摘要
在简要分析多聚焦图像融合问题后,详细推导并提出一种新的基于全变分的图像融合算法。相比以前基于多分辨率框架进行融合的金字塔分解、小波变换等方法,全变分图像融合方法将多聚焦图像问题转化为最优化问题,直接在空间域进行计算操作,简单直观,便于进行改进,并且取得很好的融合效果。
Based on brief analysis of the multifocus image fusion problem,in this paper a detail derivation is made which lead to a proposal of new algorithm of total variation-based image fusion （ TV method）. Comparing with the methods of pyramid decomposition fusion based on multiple resolution framework and wavelet transform used before, TV method turns multifocus image problem into optimization problem and computes directly in spatial domain, so it is visual and simple, easy to be modified, and have gotten quite good fusion results.
出处
《计算机应用与软件》
CSCD
2009年第5期225-227,共3页
Computer Applications and Software
关键词
多聚焦图像融合
全变分
小波变换
Multifocus image fusion Total variation Wavelet transform
