A content-based image retrieval system using extended SQL in RDBMS

被引：1

作者：

Lee, SH ^{[1
]}

Jang, GH ^{[1
]}

Lee, SH ^{[1
]}

Jung, SH ^{[1
]}

Woo, YT ^{[1
]}

机构：

[1] Changwon Natl Univ, Dept Comp Sci, Changwon, South Korea

来源：

ICICS - PROCEEDINGS OF 1997 INTERNATIONAL CONFERENCE ON INFORMATION, COMMUNICATIONS AND SIGNAL PROCESSING, VOLS 1-3: THEME: TRENDS IN INFORMATION SYSTEMS ENGINEERING AND WIRELESS MULTIMEDIA COMMUNICATIONS | 1997年

关键词：

D O I：

10.1109/ICICS.1997.652145

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

In this paper, we implement an efficient image retrieval system via content-based approach in the relational database system. In our system, as the preprocessing stages, we extract the color and texture feature from the input image to get the retrieval keywords in the image database. First, our system classifies the images roughly by the color feature using the ART2 neural network model. The classified information is stored in a relational database with the original image. To retrieve images similar to a query image, we extend relational SQL statement in coupled with the ART2 function in the WHERE condition clause. Next, to retrieve final results of images, we compare the query image with the candidate images which have the same category information by the texture feature. The proposed system is implemented on Oracle DBMS. It shows satisfactory retrieval results in the test of the sample image database.

引用

页码：1069 / 1072

页数：2