计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (31): 199-201.DOI: 10.3778/j.issn.1002-8331.2008.31.057

• 图形、图像、模式识别 • 上一篇    下一篇

小波变换和清晰块搜索结合的图像融合算法

曾志峰,马华山   

  1. 西南电子电信技术研究所,成都 610041
  • 收稿日期:2007-12-06 修回日期:2008-03-19 出版日期:2008-11-01 发布日期:2008-11-01
  • 通讯作者: 曾志峰

Method for multi-focus image fusion based on union of wavelet transform and block search

ZENG Zhi-feng,MA Hua-shan   

  1. Southwest Electronics and Telecommunication Technology Research Institute,Chengdu 610041,China
  • Received:2007-12-06 Revised:2008-03-19 Online:2008-11-01 Published:2008-11-01
  • Contact: ZENG Zhi-feng

摘要: 在基于小波变换的多聚焦图像融合中,低频系数选取决定融合图像的轮廓,同时影响融合图像的边缘细节,因此小波变换法是以损失图像的局部清晰来换取整幅图像的平均清晰;而基于空域块选择的融合方法能很好地保留利用聚焦图像的清晰部分,但是图像模糊区域效果不佳。因此提出一种新的融合算法,先对图像进行小波变换融合,之后对原图像进行清晰块搜索,进行基于清晰块选择的图像融合。实验结果表明,该算法融合结果优于单一的小波变换算法。

关键词: 多聚焦图像融合, 块搜索, 小波变换

Abstract: For the multi-focus image fusion using 2D Discrete Wavelet Transform(DWT),the selection of low frequency coefficients decides the contour and affects the edge details of the fused image,therefore,DWT get the average definition of the whole image,losing the partial definition;while the method based on block pasting can preserve the local clear part.Because of this,this paper proposes a new method used for image fusion,first using DWT method,then searching the clear blocks,using the block pasting method.Experiments show that the method proposed is better than the wavelet transform method.

Key words: multi-focus image fusion, block search, wavelet transform