Title :
Spatially enhanced local binary pattern
Author :
Kaimin Yu ; Zhiyong Wang ; Lei Yue ; Dagan Feng
Author_Institution :
Sch. of Inf. Technol., Univ. of Sydney, Sydney, NT, Australia
Abstract :
The local binary pattern (LBP) is one of the state-of-the-art texture descriptors for texture classification. However, the spatial distribution of the LBPs is not taken into account. Proposed is a novel texture descriptor, namely the spatially enhanced LBP, to encapsulate the spatial distribution of the LBPs with a shape context based method. Also, the shape context based method is adapted to be rotation invariant. Experimental results on three widely used benchmark datasets show that the proposed descriptor outperforms the original LBP descriptor.
Keywords :
image classification; image texture; LBP descriptor; LBP spatial distribution; context based method; image texture classification; spatially enhanced LBP; spatially enhanced local binary pattern; texture descriptors;
Journal_Title :
Electronics Letters
DOI :
10.1049/el.2012.2054