作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (9): 196-197,200. doi: 10.3969/j.issn.1000-3428.2011.09.068

• 人工智能及识别技术 • 上一篇    下一篇

一种新型GVF-Snake模型

杜海龙,吴锡生   

  1. (江南大学物联网工程学院,江苏 无锡 214122)
  • 出版日期:2011-05-05 发布日期:2011-05-12
  • 作者简介:杜海龙(1986-),男,硕士研究生,主研方向:人工智能,模式识别;吴锡生,教授、博士

New GVF-Snake Model

DU Hai-long, WU Xi-sheng   

  1. (School of IoT Engineering, Jiangnan University, Wuxi 214122, China)
  • Online:2011-05-05 Published:2011-05-12

摘要: 针对GVF-Snake模型对虚假边界敏感的缺陷,在原模型基础上提出一种有效梯度向量流外力场的概念,并引入自适应性气球力。有效梯度向量流外力场是对GVF模型外力场的优化,它将演化曲线过程中曲线和梯度向量之间的夹角作为模型的一个新因素,降低演化过程中噪声和虚假边界对演化曲线的干扰。自适应性气球力的引入加快了演化曲线收敛速度。实验结果表明,与原模型相比,新模型曲线在演化过程中对噪音和虚假边界的抗性得到提高。

关键词: 主动轮廓模型, 动态法梯角, 有效梯度向量流, 相容性角

Abstract: Aiming at the drawback that GVF-Snake model is susceptible to false boundary, and on the base of original model, this paper proposes a novel concepts called Valid-GVF model, and introduces adaptivity balloon force into the model. Valid-GVF model optimize the outside force field of GVF model, factoring the angle between the curve and gradient vector into the model, which reduces the interference produced by noise and false boundary. Adaptivity balloon force can speed up convergence of the evolution curve. Comparative experimental results show the resistance to noise and false boundary of the new model is enhanced.

Key words: active countour model, dynamic method gradient angle, valid gradient vector flows, consistency angle

中图分类号: