局部二值统计模型在图像检索中的应用探析

The Local two Statistical Model in the Application of Image Retrieval

互联网上的图像资源呈现海量增长的态势,网络图像检索技术的研究和应用更显重要。怎样实现基于特征的合理高效的图像检索是多媒体信息检索领域的热点课题,图像的常见特征有颜色、形状、纹理、空间关系等。该文在介绍关于图像纹理的基本检索原理和小波理论的基础上,重点讨论通过局部二值统计模型解决纹理图像旋转变化时出现检索率降低问题的方法及实验分析结果。

Image resources on the Internet are increasing massively,hence the importance of research and application of image re trieval.The current researches focus on content-based image retrieval,especially the feature-based retrieval.The most common features including color,texture,shape and spatial location.Based on the introduction to the basic principles of image texture re trieval and the theory of wavelet transform,this study explores the method to solve the problem of retrieving rate reduction by adopting two values local statistical model when the texture image rotation changes and discusses the experimental results.