基于快速清晰度估计的多聚焦图像融合  

Multi-focus image fusion based on fast sharpness estimation

作　　者：费春[1] 张萍[2] 李建平[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,成都610054 [2]电子科技大学光电信息学院,成都610054

出　　处：《计算机应用研究》2016年第3期896-898,921,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(61308102);高等学校博士学科点专项科研基金资助项目(20130185120038);四川省科技计划基金资助项目(2014GZ0005;2014JY0228);中国博士后科学基金资助项目(2013M531946);中央高校基本科研业务费资助项目(ZYGX2013J059)

摘　　要：为了准确地估计源图像的清晰区域,提高多聚焦图像融合的效率,提出了一种新的基于清晰度估计的图像融合方法。首先利用基于离散小波的清晰度估计方法获取源图像的聚焦区域;然后使用均值滤波和空洞填充进一步优化该聚焦区域;最后结合清晰度估计和相似性特性,将不同聚焦区域合并生成融合图像。该方法获得的融合图像在客观评价和主观质量上都优于以往基于清晰度的图像融合方法。This paper proposed a novel method multi-focus image fusion based on sharpness estimation in order to accurately estimate the sharpness area of source images and improved the efficiency of multi-focus image fusion. First,the proposed method obtained the focus areas from source images by using a sharpness estimation method based on discrete wavelet transform.Then,it used mean filter and hole-filling to optimize the focus areas. Finally,the proposed method selected different focus areas to generate fusion image combining sharpness estimation and similarity characteristics. The experimental results show that the proposed method outperforms to conventional fusion methods based on sharpness estimation in terms of objective evaluation and subject quality.

关 键 词：多聚焦 清晰度估计 图像融合 相似性 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]