摘要
图像的不同特征从不同的角度反映出图像像素间的差异。本文在Contourlet变换下,利用D-S证据理论的聚集性,将图像不同邻域特征的作为基本概率赋值的依据合成融合系数,提出基于证据理论的多聚焦图像融合算法。仿真实验表明,该算法能有效地综合图像的多个邻域特征,具有较好的融合效果。
The different characteristics of the image reflect the difference between the image pixels from different angles.We synthesis fusion coefficient using characteristics of different neighborhood of images as the hasis of basic probability assignment making use of clusters of D-S evidence theory under Contourlet transformation in my thesis and propose the algorithm of multi-focus image fusion based on the evidence theory.It shows that the algorithm can effectively integrate multiple neighborhood features of the image which is provided with a good fusion effect according to the simulation experiment.
出处
《科技视界》
2015年第2期189-190,共2页
Science & Technology Vision
