一种新的指纹图像复合分割方法

A New Compound Method on Fingerprint Image Segmentation

本文在分析了方差法,方向图法和角部灰度均值法的优缺点后,将它们合理的结合,提出了一种新的基于方向图和角部灰度均值的复合分割方法。本算法的优点是充分利用了指纹图像特有的方向性和灰度特性,对于方向图法和方差方法单独应用时无效的情况都适用,能准确地将背景区域从指纹图像中分离出来,从而提高了分割的精确度。

This paper has been analyzing the advantages and disadvantages of the gray variance ,directional image and the angle gray average segmentation method, and their reasonable union, then it presents a new approach namely uses one kind based on the directional image and angle gray average compound segmentation method. The advantages of this method is fully utilizes the gray and direction characteristics of the fingerprint image, regarding when directional image and gray variance alone application the invalid situation all is suitable, thus it can precisely separate the background area from the fingerprint image and improve the segmentation precision.