指纹图像分割方法综述

在自动指纹识别系统中,精确的指纹图像分割可以加快后续处理速度,提高识别的准确性。按照图像特征的定义范围,将指纹图像分割方法分为基于像素特征、块特征和图像全局特征的方法。对每类方法分别从特征层面与分类器层面进行了归纳总结,并简要分析了3类方法的分割错误率与时间复杂性。最后指出了目前指纹分割算法中存在的主要问题和未来的研究方向。

A review of fingerprint image segmentation methods

In an automatic fingerprint identification system, precise segmentation of fingerprint image contributes to speeding up subsequent processes as well as improving the recognition accuracy. Fingerprint segmentation methods can be roughly divided into three groups, including the methods based on pixel features, the methods based on block features and the methods based on global features. Here each group was introduced according to the used features and in turn classifiers. Segmentation accuracy and time complexities of the methods were also provided for brief comparison. In conclusion,  the remaining problems and list some future research directions were detailed.