Statistical Wavelet Subband Characterization Based on Generalized Gamma Density and Its Application in Texture Retrieval

Abstract

The modeling of image data by a general parametric family of statistical distributions plays an important role in many applications. In this paper, we propose to adopt the three-parameter generalized Gamma density (GGammaD) for modeling wavelet detail subband histograms and for texture image retrieval. The advantage of GGammaD over the existing generalized Gaussian density (GGD) is that it provides more flexibility to control the shape of model which is critical for practical histogram-based applications. To measure the discrepancy between GGammaDs, we use the symmetrized Kullback-Leibler distance (SKLD) and derive a closed form for the SKLD between GGammaDs. Such a distance can be computed directly and effectively via the model parameters, making our proposed scheme particularly suitable for image retrieval systems with large image database. Experimental results on the well-known databases reveal the superior performance of our proposed method compared with the current existing approaches.

A performance evaluation of gradient field HOG descriptor for sketch based image retrieval

“…These systems were based primarily on global color histograms [7], spatial patterns [19] or region adjacency [20,21]. Robust color blob-based matching was later proposed using spectral descriptors, such as wavelets [22,23] which have more recently proven useful for more general image retrieval using colour texture [24,25].…”

Multivariate texture retrieval using the Kullback-Leibler divergence between bivariate generalized Gamma times an Uniform distribution

“…This histogram is then modeled by the corresponding distributions of u (see (3)) for the BGΓU, the BGGD, the Laplace and the Gaussian distributions as shown in Fig. 1.…”

“…Those works have further been extended by the use of the generalized Gamma (GΓ) distribution which generalizes these two models [3]. However, these approaches do not fully exploit the texture information in the image.…”

An Improved AP-Wishart Classifier for Polarimetric SAR Images by Incorporating a Textural Features

“…The flowchart of the AP-Wishart classifier is shown in The process of the improved classification scheme is as follows: A wavelet extraction algorithm is used to extract texture from the PolSAR image. The wavelet as a kind of information extraction method is widely used in signal processing and image analysis [21], [22]. After Daubechies' wavelet basis function is chosen, a pyramid algorithm from Mallat is used to extract texture from a polarimetric span image.…”