【摘要】 研究大型图像数据库的检索速度优化问题。在大型图像数据库中,存在海量特征的图像资源,图像之间不存在直接性的联系,再加上数量巨大,给建立快速的图像检索带来了困难。传统的数据库信息检索方法,通过简单的属性对应关联进行检索,建立在图像存在单一属性特征的基础上,随着图像复杂性的增加,上述方法很难对图像作进一步的特征提取,造成图像检索耗时大量增加。为解决上述问题,提出一种大型图像数据库快速检索方法。利用局部二维模式纹理分析方法,提取数据库中的纹理特征,从而为数据库的检索提供依据。引入渐变计算方法,在大型图像数据库中,根据数据库中图像纹理特征,完成目标检索。实验结果表明,利用改进算法进行大型图像数据库检索,能够提高检索速度,从而提高大型图像数据库的性能。更多还原

【Abstract】 The large image database retrieval speed optimization problems were studied. A large image database quick retrieval method based on cloud computing was proposed in the paper. Local 2-d pattern texture analysis method was employed to extract texture features in the database,so as to provide the basis for the database retrieval. Using cloud computing methods,in the large image database,based on image texture features in database,the goal retrieves were achieved. The experimental results show that the algorithm can improve the retrieval speed and performance of large image databases.更多还原