【摘要】 在刑侦应用中,常需要将现场得到的多个重叠在一起的指纹分离开来。论文提出一种基于分区模板的重叠指纹分离方法。不同类型指纹脊线走向具有特定的模式,定义与每种类型指纹对应的指纹分区模板。这样,将分区模板施用于所对应类型的指纹后,每一分区内的指纹脊线成为走向基本一致的平行曲线。因而采用具有方向和频率选择性的Gabbor滤波器,选择合适的参数,可提取各个区内所要获得的指纹脊线。通过对人工合成图像和现场采集指纹图像的实验结果表明,提出的方案可有效地提取重叠在一起的指纹,保留绝大多数指纹特征点。因此,该文的方案较好地解决了指纹预处理中很少涉及,但在实践中却经常遇到的问题。更多还原

【Abstract】 There is a need for crime reconnaissance to extract individual fingerprints from an image with several over-lapped fingerprints acquired in the crime commitment zone.However,little literature addresses such a need.A scheme for separating the overlapped fingerprints is presented in this paper.Noting the fact that the orientation fields of six pre-specified types of fingerprint ,partition masks corresponding to the types are defined.Partitioning the wanted fingerprint based on the corresponding mask,the ridge and furrow in each partitioned sector present a consistent direction and have a certain spatial frequency.Then,the wanted fingerprint is obtained by applying Gabor filters with properly tuned direction and frequency parameters to all the sectors.Experiments on a synthetic image and a fingerprint image collected from crime scene show that our scheme is capable of extracting overlapped fingerprints effectively,and preserving most of fingerprint minutiae.That is,our scheme can solve the problem quite well.更多还原