Rotation Invariant Texture Classification using Fuzzy Local Texture Patterns
Easwar Srinivasan, K. Ramar, A. Suruliandi · 2012
Texture is one of the basic image properties, which is useful in many computer vision applications. Though there are many approaches to describe textures, no particular representation has been shown to give great results on a wide range of textures due to the variations in texture properties and vagueness in describing them. Fuzzy logic is an excellent tool to handle uncertainties which arise due to ambiguous or incomplete information. A fuzzy logic based texture descriptor Fuzzy Local Texture Patterns (FLTP) has been proposed in our earlier work and in this paper, rotation invariant property of the FLTP descriptor is demonstrated with texture classification experiments. The results show that the FLTP method is robust against rotational variations.