Statistical Modeling of Gabor Filtered Magnitude for Batik Image Retrieval

Abstract

This paper presents an effective and efficient way on statistical modeling of Gabor filtered magnitude to generate an image feature descriptor. Two statistical distributions, i.e. Gaussian and Rayleigh distributions, are considered in the image feature extraction. The image feature is simply constructed by concatenating the distribution estimators of Gabor filtered magnitudes under different scales and orientations. As documented in the experimental section, the proposed method yields good performance in the Batik image retrieval system. In addition, the performance of Gabor feature can be improved by injecting the color feature in order to capture the color richness of an image.