Gabor变换实现的难点是Gabor滤波器纽的参数选择。本文提出了一种适用于虹膜纹理特征提取的Gabor滤波器组参数选择方法。该方法根据图像分块确定Gabor滤波器的位置因子取值;借助海明距离均值曲线确定滤波器尺度因子;通过建立尺度因子与频率调制因子的关系,最终确定频率因子的取值。实验证明,依据该方法设计的滤波器,能有效提取虹膜纹理特征,得到较高识别准确率达到虹膜识别的目的。
The algorithm for Iris recognition based on Gabor transform is difficult to choose Gabor parameters. A method for parameters selection in Gabor filter which used in iris recognition is proposed. In this method, the position parameters can be selected according to partitions of image, the size parameters can be chosen by the curve of hamming distance; and the frequency parameters can be gotten based on the proportion to size parameters. With the Gabor filters designed by the method to iris feature extraction, experimental results show that it can extract effectively the characteristics of iris texture for iris recognition and has high accuracy.