Large Scale Image Searching Using Two Dimensional Binary Trees Combined with Feature Selection

Abstract

-The challenging task in image retrieval is to find similar images based on the distance between images. Content based image retrieval system is used to identify the similarity between multidimensional image feature vectors. If the database is large, searching the image linearly will be too slow. We have proposed the two dimensional binary tree approach for searching the images based on feature detectors and they have improve the retrieval efficiency significantly.