基于harris-like角点算子的指纹分离

Image separation based on harris-like operator

Harris-like角点算子用于检测图像中的交角,而重叠指纹图像则由于指纹与背景纹理的重叠产生了大量的额外的交角,因此为了实现指纹分离,首先采用梯度下降法使得指纹图像的harris-like角点算子最小,然后加入全变差调整,更好地指导分离过程。MATLAB仿真结果表明,此算法能够实现指纹分离。

Harris corner operator is used to detect the corners in the image, and overlapping fingerprint image produces many extra corners, so to separate the overlapping fingerprint image, first, minimize the harris-like corner operator： then total variation regulation is added to supervise the separation processing. MATLAB simulation indicates that this algorithm can separate the overlapping fingerprint image.