摘要
小波变换应用于多聚焦图像融合时分解层数的选取是一个关键问题。根据区域标准偏差最小准则对小波变换后的高、低频图像分别进行融合。利用峰值信噪比、偏差度、熵等标准对融合结果进行评价。对比实验结果表明,多聚焦图像融合的最佳小波分解层数为5层。
Selection of decomposition level is a key problem when wavelet transform is applied to multi-focus image fusion. Fusion for high-frequency and low-frequency images is carried out, respectively according to root mean standard error rule based on region. The fusion results are evaluated by using the parameters such as peak signal-noise-ratio, departure and entropy. The comparison experiments show that the optimal wavelet decomposition level for multi-focus image fusion of 5 levels.
出处
《光电工程》
CAS
CSCD
北大核心
2004年第3期64-67,共4页
Opto-Electronic Engineering
基金
暨南大学自然科学基金资助项目
关键词
图像融合
分解层数
小波变换
Image fusion
Decomposition level
Wavelet transform
