character string, a sixth step of storing the index word extracted in the fifth step, And a seventh step of correcting the text using the character string of the index word stored in the six steps.

In the third step, a photographic image is transmitted from a server in a first step, and an index is extracted from the photographic image material transmitted in the first step according to form, hue, texture, brightness and saturation. And a third step of transmitting the index of the photographic image extracted in the second step to the server.

[0026]

Embodiments of the present invention will be described below with reference to the drawings. In the automatic indexing robot system according to the present invention, as shown in FIG. 1, a server 10 that stores information such as XML documents, WPS materials, image scans, moving image materials, photographing, index words and image indexes, A robot PC 20 that searches information stored in the server 10 using an index word search or a character recognizer, or extracts vector image data.

In the thus configured automatic indexing robot system according to the present invention, the server 10 has a function of communicating with the robot PC 20 and a function of communicating with the robot PC 20.
And a function of systematically storing an index word or an image index by transmitting an index from the system.

On the other hand, the robot PC2O should have a hardware-based or software-based index term search function in the case of a text material, and a character recognition board (in the case of an original image material). Boar
d) Or character recognition software should be provided, and for photographic images, vector data extraction function should be provided.

A processing method using the automatic indexing robot system according to the present invention using the server 10 having the above functions and the robot PC 2O will be described below. First, if the information to be indexed by the user is a document, when text data is transmitted from the server 10, the robot PC 20 first analyzes a morpheme using a morphological analysis system, and then analyzes the morpheme. Then, unnecessary words that are not suitable as an index word are filtered from to finally extract an index word.

Here, the unnecessary word (Stop word) means an unnecessary word that does not need to be processed as an index word among the analyzed index words.

Next, the extracted index words are transmitted to the server 10 and stored. On the other hand, if the information to be indexed by the user is a scanned original image, and the scanned image data is transmitted from the server 10, the robot PC 20 operates a character recognition device (not shown) to transmit the text. After extracting a character string by recognizing, a morpheme is analyzed using a morphological analysis system, and unnecessary words that are not suitable as an index word are filtered from the analysis result to finally extract an index word. Next, the index word or character string extracted in this way is
When the character string extracted by the robot PC 20 is transmitted, the server 10 extracts an index word from the transmitted character string, stores the index word, and uses the character string to convert a text. to correct.

On the other hand, if the information to be indexed by the user is a photographic image, when the photographic image is transmitted from the server 10, the robot PC 20 extracts a vector of the photographic image, and An index is extracted based on hue, texture, brightness, and saturation.

Next, the index of the photographic image thus extracted is transmitted to the server 10 and stored therein. The processing method using the automatic indexing robot system according to the present invention that operates as described above will be briefly described as follows. That is, a first process 100 for automatically indexing the text material type according to the material stored in the server, a second process 200 for automatically indexing the original image scanned by the material stored in the server, and storing the data in the server. 3rd step 300 for automatically indexing photographic images with selected material
Are sequentially performed.

Then, in the first step 100, FIG.
As shown in (1), a first step 101 in which the character data type data is transmitted from the server, and a second step 102 in which an index term character string is extracted from the character material type data transmitted in the first step 101
And a third step 103 of transmitting the index term extracted as a character string in the second step 102 to the server.

In the second step 200, as shown in FIG. 3, a scanned original image data is transmitted from a server in a first step 201, and the scanned image data is transmitted from the image data transmitted in the first step 201. A second step 202 of operating a character recognizer to extract a character string; and a third step of extracting an index term character string by the character recognizer operated in the second step 202.
Step 203, a fourth step 204 of transmitting the index word or the character string extracted in the third step 203 to the server, and if the character string extracted in the fourth step 204 is transmitted, the transmitted A fifth step 205 for extracting an index term from the character string, and a sixth step 206 for storing the index term extracted in the fifth step 205
And a seventh step 207 of correcting a text using the character string of the index word stored in the sixth step 206 are sequentially performed.

Also, in the third step 300, as shown in FIG. 4, a first step 301 in which a photographic image is transmitted from the server, and a form, based on the photographic image material transmitted in the first step 301, A second step 302 of extracting an index based on hue, texture, brightness and saturation, and a third step 3 of transmitting an index of the photographic image extracted in the second step 302 to the server
03, and so on.

[0037]

As described above, in the automatic indexing robot system and the processing method using the same according to the present invention, it is possible to provide a user with a visually accurate, convenient and high-quality search service. Therefore, the process can be shortened due to automation using a robot, and the 24-hour operation system eliminates the need to build a large number of recognition systems, thus reducing costs. is there.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing an automatic indexing robot system according to the present invention.

FIG. 2 is a flowchart showing a first step (100) of a processing method using the automatic indexing robot system according to the present invention.

FIG. 3 is a flowchart showing a second step (200) of the processing method using the automatic indexing robot system according to the present invention.

FIG. 4 is a flowchart showing a third step (300) of the processing method using the automatic indexing robot system according to the present invention.

[Explanation of symbols]

 1O: Server 20: Robot PC

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06T 1/00 200 G06T 1/00 200E

Claims (5)

[Claims] 1. An XML document, a WPS material, an image scan,
A server 10 for storing information such as moving picture materials, photographing, and index words or image indexes; and searching for information stored in the server 10 using an index word search or a character recognizer, and a vector image. An automatic indexing robot system, comprising: a robot PC 20 for extracting data; 2. A first step 100 for automatically indexing textual material type data based on materials stored in a server, a second step 200 for automatically indexing original images scanned by the materials stored in the server, A third method 300 for automatically indexing photographic images based on materials stored in a server. 3. The first step 100 is a first step 1O1 in which character data type data is transmitted from a server.
A second step 102 of extracting an index term character string from the character data type data transmitted in the first step 101, and a second step of transmitting the index term extracted as the character string in the second step 102 to the server. 3. The processing method using the automatic indexing robot system according to claim 2, wherein the following three steps are sequentially performed. 4. The second process 200 includes a first step 201 of receiving the scanned original image material from the server, and a character recognition for extracting a character string from the image material transmitted in the first step 201. A second stage 202 for operating the machine, a third stage 203 for extracting an index word character string by the character recognizer operated in the second stage 202, and an index word or character string extracted in the third stage 203. To the server, and if the character string extracted in the fourth step 204 is transmitted,
A fifth step 205 for extracting an index term from the character string, and a sixth step 2 for storing the index term extracted in the fifth step 205
The automatic indexing robot according to claim 2, wherein the following steps are sequentially performed: 06, and a seventh step 207 for correcting a text using the character string of the index word stored in the sixth step 206. Processing method using the system. 5. The third process 300 includes a first step 301 of transmitting a photographic image from a server, and a form, a hue, a texture, a brightness, and a saturation of the photographic image material transmitted in the first step 301. Index extraction
The automatic indexing robot system according to claim 2, wherein a second step 302 and a third step 303 of transmitting an index of the photographic image extracted in the second step 302 to the server are sequentially performed. Processing method.