摘要
为了获得理想的大规模图像检索结果,设计基于云平台和分布式计算的大规模图像检索方法。采用云平台和分布式计算技术构建图像检索环境,使用SIFT算法提取图像检索特征,通过k-means聚类进行特征向量降维,根据视觉词典与图形向量的对应关系构建倒排索引结构,初步实现图像的大规模检索,利用多特征融合方法进一步提升图像检索的准确性。测试结果表明,该检索方法具有较好的扩展率与加速比,图像检索准确性高,图像检索效率高。
In order to obtain the ideal large-scale image retrieval results,a large-scale image retrieval method based on cloud platform and distributed computing is designed.The cloud platform and distributed computing technology are used to build the image retrieval environment,the SIFT algorithm is used to extract the image retrieval features,the dimension of the feature vector is reduced through k-means clustering,the inverted index structure is constructed according to the corresponding relationship between the visual dictionary and the graphic vector,the large-scale image retrieval is preliminarily realized,and the multi-feature fusion method is used to further improve the accuracy of image retrieval.The test results show that the retrieval method has better expansion rate and acceleration ratio,high image retrieval accuracy and high image retrieval efficiency.
作者
李薇
LI Wei(Academic Affairs Office,Shaanxi Post and Telecommunication College,Xianyang 712000,China)
出处
《微型电脑应用》
2024年第8期211-215,共5页
Microcomputer Applications
