LAUSR

In this paper, a novel feature descriptor named Z with Tilted Z Local Binary Pattern (Z⊕TZLBP) is proposed for extracting coral reef image features efficiently. The aim is to reduce LBP’s computational complexity by reducing the size of the feature vector. This is achieved in the proposed Z⊕TZLBP by dividing the neighborhood pixels into two non overlapped groups of Z and TZ (Tilted Z), and computing LBP wherein the centre pixel is also treated as one of the neighbors. KNN classification with four different distance metrices has been used for classification purpose. Metric F-measure is used to evaluate the performance of the proposed system. Experiments conducted with various coral image and video data sets show that the proposed feature descriptor outperforms Local Binary Pattern (LBP), Uniform Pattern, Center-Symmetric Local Binary Pattern and Orthogonal Combination of Local Binary Pattern and also guarantees accurate and efficient feature extraction.