基于内容的图像检索的方法研究

On the Methods of Content Based Image Retrieval

一、前言近几年来,随着大容量存储器和数字化图像设备,诸如扫描仪、数字相机等技术的发展和普及应用。图像数据呈现级数的增长。由于无论是WWW还是企业单位部门都拥有大量的图像,因此从图像中提取有用的信息的前提是图像的有效检索,这也是数据转化为信息的关键一步。图像检索的传统方法是基于文本的。使用关键字注释是最常用的方法。对图像的检索变成了对关键字的查找。这种方法可以利用传统的DBMS实现,简单易行,而且能够从用户角度表达图像内容的高层语义。

Content Based Image Retrieval is the most active research area. This paper summarizes the important result reached by now in this area,and introduces some famous systems. At his childhood, CBIR has a lot of problems remained unsolved. Pointing out the default in the existing methods, this paper put over our ideas and suggested some future research directions.