指纹图像增强算法研究

指纹识别是一种重要的利用生物特征的个人身份识别技术。指纹图像增强是指纹识别系统中的必要步骤。针对指纹图像中的噪声干扰,提出一种基于指纹局部方向的指纹图像增强算法。首先利用图像的局部灰度分布特点计算局部方向角;然后利用局部正交方向轴线上的灰度投影量估计局部脊线频率;最后利用自适应的具有方向选择和频率选择的Gabor滤波器对原图像进行滤波增强。实验结果表明,此算法能有效地抑制噪声,而保持真实的脊线结构。

A Fingerprint Image Enhancement Algorithm

Fingerprint image enhancement is a key procedure in an automatic fingerprint identification system.An algorithm based on the local direction of fingerprint is proposed to improve the quality of the fingerprint image in noise background.Firstly,local ridge orientation is calculated by making use of the local gray distribution.Then,the gray projection on the local orthogonal axis was used for estimating the local ridge frequency.At last,a set of adaptive Gabor filter which have both orientation-selective and frequency-selective properties was used to enhance the fingerprint image.Simulation shows this algorithm can suppress noise effectively and preserve true ridge structure.