摘要
通过分析图像经多小波分解后小波系数的多个统计特征,构造出多小波系数区域相似度,结合多聚焦图像的成像机理,提出了一中基于多小波变换和区域相似度的融合算法,实验结果表明,该算法可有效融合多聚焦图像,从而获得场景内所有物体都清晰的图像。
Defines the regional similarity after analyzing character of muhiwavelet decomposed coefficients, proposes a new fusion based on muhiwavelet transform and regional similarity with principles of muhifocus imaging .The experiments show that the fusion method can efficiently fuse muhifocus images and give a satisfactory result.
出处
《现代计算机》
2007年第10期47-50,共4页
Modern Computer
关键词
多小波变换
区域相似度
图像融合
Multiwavelet Transform
Regional Similarity
Image Fusion
