Segmentation of Fingerprint Image Based on Automatic-parameter Normalization
基于自动参数标准化的指纹分割方法

A combined method to segment fingerprint images,based on automatic-parameter normalization,is presented in this paper.Taking directional field and gray variance information in the fingerprint image into account,the method holds an efficient and robust feature.Compared with fixed-parameter normalization in former segmentation methods,the automatic-parameter normalization presented in this paper can normalize the fingerprint image to a maximum extent while not deteriorating any local image feature.