A compact image signature for RTS-invariant image retrieval

Pratim Ghosh*, B.S.Manjunath$, K.R.Ramakrishnan*
*Indian Institute of Science, Bangalore, India, (prathim,krr [at] ee.iisc.ernet.in) ,
$University of California Santa Barbara, USA, (manj [at] ece.ucsb.edu).


In this paper we propose a novel framework to obtain a very compact image signature (32 bits) which is invariant to rotation, translation, scaling and other minor perturbations like smoothing, random noise addition, JPEG compression etc. The framework involves Fourier-Mellin transform, conventional PCA and non-uniform scalar quantization. The high retrieval efficiency and low space consumption demonstrates the significance of our signature in duplicate image retrieval and large image database indexing.
[PDF] [BibTex]
Pratim Ghosh, B.S. Manjunath and K.R. Ramakrishnan,
IEE International Conference on Visual Information Engineering (VIE 2006), Bangalore, India, Sep. 2006.
Node ID: 450 , DB ID: 254 , Lab: VRL , Target: Proceedings
Subject: [Object-Based Retrieval] « Look up more