作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

›› 2003, Vol. 29 ›› Issue (17): 13-.

• 论文 • 上一篇    下一篇

一种有效的运动模糊图像恢复算法

王晓红,陈秀万,谭仲军,赵荣椿   

  1. 北京大学遥感所,北京大学遥感所,天创科技发展有限公司,西北工业大学计算机科学与工程系 北京 100871
  • 发布日期:2003-09-15

An Efficient Algorithm for Restoration of Motion-blurred Images

WANG Xiaohong1, CHEN Xiuwan1, TAN Zhongjun2, ZHAO Rongchun3   

  1. (1. Institute of Remote Sensing, Peking University, Beijing 100871; 2. Tianchuang High-tech Development Co. ltd., Beijing 100031; 3. Department of Computer Science and Engineering, Northwestern Polytechnical University, Xi'an 710072)
  • Published:2003-09-15

摘要: 探讨了任意方向的匀速直线运动模糊图像的恢复问题。证明了对任意方向的直线运动模糊图像只能直接设置二维点扩展函数进行二维恢复,不可先进行x方向的一维恢复再进行y方向的一维恢复。给出了估计PSF的方向和模糊范围D的频谱方法。并用实验验证了PSF谱估计方法的有效性,以及带窗的维纳滤波方法相比于标准维纳滤波方法的优越性。

关键词: 点扩展函数, 运动模糊, 维纳滤波

Abstract: This paper discusses the restoration problem of motion-blurred images caused by constant-speed straight-line movement of arbitrary direction. And it proves that blurred images must be restored in 2D form by directly setting 2D PSF, instead of first restored in X direction and then in Y direction. Then it proposes the spectral method for estimating the direction of PSF and blur extent D. The validity of the PSF estimating method and superiority of windowed Wiener filter over standard Wiener filter are clearly demonstrated by experiments.

中图分类号: