关键词: 指纹分割, 方差, 均值, 纹理特征, 自适应阈值

Abstract: Fingerprint segmentation is an important step in the process of image preprocessing.The traditional fingerprint image segmentation algorithms are analyzed and compared，and a new adaptive fingerprint segmentation algorithm based on the variance based fingerprint segmentation algorithm is proposed.In the new algorithm，the texture feature which is represented by the uniform ridge and valley intervals is used to compute mean and threshold adaptively.The problem of determining threshold experientially is avoided.The experimental results show that the effect of segmentation of the proposed algorithm is better than the variance based fingerprint segmentation，and it also has stronger resistance to noise and higher adaptability to different types of fingerprint images.

Key words: fingerprint segmentation, variance, mean, texture feature, adaptive threshold