摘要： 为提取有效的特征用于纹理描述和分类，提出一种融合局部二进制模式(LBP)和灰度共生矩阵(GLCM)的纹理特征提取方法。利用旋转不变的LBP算子处理纹理图像，得到LBP图像及其GLCM，采用对比度、相关性、能量和逆差矩描述图像的纹理特征。实验结果表明，与其他方法相比，该方法提取的纹理特征具有更强的纹理鉴别能力，平均分类正确率达到93%。

关键词: 纹理分析, 特征提取, Haralick特征, Gabor滤波器, 局部二进制模式, 灰度共生矩阵

Abstract: In order to extract effective features for texture description and classification, this paper proposes a texture feature extraction method fused with Local Binary Pattern(LBP) and Gray-level Co-occurrence Matrix(GLCM). The texture image is processed by rotation invariant LBP operator. The LBP image is obtained and its GLCMs are calculated. Contrast, correlation, energy and inverse difference moment are imposed for texture description. Experimental results show that, compared with other methods, the proposed method is more effective in texture feature extraction and the average classification accuracy reaches to 93%.

Key words: texture analysis, feature extraction, Haralick feature, Gabor filter, Local Binary Pattern(LBP), Gray-level Co-occurrence Matrix (GLCM)

中图分类号: 

• TP18