基于能量最大响应方向的虹膜识别算法

Iris recognition method based on maximal responding orientation of energy

为提高虹膜识别算法的正确识别率,给出了基于能量最大响应方向的虹膜识别算法。该算法首先对人眼图像进行预处理得到归一化虹膜图像,并依据归一化虹膜纹理的分布特点,选择多尺度多方向滤波技术对归一化虹膜图像进行分解;然后提取最优分解尺度下像素点的能量最大响应方向数,进行二进制虹膜编码和去噪;最后通过计算两个虹膜编码间的加权海明距离和衡量两者的差异性,给出识别结果。实验证明,在等误率为0.1176时,正确识别率达到99.997%。与经典的虹膜识别算法相比较,该算法具有更高的识别率。

In order to improve the accuracy of iris recognition further,this paper proposed an iris recognition method base on the maximal responding orientation of energy.Firstly,obtained the normalized iris images by preprocessing eye image,selected multiscale and multidirectional filter by the feature of iris texture,decomposed the normalized iris image through the filter.Se-condly,extracted and encoded the maximal responding orientation on the pixel point into a simple binary iris code,performed denosing at the same time. The last,measured the difference of tow iris code through weighted hashiymming distance and gave the recognition results.The experimental results demonstrate that the proposed method has the equal error rate of 0.1176% and the correct recognition rate of 99.997%. Compared with the classical methods, the proposed method has a higher accuracy.