Local Binary Pattern Technique For Soil Texture Classification

Abstract

In Image processing programs texture classification is one of the giant and beneficial challenge. Many texture models were implemented over the last few years and Local Binary Pattern (LBP) approach is one of the easy and dynamic method amongst them. A quantity of references to the LBP technique have been additionally supplied, but the problem stays difficult in feature vector generation and evaluation. Parameters like gray-scale variation, rotation variant, illumination version and noise effectively handled by way of texture version. We proposed a revolutionary model (ERLBP) that allows classifying and characterizing texture in digital images. The outcomes are in comparison with other extensively used texture models by applying classification tests to various texture images. Experimental consequences shows that the proposed texture model (ERLBP) is robust to gray-scale variation which improves its discriminative functionality and decreases the noise.