Combining LBP and contourlet features for image retrieval

摘要

Features for image retrieval application can either be extracted locally or globally. In this paper, a combined approach is proposed in which image retrieval is done by combining the Local Binary Pattern (local feature) and Contourlet features (global feature). The proposed approach is applied to a COREL database of 1000 sub-images from ten different sets. Experimental result shows that the proposed technique achieves good retrieval rate.