指纹自动识别中图象分割方法的研究

图象分割的一般方法是基于图象灰度特性的，但这种方法不适合于低对此度或增强后的图象，对于指纹图象的分割，基于方向信息的方法更为适合，但它也有一些限制。而将两种方法合理的结合，则可较好地解决这些问题，实验表明，这种复合方法对于指纹图象的处理是很有效的。

Research on Image Segmentation in Automatic Fingerprint Identification

Image segmentation is usually based on gray-level- statistics of images. However, it is not valid for enhanced images or those with low contrast. The method based on direction information is suitable for fingerprint segmentation but still with some limitations. This paper proposes a composite method, which takes advantages of both direction and variance. The experimental result shows that this method is more effective and robust for fingerprint segmentation.