Second order-based image retrieval algorithm

摘要

Under the environment of big data, retrieval becomes a crucial technology and image retrieval is paid more attention and widely used. The paper proposes a second-order retrieval algorithm, of which can be used to retrieval the similar images. Firstly, extracting image sift features. Then, build frequency table of characteristic words by k-means clustering and bag of word algorithm. Finally, based on word frequency table, first retrieve the images that have similar distribution characteristics of the structure. The second-order retrieval implement accurate retrieval of images according to the proportion of the corresponding feature points that belonging to the same class. The experimental results show that this method has good recall factor and good effect on query efficiency. It's a kind of method can be used.