计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (5): 206-208.

• 工程与应用 • 上一篇    下一篇

基于改进的GVF Snake模型CT图像轮廓提取技术

张伟伟,赵歆波,张 涛,王 凯   

  1. 西北工业大学 现代设计与集成制造教育部重点实验室,西安 710072
  • 收稿日期:2007-06-06 修回日期:2007-08-22 出版日期:2008-02-11 发布日期:2008-02-11
  • 通讯作者: 张伟伟

Technology of contour extraction based on improved GVF Snake algorithm

ZHANG Wei-wei,ZHAO Xin-bo,ZHANG Tao,WNAG Kai   

  1. The Key Lab of Contemporary Design and Integrated Manufacturing Technology,Northwestern Polytechnical University,Xi’an 710072,China
  • Received:2007-06-06 Revised:2007-08-22 Online:2008-02-11 Published:2008-02-11
  • Contact: ZHANG Wei-wei

摘要: CT图像轮廓提取是逆向工程中的关键技术之一,由于基于边界检测算法很难提供一条连续的无间断的边界轮廓。采用一种改进的GVF Snake算法,首先通过人工取点,B样条拟合的方法,提取其初始轮廓,再利用改进的GVF Snake方法进行边界轮廓逼近,以获得比较精确的叶片边界轮廓,克服了传统活动轮廓模型梯度势能场捕捉范围小的缺点,并通过实验证明了该方法的有效性。

关键词: 活动轮廓模型, 轮廓提取, GVF Snake, CT图像

Abstract: In reverse engineering,the contour extraction of CT image is one of the key Technologies.It is difficult for algorithms based on edge detection to provide a continuously solid edge contour.This paper adopts an improved algorithm of GVF Snake.First,it extracts original contour by man-made sampling and the method of B spline drafting,then,recycles improved GVF Snake for approach of edge contour,obtaining relatively precise edge counter of blade image.It conquers the error of tradition Snake which catches area small,and conforms to this algorithm effectively.

Key words: snake model, contour extraction, GVF Snake, CT image