Combining frequent itemsets and statistical features for texture classification in relative phase domain

Relative phase is a newly developing technology for extracting features of images from phase domain and this paper studies a method of texture classification in relative phase domain. Because relative phase information can be obtained only in complex wavelet, we select DTCWT (Dual Tree Complex Wavelet Transform) and PDTDFB (Pyramidal Dual Tree Directional Filter Bank) to decompose images into different subbands at different levels and directions, and then the wavelet coefficients are mapped into relative phase domain. In relative phase domain, we calculate the frequent 2-itemsets and statistical characteristics mean and standard deviation of each subband as image features for texture classification. The experimental results show that our texture classification method has better performance in relative phase domain built from either DTCWT or PDTDFB.