计算机工程 ›› 2010, Vol. 36 ›› Issue (3): 213-214,.doi: 10.3969/j.issn.1000-3428.2010.03.071

• 人工智能及识别技术 • 上一篇    下一篇

一种改进的Bayes抠图算法

吕巨建1,战荫伟2   

  1. (1. 广东工业大学信息工程学院,广州 510006;2. 广东工业大学计算机学院,广州 510006)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2010-02-05 发布日期:2010-02-05

Improved Bayes Matting Algorithm

LV Ju-jian1, ZHAN Yin-wei2   

  1. (1. Faculty of Information Engineering, Guangdong University of Technology, Guangzhou 510006; 2. Faculty of Computer, Guangdong University of Technology, Guangzhou 510006)
  • Received:1900-01-01 Revised:1900-01-01 Online:2010-02-05 Published:2010-02-05

摘要: 针对传统Bayes抠图算法容易在生成的不透明度通道上产生脉冲噪音的问题,提出一种改进算法,充分考虑分簇采样点所占比重对α值估算准确度的影响,在贝叶斯框架中嵌入一个含比重因子的平滑约束项表达式。仿真实验结果表明,该改进算法能够有效消除脉冲噪音,使抠像结果更为光滑。

关键词: Bayes抠图, 数字抠图, 自然图像抠图

Abstract: Aiming at the problem that traditional Bayes matting algorithm usually generates impulse noises in the resulted opacity channel, an improved algorithm is presented, which considerates the effect on α by clustering sampling points weight and introduces a term of smoothness constraint into the Bayes matting framework. Simulation experimental results show that the impulse noise is eliminated and matting quality is improved.

Key words: Bayes matting, digital matting, natural image matting

中图分类号: