특징 기술자와 의미론적 키워드를 이용한 웹기반 이미지 검색 시스템 설계

Metadata Downloads
Issued Date
A variety of web services are needed due to development of information and communication network and rapid increase of the Internet demand, but as data of enormous volume are needed for services on the multimedia data, which causes overload in network, it is hard to provide quick and various services.
Therefore, as a rapid development of multimedia technology and increase of population to use internet, the necessity of retrieval system for various images on the web is on the rise.
There are two kinds of image retrieval system on the web. One is a system that retrieval to present images from the image database contained in the server in itself through web interface. The other is a system that present the results of retrieval by the text-based retrieval method for the images on the web. The former is high in the confidence level, but is not satisfactory for the web user's needs that make request for various images on the web. The latter can retrieve the number of various images on the web, but has the problem of the fall in responsibility for retrieval.
To solve the problem, MPEG-7, which is an international standard of meta data representation in multimedia, was established in early 2002. accompanied by development of various types of services.
Also, an incorporating service is becoming a matter of common interest because a variety of users require One-Source Multi-Use which was designed to apply multimedia services at various platform.
In this background, this paper suggested the way of solving the problem that has the incorrectness of the existing keyword-based multimedia retrieval system. Based on it, this paper try to construct the paper with the web-based multimedia retrieval system.
For the solution of the defect of the exiting simple keyword-based multimedia retrieval system, first of all, we applicate the semantic-based keyword method instead of it. Such makes a syntactic retrieval to be able through a input with the keyword that has the information matched a descriptor information instead of the simple keyword. The color keyword can accept only registered colors and gets rid of problem that the retrieval has exception though it is the same meaning of a keyword. Also, it is easy for you to draw a objective keyword and refer to syntactic retrieval when you make a input with segmenting texture, shape, object etc.
Secondly, the suggested system can refer to information including the descriptor information in case of necessity of a keyword-based. Descriptors can use color, texture, shape and add or remove various algorithm in case of need.
Thirdly, it provides a function of a integrative retrieval by using combination of a keyword and each descriptor information. As a rate set up by users, each method draws a degree of similarity and refers to the image of the highest similarity.
Lastly, all of the image information are stored in database and referred results are consisted of XML data to support the function such as retrieval efficiency, high- speed retrieval, various output methods.
Proposed system constructed web-based multimedia retrieval system with JSP to verify the suggested system and simulated it by using several images.
Alternative Title
Design of the Web based Image Retrieval System Using Feature Descriptors and Semantic Keyword
Alternative Author(s)
Kang, Sung-Kwan
朝鮮大學校 大學院
일반대학원 전자공학과
Awarded Date
Table Of Contents
그림목차 = ⅲ
표목차 = ⅴ
Abstract = ⅵ
Ⅰ. 서론 = 1
A. 연구의 배경 및 목적 = 1
B. 연구의 내용 및 구성 = 5
Ⅱ. 멀티미디어 관리 시스템 = 7
A. MPEG-7 (Moving Picture Expert Group) = 7
1. MPEG의 표준화 동향 = 7
2. MPEG-7 표준 = 10
B. 멀티미디어 데이터 정보 추출 = 17
1. Color = 17
2. Texture = 27
3. Shape = 32
4. 기술자 설계 = 37
Ⅲ. 검색 엔진 = 45
A. 이미지 정보의 검색 방법 = 45
1. 문자 기반 검색 = 45
2. 의미론적 기반 검색 = 46
3. 내용 기반 검색 시스템 = 46
B. 웹 기반 이미지 검색 시스템 = 54
Ⅳ. 시스템 설계 = 56
A. 기술자 설계 = 56
1. 색상 기반 검색 시스템 설계 = 56
2. 질감 기반 검색 시스템 설계 = 60
B. 검색 시스템의 설계 = 62
C. 데이터 구조 및 데이터베이스의 설계 = 64
1. 데이터베이스 설계 = 64
2. 데이터 구조 설계 = 64
D. 웹 기반 멀티미디어 데이터 관리 시스템의 설계 = 67
1. 시스템의 구성 = 67
2. 인터페이스 설계 = 68
Ⅴ. 구현 및 분석 = 70
A. 구현 = 70
1. 시스템 구성 = 70
2. 사용 이미지 = 70
3. 구현 시스템 = 74
B. 분석 = 77
1. 색상 기반 검색 = 77
2. 질감 기반 검색 = 89
3. 키워드 기반 검색 = 92
Ⅵ. 결론 및 향후 연구과제 = 94
참고문헌 = 96
朝鮮大學校 大學院
姜成官. (2005). 특징 기술자와 의미론적 키워드를 이용한 웹기반 이미지 검색 시스템 설계.
Appears in Collections:
General Graduate School > 4. Theses(Ph.D)
Authorize & License
  • AuthorizeOpen
  • Embargo2005-10-20
Files in This Item:

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.