计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (9): 178-180.

• 图形、图像处理 • 上一篇    下一篇

基于微粒群算法的纹理合成图像修补方法

朱 为,李国辉,来 旭   

  1. 国防科技大学 信息系统与管理学院,长沙 410073
  • 收稿日期:2007-10-19 修回日期:2008-01-17 出版日期:2008-03-21 发布日期:2008-03-21
  • 通讯作者: 朱 为

Texture synthesis image completion based on PSO

ZHU Wei,LI Guo-hui,LAI Xu   

  1. Department of Information System & Management,National University of Defense Technology,Changsha 410073,China
  • Received:2007-10-19 Revised:2008-01-17 Online:2008-03-21 Published:2008-03-21
  • Contact: ZHU Wei

摘要: 提出了一种基于微粒群算法(PSO)的纹理合成图像修补方法。针对基于块的纹理合成算法中常用的匹配块全搜索效率低下的问题,对候选区域采样确定初始微粒,利用PSO算法搜索得到最优匹配块。实验结果表明了方法的有效性。

关键词: 图像修补, 纹理合成, 微粒群算法

Abstract: In this paper,a texture synthesis image completion based on Particle Swarm Optimization (PSO) is proposed.According to the inefficient matched patches full search often used in exemplar-based texture synthesis,initial sampling particles are selected in candidate area,and PSO algorithm is used to search the optimal matched patch.Experiment results demonstrate the effectiveness of the proposed approach.

Key words: image completion, texture synthesis, PSO