In this paper,pattern recognition feature extraction based on discrete wavelet transform(DWT)and local binary pattern(LBP)integration method,made the two-dimensional image after DWT transform low frequency,high frequency wavelet co?efficients,used of LBP extract DWT feature vector image obtained by KNN classifier pattern classification.In Brodatz,Outex,UMD Gallery experimental results show that better than DWT,LBP resolution,with good prospects for the study.%论文提出了基于离散小波变换(DWT)和局部二进制模式(LBP)融和的模式识别特征提取方法,将二维图像经过DWT变换后得到低频,高频小波系数,利用LBP提取DWT得到图像的特征向量,通过KNN分类器来模式分类,在Brodatz, Outex,UMD图库中实验,实验结果表明,优于DWT,LBP分辨率,有良好的研究前景.
展开▼
