Characterization of Complex Image Spatial Structures Based on Symmetrical Weibull Distribution Model for Texture Pattern Classification

Table 2

Summary of the comparative methods.

Methods	Feature characteristic	Feature dimension per image	Classifier
GLCM [48]	Statistical measures		PLS-DA
GWT [49]	Statistical measures		PLS-DA
MRMRF [50] +PLS-DA	Statistical model parameters		PLS-DA
MRMRF+NLC	Statistical model parameters		NLC approach [50]
VZ algorithm [11]	Frequency histogram of learned textons		Nearest neighbor classifier(NNC) with the statistic for histogram distance measurement
PATCH [28]	Frequency histogram of learned textons		NNC, Joint classifier
PATCH-MRF [28]	Two-dimensional histogram		NNC, Joint clssifier
WLD [12]	One-dimensional histogram encoded from a differential excitation and gradient orientation-paired two-dimensional histogram		-nearest neighbor classifier, , with normalized histogram intersection as the similarity measurement of two histogram
PRICoLBP [16] +SVM	Two-dimensional histogram: Pairwise rotation invariant cooccurrence LBP feature		SVM
PRICoLBP+PLS-DA	Two-dimensional histogram: Pairwise rotation invariant cooccurrence LBP feature		PLS-DA
ISM-TPI	Statistical model parameters		PLS-DA