Modeling Micro-patterns for Feature Extraction

Currently, most of the feature extraction methods based on micropatterns are application oriented. The micro-patterns are intuitively userdesigned based on experience. Few works have built models of micro-patterns for feature extraction. In this paper, we propose a model-based feature extraction approach, which uses micro-structure modeling to design adaptive micro-patterns. We first model the micro-structure of the image by Markov random field. Then we give the generalized definition of micro-pattern based on the model. After that, we define the fitness function and compute the fitness index to encode the image’s local fitness to micro-patterns. Theoretical analysis and experimental results show that the new algorithm is both flexible and effective in extracting good features.