耳纹图像特征的三维识别技术研究  

Research on 3D recognition technology of ear print image feature

作　　者：尤鸿霞 黄艳艳[2] 

机构地区：[1]江苏工程职业技术学院,江苏南通226007 [2]南通大学理学院,江苏南通226007

出　　处：《现代电子技术》2017年第22期91-94,共4页Modern Electronics Technique

基　　金：国家自然科学基金(11604163)

摘　　要：为了提高对耳纹图像的三维识别能力,提出一种基于脊波变换的耳纹图像分块特征提取和三维识别技术。采用二维傅里叶变换方法进行耳纹图像的亚像素级特征分解,结合区域网格径向划分方法进行图像分割,对耳纹图像的子块像素点进行自适应融合滤波,实现图像降噪,使用脊波变换进行耳纹图像三维特征重构,使用多尺度Retinex信息增强方法输出图像的辨识特征点,实现耳纹图像三维识别。仿真结果表明,采用该算法进行耳纹图像特征识别的精度较高,图像输出质量较好。In order to improve the 3D recognition ability for the ear print image,an ear print image's block feature extraction and 3D recognition technology based on ridgelet transform is proposed.The two-dimensional Fourier transform method is used to decompose the sub-pixel feature of the ear print image,and combined with the regional grid radial partition method for image segmentation.The subblock pixel points of the ear print image are performed with the adaptive fusion filtering to denoise the image.The ridgelet transform is adopted to carry out the 3D feature reconstruction of the ear print image.The multi-scale Retinex information enhancement method is employed to output the identification feature points of the image to realize the ear print image 3D recognition.The simulation results show that the algorithm has high precision for ear print image feature recognition and high image output quality.

关 键 词：耳纹图像 特征提取 三维识别 脊波变换 图像分割 

分 类 号：TN919-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]