Multiresolution Histogram Based Image Retrieval with Pseudo-Multiple Examples
XU Sheng-rong · 2007
The paper presents a new content-based image retrieval method.The method decomposes an image with multiresolution to get a set of sub images,which depicts the information the index image from global to local precision.Histograms are computed as the retrieval features.We use the multiple examples retrieval method to fuse the result.The paper makes the analysis of the retrieval results with different fusion policies and shows that our method is promising in the content-based image retrieval.