Binary and ternary patterns for image classification

There is a growing interest in CBIR (Content based image retrieval) because of the limitations inherent in metadata-based systems, as well as the large range of possible uses for efficient image retrieval. Textual information about images can be easily searched using existing technology, but requires humans to personally describe every image in the database. Texture content plays a prominent role in classification of the image. In this paper we use binary and ternary patterns for the texture analysis which has high tolerance against illumination variation. The experimental results are evaluated for different binary patterns like local binary pattern (LBP), simplified binary pattern (SLBP) and local line binary pattern (LLBP) and Ternary pattern (TP).