New Method of Content Based Image Retrieval based on 2-D ESPRIT Method and the Gabor Filters

Abstract

We propose, in this paper, a new method for Content Based Image Retrieval (CBIR) by exploiting the digital image content. Our method is based on the representation of the digital image content by a characteristics vector of the indexed image. Indeed, we have exploited the image texture to extract its characteristics and for constructing a new descriptor vector by combining the Bidimensional High Resolution Spectral Analysis 2-D ESPRIT (Estimation of Signal Parameters via Rotationnal Invariance Techniques) method and Gabor filter. To evaluate the performance, we have tested our approach on Brodatz image database. The results show that the representation of the digital image content appears significant in research of imaging information.