Abstract: For the problem that the sole remaining fingerprint cannot be segmented easily，an application segmentation algorithm is proposed for remaining fingerprint based on quality estimation.The algorithm firstly selects effective area，the extent of wet and dry，position offset as estimation factor，considering all estimation factors in order to obtain the total quality point，secondly the ultimate estimation result is obtained based on above，finally isolated block is removed by calculating the area of the remaining fingerprint.The algorithm has following improvements：On the one hand，it determines the quality before the pre-process，so it controls input fingerprint quality well.On the other hand，it calculates remaining fingerprint area by the Freeman chain code.Experimental results show that the algorithm not only can filter unqualified fingerprint image effectively，but also can segment the remaining fingerprint from the foreground.

Key words: fingerprint segmentation, quality estimation, remaining fingerprint, estimation factor, freeman code

摘要： 针对孤立的残留指纹往往具有清晰的纹路而不易被分割的问题，提出一种质量评估在残留指纹分割中的应用算法。该算法首先选取指纹图像的有效面积、干湿程度、位置偏移量作为评价因子，然后综合考虑各种评价因子求取总的质量分得到最终的评价结果；最后通过计算面积割除孤立的残留指纹。该算法的改进之处：首先在指纹预处理之前先判断指纹的质量，很好地控制了录入指纹的质量；其次使用Freeman链码表示法计算残留指纹的面积。实验结果表明，该算法不仅对不合格的指纹有较好的过滤效果，而且可以有效地分割残留指纹。

关键词: 指纹分割, 质量评估, 残留指纹, 评价因子, Freeman码