计算机工程 ›› 2011, Vol. 37 ›› Issue (17): 208-210.doi: 10.3969/j.issn.1000-3428.2011.17.070

• 图形图像处理 • 上一篇    下一篇

一种指纹图像的快速分割方法

梁媛媛,温佩芝,黄文明,任亚恒,朱叶青,罗桂海   

  1. (桂林电子科技大学计算机与控制学院,广西 桂林 541004)
  • 收稿日期:2011-03-24 出版日期:2011-09-05 发布日期:2011-09-05
  • 作者简介:梁媛媛(1984-),女,硕士研究生,主研方向:图像处 理;温佩芝,教授、博士后;黄文明,副教授;任亚恒、朱叶青、罗桂海,硕士研究生
  • 基金项目:
    广西研究生教育创新计划基金资助项目(20091059508 12M27)

Fast Segmentation Method of Fingerprint Image

LIANG Yuan-yuan, WEN Pei-zhi, HUANG Wen-ming, REN Ya-heng, ZHU Ye-qing, LUO Gui-hai   

  1. (School of Computer and Control, Guilin University of Electronic Technology, Guilin 541004, China)
  • Received:2011-03-24 Online:2011-09-05 Published:2011-09-05

摘要: 利用数学形态学变形虫法进行指纹图像分割时计算量大,导致分割速度变慢。针对该问题,提出一种将数学形态学变形虫法和离散小波变换相结合的指纹图像快速分割方法。先对原始图像进行小波分解,再对获取的低频图像采用形态学变形虫分割。实验结果表明,该方法在保证良好分割精度的同时分割速度较快。

关键词: 数学形态学, 变形虫, 自适应分割, 离散小波变换, 指纹分割

Abstract: Mathematical morphology amoebas method can segment the fingerprint, but the run speed is slow for the large computation. Aiming at this problem, this paper proposes a fast segmentation method of fingerprint image based on the mathematical morphology amoebas and Discrete Wavelet Transform(DWT). It uses the wavelet decomposition, and segments the low frequency image using morphology amoebas. Experimental results show that this method ensures good segmentation accuracy and realizes the fast segmentation of fingerprint image.

Key words: mathematical morphology, amoebas, self-adaptive segmentation, Discrete Wavelet Transform(DWT), fingerprint segmentation

中图分类号: