论文部分内容阅读
Many recent state-of-the-art image retrieval approaches are based on Bag-of-Visual-Words model and represent an image with a set of visual words by quantizing local SIFT(scale invariant feature transform)features.Feature quantization reduces the discri