大鱼际掌纹图像特征提取方法研究

Study on Feature Extraction of Thenar Palmprint Image

为了更客观更准确的判断出患者的大鱼际掌纹的级数,可以采用图像处理技术对大鱼际掌纹进行预处理、特征提取和分类,以实现大鱼际掌纹的量化与客观识别.文中提出一种基于改进的二维主成分分析技术(2DPCA)再结合Gabor滤波的特征提取方法.以定位分割并经增强处理的大鱼际掌纹图像为基础,获得图像的特征矩阵,作为下一步量化分级的特征输入量.仿真实验结果表明该方法是适用有效的.

In order to more objectively determine more accurately in patients with thenar palmprint of the series,image processing techniques can be used on the thenar palmprint for preprocessing,feature extraction and classification.This paper presents an improved two-dimensional principal component analysis（2DPCA） combined with Gabor filter feature extraction method,Image processing techniques can be used on the thenar palmprint for preprocessing,feature extraction and classification.The experimental results shows that the method which use combined algorithm can give the best performance.