Compact multi-dimensional LBP features for improved texture retrieval

Content-based image retrieval has become an important research area and consequently well performing retrieval algorithms are highly sought after. Texture features are often crucial for retrieval applications to achieve high precision, while local binary pattern (LBP) based texture descriptors have been shown to work well in this context. LBP features decsribe the texture neighbourhood of a pixel using simple comparison operators, and are often calculated based on varying neighbourhood radii to provide multi-resolution texture description. Furthermore, local contrast information can be integrated into LBP leading to LBP variance (LBPV) features. In conventional LBP methods, the histograms corresponding to different radii are simply concatenated resulting in a loss of information between different resolutions and added ambiguity. In this paper, we show that multi-dimensional LBP and LBPV histograms, which preserve the relationships between scales, provide improved texture retrieval performance. To cope with the exponential increase in terms of feature length, we show that application of principal component based feature reduction leads to very compact texture descriptors with high retrieval accuracy.