摘要
基于内容的图像检索(CBIR)技术依赖于对图像特征(例如颜色直方图、纹理、草图、形状等)的提取,相对于传统的基于文本的图像检索方式,这种方式提高了检索效率和检索的准确率。文中主要介绍了基于颜色和基于纹理特征这两种特征提取方法。这两种方法既能够反映全局特征,又能够兼顾所感兴趣区域的局部特征,是基于内容的图像检索的两种非常有效的方法。
Gontent- based image retrieval（CBIR） techniques rely on the features that are abstracted from the images. This method is more efficient and more accurate than the traditional method of text - based image retrieval. This paper introduces two methods to abstract the features of the image. The one is color - based image retrieval, the other is texture - based image retrieval. At the same time those two methods can reflect global feature and regions - of- interest feature. At the same time those are effective ways to solve the problems about content - based image retrieval.
出处
《计算机技术与发展》
2006年第6期191-192,195,共3页
Computer Technology and Development
关键词
颜色
纹理
图像检索
color
texture
image retrieval
