点击次数：

论文类型：期刊论文

发表时间：2022-06-29

发表刊物：光电子 激光

所属单位：机械工程学院

卷号：20

期号：7

页面范围：959-963

ISSN号：1005-0086

摘要：This paper first introduces the Maximally Stable Extremal Region (MSER)
   detector,which is affine invariant.According to disjoint-set forests
   data structure and union-find,the extremal regions are
   extracted.And,combing the component tree and maximally stable extremal
   condition,the MSERs are obtained.Then the SIFT descriptors,which are
   used as local feature at low level,are produced in the MSER and then
   clustered into the visual "words".By using standard weighting,the query
   region is selected by the rectangle in the retrieval image.Based on the
   similarity of database images and the query regions,the image retrieval
   results are ranked.With spatial consistency measurement rule of regions
   matching method of search unit,the image retrieval results are
   obtained.The experiment shows these distinctive feature regions are
   invariant to the change in scale,rotation,translation and viewpoint,and
   the retrieval mechanism also improves the performance and reliability of
   image retrieve system greatly.

备注：新增回溯数据