摘要
提出了1种针对煤岩显微组分的RILBP-GLGM纹理特征提取算法。利用RILBP算子获取原图像的RILBP图谱,获取该图谱的灰度共生矩阵,提取4个方向上的能量、熵、惯性矩、相关性特征构成图像纹理特征,利用多分类支持向量机对图像纹理特征进行训练和分类。结果表明,基于RILBP-GLCM算法的煤岩显微组分纹理特征分类是有效的。
A RILBP-GLCM texture feature extraction method for coal rock macerals is presented. Operating the origin images with RILBP to get a hapmap. Getting the statistics GLCM and extracting the value of energy, entropy, moment of inertia and correlation with four orientation, training and classifying the image texture features by MSVM. Experimental results show that the effectiveness by using RILBP-GLCM method to extract coal rock macerals texture feature.
出处
《煤炭技术》
CAS
北大核心
2017年第3期142-144,共3页
Coal Technology
基金
山西省自然科学基金(2014011013-1)
