【摘要】 该文提出了一种高精度的基于虹膜的身份识别的方法。该方法包括虹膜图像的预处理、虹膜的特征提取和虹膜的匹配与识别三个部分。虹膜图像的预处理包括虹膜图像的定位、归一化、增强和去噪。利用虹膜图像提供的丰富的纹理信息,特征提取采用Daubechies-2小波变换的纹理分析方法。匹配采用k近邻分类器。该算法对光照和噪声不敏感,且具有平移、缩放、旋转不变性。由实验结果表明,本文提出的方法提高了虹膜识别的有效性和准确性,是一种实用的身份识别方法。更多还原

【Abstract】 High accuracy arithmetic based on iris patterns is proposed for personal identification. The method consists of three major components: image preprocessing,, feature extraction and matching. It is composed of iris localization, iris normalization, iris image enhancement and denoising in image preprocessing. The abundant texture information of the iris is employed and Daubechies-2 wavelet transform is used in the extraction of iris feature. Knn classifier is adopted while iris is matched. The method is not sensitive to illumination and noise and translation, zoom and rotation invariant. The experimental results show that the arithmetic presented in this paper improves the efficiency and accuracy of iris recognition and is a practical arithmetic for personal identification.更多还原