Method for fingerprint image binarization using difference
指纹图像差分二值化算法

Binarization is one important step in automatic digital fingerprint identification.Selection of threshold will greatly influence the effect of binarization,as well as the number of valid features.In this paper,we proposed a new phase remove smoothing and difference method for fingerprint image binarization,which moves the threshold of background and foreground toward 0 point.Experiments show that this method can reduce the sensitivity of threshold selection,and at the same time can keep the valid fingerprint features after binarization.