扩展统一局部二值模式及图像纹理特征提取  

Extended Uniform Local Binary Patterns and its application for image texture feature extraction

作　　者：李江美 陈熙 LI Jiangmei;CHEN Xi(School of Big data and Computer Science,Guizhou Normal University,Guiyang 550025,Guizhou,China)

机构地区：[1]贵州师范大学大数据与计算机科学学院,贵阳550025

出　　处：《智能计算机与应用》2025年第1期151-157,共7页Intelligent Computer and Applications

基　　金：国家自然科学基金(61762022)。

摘　　要：统一局部二进制模式(ULBP)及其许多变体,已显示出对纹理分类的有效性。然而,这些ULBP方法中的大多数专注于编码中心像素与其相邻像素间的统一模式特征。因此存在无法捕捉非统一模式发生的像素间图像特征,忽略了不同邻域半径下像素间的相互作用这两个主要问题。针对于此,本文提出了扩展统一局部二值模式(EULBP)。EULBP统计图像相邻像素间的非统一模式,并逐步编码相邻采样点间的非统一模式;其次,在不同邻域下提取图像特征进行融合;最后,使用直方图交叉距离计算特征向量的相似度,得到在不同数据集中的识别率。经实验,验证了该算法的有效性。The Uniform Local Binary Pattern(ULBP) and its many variants have shown effectiveness in texture classification.However,most of these ULBP methods focus on the uniform pattern features between the coding center pixel and its adjacent pixels.Therefore,they cannot capture the image features between pixels that occur outside the non-uniform mode,and they ignore the interaction between pixels at different scales.In response to these issues,the Extended Uniform Local Binary Pattern(EULBP) is proposed.EULBP counts the non-uniform patterns between adjacent pixels of the image,and gradually encodes the non-uniform patterns between adjacent sampling points.Secondly,extract image features at different scales for fusion.Finally,the similarity of feature vectors is calculated using the cross distance of histograms to obtain recognition rates in different datasets.The effectiveness of the algorithm has been verified through experiments.

关 键 词：扩展统一局部二值模式 图像特征提取 特征融合 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]