Rotation Invariant Texture Classification Using Texton Co-occurrence Matrix Derived from Texture Orientation 5.1 Rotation Invariant Texture Classification Based on Texton Co-occurrence Matrix

In the previous chapter, an integrated approach for texture classification using ILCLBP-T is proposed. In continuation to that, the present chapter derived a new co-occurrence matrix based on textons and texture orientation for rotation invariant texture classification of 2D images. The new co-occurrence matrix is called as Texton and Texture Orientation Co-occurrence Matrix (T&TO-CM). The Co-occurrence Matrix (CM) characterizes the relationship between the values of neighboring pixels, while the histogram based techniques have high indexing performance. If the CM is used to represent image features directly, then the dimension will be high and the performance is decreased. On the other hand, if histogram is used to represent image features, the spatial information will be lost. In order to overcome this and to precisely represent the spatial correlation of color and texture orientation, the present thesis integrates the advantages of co-occurrence matrix and histogram by representing the attribute of co-occurrence matrix using histogram. The proposed method is computationally attractive as it computes different features with limited number of selected pixels.