利用2D-Gabor滤波器提取纹理方向特征的虹膜识别方法

Iris recognition method based on iris texture orientation feature extracted by 2D-Gabor filter

目前基于2D-Gabor滤波器的虹膜识别算法主要是使用虹膜的相位信息或能量信息作为特征,但在虹膜可用区域减少时这些算法的识别效果明显下降。虹膜纹理除具有上述特征外,还有很强的方向性。在分析了2D-Gabor滤波器的方向和频率选择性后,提出了一种利用2D-Gabor滤波器提取纹理方向特征的虹膜识别方法。实验表明该方法提取的虹膜纹理方向特征可以在较小区域内提取出足够丰富的可区分性特征,实现高准确性的虹膜识别,说明方向特征是一种有效的虹膜识别特征。

The algorithms of iris recognition based on the 2D-Gabor filter usually used the phase features or energy features of the iris, but when the usable area of the iris decreased, the recognition rate of these algorithms obviously would go down. Iris not only has phase features or energy features, but also has orientation features. After analyzing the directional and frequency selection of 2D-Gabor filter, proposed an iris recognition method based on iris orientation feature using 2D-Gabor filter. Experimental results show that these orientation features of the iris texture which are extracted in very small region of iris can achieve an iris recognition which has high accuracy, this result demonstrates that orientation features of the iris texture is an effective features which is used in iris recognition.