摘要
针对显微图像灰度对比度低,内容复杂度高的特点,提出一种基于小波的显微图像融合算法.将图像变换到小波域,并在此基础上进行四叉树分解融合,以消除块效应,不同于现有的小波几何分块算法,可依据图像实际情况调整,符合显微图像结构多样化、内容多变化的实际情况.实验结果表明,本文算法的融合效果理想,与同类算法相比具有更强的稳定性和适用性.
Based on the fact that a micrograph is characterized with low contrast and high content complexity, the author proposes a wavelet transform based algorithm for micrograph fusion. The images are first transformed by wavelet, then fused based on quadtree decomposition to eliminate the ‘block effect＇. Different from geometric-block method, block size in quadtree is changeable, adjustable as needed. The proposed method can deal with the problems normally identified with micrographs such as structural diversity and content complexity. The fusion is proved to be more stable and effective than those generic algorithms.
出处
《宁波大学学报（理工版）》
CAS
2006年第3期286-289,共4页
Journal of Ningbo University：Natural Science and Engineering Edition
基金
宁波大学学科基金(XK200438)
关键词
小波变换
四叉树分解
多聚焦融合
wavelet transform
quadtree decomposition
multi-focus image fusion
