doiSerbia

MFI-tree: An effective multi-feature index structure for weighted query application

He Yunfeng (School of Computer Science & Technology, Huazhong University of Science & Technology, Wuhan, China)
Junqing Yu (School of Computer Science & Technology, Huazhong University of Science & Technology, Wuhan, China)

Multi-Feature Index Tree (MFI-Tree), a new indexing structure, is proposed to index multiple high-dimensional features of video data for video retrieval through example. MFI-Tree employs tree structure which is beneficial for the browsing application, and retrieves the last level cluster nodes in retrieval application to improve the performance. Aggressive Decided Distance for kNN (ADD-kNN) search algorithm is designed because it can effectively reduce the distance to prune the search space. Experimental results demonstrate that the MFITree and ADD-kNN algorithm have the advantages over sequential scan in performance.

Keywords: multi-feature index tree, kNN, aggressive decided distance for kNN, video retrieval