基于扩展的LBP算子地板块纹理分类研究

Research of the Plate Texture Classification Based on Extended LBP Operator

针对地板块纹理分类问题,首次引入局部二值模式LBP算子提取地板块纹理特征。本文提出一种基于扩展的LBP算子地板块纹理分类方法。在阐述LBP算子基本原理的基础上,采用中值滤波法去除图像噪点,以减少噪点对图像纹理特征的干扰,将均匀模式和旋转不变性与传统的LBP算子相融合,提取地板块纹理特征,经KNN分类器实现地板块纹理分类。实验结果表明该方法识别速度快、辨识准确率高,优于传统的灰度共生矩阵法,为地板块纹理分类的研究提供新思路。

To the question of plate texture classification, the paper firstly imports LBP operator to extract texture feature, and proposes the plate texture classification algorithm based on extended LBP operator. Based on the principle of LBP operator, this paper adopts median filter to wipe off the image noise in order to reduce the interference, and combines the rotation invariance, uniform pat- tern with traditional LBP operator to extract plate texture feature. Finally, plate texture classification is achieved using KNN classifier. The experimental results show that the algorithm has the advantages of faster speed and higher accuracy of recognition than the tradition- al Grey-Level Co-occurrence Matrix. The algorithm provides a new thought for the research on the plate texture classification algorithm.