Volume 10 Issue 1 - March 2018

  • 1. Texture analysis and clustering with elbp-var and mean shift for classification of high resolution images

    Authors : Sourabh Singh, Debasish Chakraborty

    Pages : 27-34

    DOI : http://dx.doi.org/10.21172/1.101.05

    Keywords : High resolution image,clustering,Mean Shift,IKONOS,LBP, classification

    Abstract :

    In this study, an enhanced local binary pattern (ELBP) operator is evolved for quantifying spatial structure. Variance (VAR) is used for measuring contrast around the pixel of the image. Thereafter, the quantified ELBP and VAR value are used together to transform the image for measuring the texture. The Mean-Shift based two-step iterative procedure for clustering the transformed image is adapted to (i) identify the range of texture that is densely occupied in the kernel (ii) partition the textures into a cluster that matches with the range. Subsequently similar type of clusters are grouped together to get classified image. Texture values [noise or not associated with the other cluster] are clubbed to a nearest possible cluster using the contextual clustering. IKONOS 1m PAN images are classified using the proposed clustering algorithm and found that the classification accuracy is more than 89%.

    Citing this Journal Article :

    Sourabh Singh, Debasish Chakraborty, "Texture analysis and clustering with elbp-var and mean shift for classification of high resolution images", https://www.ijltet.org/journal_details.php?id=928&j_id=4373, Volume 10 Issue 1 - March 2018, 27-34, #ijltetorg