Description

The aim of this project is used to classify texture based on scale, rotation invariant and sub uniform based circular shift of lbp. In an existing system, classify texture based on local binary pattern which is used local neighbourhood property of pixel. If the image is rescaled, the neighbourhood pixels can be changed, so the original lbp feature also changed. We propose a system, scale adptive texton which represents adaptive circularly neighbour hood region of pixel. Implement the system the feature of scale and rotion invariant of lbp pattern for pixel of texture, and used the subuniform pattern for each uniform pattern of pixels for texture description and then shifted circularly to classify the texture. Experimental results demonstrate a good discrimination capability of the proposed scale- and rotation-invariant LBP in texture classification.