Segmentation of spatially variant image textures

Spatially variant image textures are characterized by a smooth variation of their local spatial-frequency and/or orientation components. These textures are non-homogeneous in the sense of having a nonstationary local spectrum, while remaining locally coherent. Segmentation of image regions with spatially varying texture patterns is complicated by the lack of global homogeneity. In the proposed method, local spatial-frequency components are accurately estimated using Gabor wavelets outputs along with the absolute magnitude of the convolution of the input image with the first derivatives of the underlying Gabor function. Discontinuities in estimated local spatial-frequencies are interpreted as corresponding to texture boundaries. Results are presented that confirm the efficacy of the proposed approach.