Classification of High Spatial Resolution Image Using Multi Circular Local Binary Pattern and Variance.

High spatial resolution satellite image comprises of textured and non-textured regions. Hence classification of high spatial resolution satellite image either by pixel-based or texture-based classification technique does not yield good results. In this study, the Multi Circular Local Binary Pattern (MCLBP) Operator and variance (VAR) based algorithms are used together to transform the image for measuring the texture. The transformed image is segmented into textured and non-textured region using a threshold. Subsequently, the original image is extracted into textured and non-textured regions using this segmented image mask. Further, extracted textured region is classified using ISODATA classification algorithm considering MCLBP and VAR values of individual pixel of textured region and extracted non-textured region of the image is classified using ISODATA classification algorithm. In case of non-textured region MCLBP and VAR value of individual pixel is not considered for classification as significant textural variation is not found among different classes. Consequently the classified outputs of non-textured and textured region that are generated independently are merged together to get the final classified image. IKONOS 1m PAN images are classified using the proposed classification algorithm and found that the classification accuracy is more than 84%.