Author(s): Keerthana.A , Ashwin M

This paper attempts in exploring the survey of various methodology for Texture classification. Now-a-days due to high availability of computing facilities, large amount of data in electronic form is generated. The data generated is to be analyzed so as to maximize the benefits, for intelligent decision making. If the data generated is in the structured form then large amount of work in analyzing such structured data is available. The survey has taken for statistical approach for learning joint distribution of filter responses for image distribution mapping, texton distribution and comparing the distribution by learnt models in classification. Then Feature extraction stage for set of small random feature are extracted from local image patches, here bag of words are used to perform texture classification. It is focused on binary images texture pattern and investigate a class of texture descriptor that characterize the probability of occurrence of the patterns associated to the neighbourhood of given size and shape. HEP(Histogram of equivalent patterns) combine the CLBP(Completed Local Binary Patterns) and ILTP(Improved Local Ternary Patterns).When texture analysis history has algorithm range from random field models to multiresolution representation based on Gabor filters. Next survey says Noise Resistant LBP(NR-LBP) and Extended Noise Resistant LBP(ENRLBP)is used to reduce the noise. Finally the survey says BRINT texture classification is better than all other classification. So BRINT Texture Classification can be used for Snake Texture classification and for future work it can be at Face Recognition.