基于非线性扩散滤波的指纹增强算法

根据指纹图像的特性提出了一种利用非线性扩散滤波增强指纹图像以提取细节点的方法.该算法利用指纹图像的结构张量构造扩散张量.使得滤波沿着指纹脊线方向扩散.在求解非线性扩散滤波中最重要是怎样选取迭代时间和迭代步长.为了解决这两个问题,根据扩散张量在大小为(2n+1)×(2n+1)窗口上的离散形式,分解扩散滤波为4n个方向的和,然后利用托马斯-高斯消元法在每个方向快速求解非线性滤波后的指纹增强图像.该算法的优点是可以有效地连接断裂脊线,而不会改变指纹脊线的连续性和奇异性.在NIST27数据库及FVC2002数据库上的实验结果表明,基于非线性扩散滤波的指纹增强算法可以有效地提高提取细节点的性能和指纹匹配率.

Fingerprint Enhancement Algorithm Based on Nonlinear Diffusion Filter

This paper proposes a nonlinear diffusion filter to enhance fingerprint imageand extract the minutiae according to the characteristic of the fingerprint. The algorithmuses the structure tensor to construct diffusion tensor, and makes the filter diffuse alongthe ridge direction. In order to improve the computing efficiency, according to the discreteformat of the diffusion tensor in a window size of (2n+1)×(2n+1), the diffusion filter issplit into 4n directions, after that, in each direction the fingerprint enhancement image iscomputed with Thomas-Gaussian elimination algorithm. The advantage of this algorithmis to connect the breaking ridges in an effective way, and without changing the continuityand singularity of the ridge. Experiments on NIST27 and FVC2002 databases show that theenhancement algorithm based on nonlinear diffusion filter could improve the performance ofextracting minutiae and correct fingerprint matching ratio.