• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2007, Vol. 29 ›› Issue (6): 42-44.

• 论文 • 上一篇    下一篇

图像搜索中基于网页分块的图像分类研究

章勤 余洋 陶文兵   

  • 出版日期:2007-06-01 发布日期:2010-06-03

  • Online:2007-06-01 Published:2010-06-03

摘要:

本文提出并实现了一种基于网页分块的图像分类模型,利用图像位于网页中的特点,将图像和网页结构有效地结合起来对网络环境中的图像进行分类。首先利用VIPS页面分块算 法对网页进行分块,然后根据分块网页结构的特点将各分块网页中的图像分为背景类、广告类和主题类,最后利用这个图像分类结果指导网络环境下的图像搜索。实验结果表明 明,这种方法能够显著提高图像搜索性能。

关键词: 网页分块 图像搜索 分类 VIPS

Abstract:

Considering the characteristies of images and the structure of web pages, an image classification model is proposed and implemented. First, a web segmentation algorithm is applied to Web pages. Then, based on using the results of segmentation, all the pictures are classified into three categories., ba ckground, advertisement and subject. Finally, the resuits of classification conduct the image retrieval based on texts and composition retrieval. Experi mental results indicate that this method can obtain a better accuracy.

Key words: page segmentation, image search, classification, VIPS