Multi-feature Index Structure for Weighted Query Applications

In the application of video retrieval by sample,video data is described by multiple high-dimensional features,and the weights of these features are changed in different queries.We propose a new indexing structure called multi-feature index tree(MFI-Tree) to index multiple high-dimensional features of video data for this retrieval application.MFI-Tree employs tree structure which is benefit for browsing application,and travels the last level aggregate node in retrieval application to improve the performance.And more,aggressive decided distance for kNN search algorithm which fast reduces the distance to prune the search space more effectively is proposed.The experimental results show that MFI-Tree and ADD-kNN algorithm offer performance advantages over sequential scan.