Multi-resolution local ternary patterns for image indexing retrieval

Abstract

A new image indexing and retrieval algorithm known as multi-resolution local ternary patterns (Mu_LTP) is presented in this paper. LTP histogram captures the distribution of edges in an image which are evaluated by taking into consideration of local difference between the centre pixel and its neighbours. Local ternary patterns (LTP) is more discriminate and less sensitive to noise in uniform regions as compared to local binary patterns (LBP). Multi-resolution texture decomposition and LTP has been efficiently used in the proposed method where multi-resolution images are computed using Gaussian filter. Eventually, feature vectors are constructed by making into play LTP on multi-resolution images. The retrieval results of the proposed method are examined on two different natural and texture image databases viz Brodatz database (DB1), and MIT VisTex database (DB2), and shows a major improvement in terms of average retrieval precision (ARP) and average retrieval rate as when weighed against with LBP, LTP and some existing transform domain techniques.