Texture feature extraction of coal-rock image based on variogram and local variance image
HUANG Lei
GUO Chaoya
Abstract：In view of problems of low classification accuracy and algorithmic running efficiency and poor robust property of rotation texture recognition existed in local binary patterns for texture feature extraction of coal-rock,a texture feature extraction algorithm of coal-rock image based on variogram and local variance image was proposed.Firstly,local variance image was got by calculating local variance with pixel by pixel in theoretic framework of local binary patterns.Then,the variogram vectors with different direction were calculated by variogram in local variance image.Finally,combination variogram vectors were taken as the texture feature,classification and recognition of texture of coal-rock was realized combining the texture feature and local binary patterns feature.Experiment results show that the algorithm can effectively extract spatial distribution information of the local variance image,realize information reuse missed by local binary patterns,and its classification results are better than other algorithms of texture extraction based on local binary patterns,the classification precision reaches 86 ％.
Keywords：coal miningcoal and rock recognitionlocal variance imagevariogramtexture feature extractionlocal binary pattern
Publication Date：2018-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 62-68 )