基于小波变换和引导滤波的多聚焦图像融合  

Multi-Focus Image Fusion Based on Wavelet Transform and Guided Filtering

作　　者：朱世松[1] 瞿佩云 ZHU Shi-song;QU Pei-yun(College of Computer Science and Technology,Henan Polytechnic University,Jiaozuo 454003,China)

机构地区：[1]河南理工大学计算机科学与技术学院,河南焦作454003

出　　处：《测控技术》2020年第9期103-107,共5页Measurement & Control Technology

基　　金：河南省国际科技合作项目(084300510065);河南省教育厅科学技术研究重点项目(13A520340);河南理工大学博士基金项目(B2010-95);河南省高等学校矿山信息化重点学科开放实验室开放基金资助项目(KZ2012-02)。

摘　　要：为了进一步提高多聚焦图像融合效果,提出了一种基于小变换和引导滤波的多聚焦图像融合方法。对源图像进行二维小波分解,得到低频子带系数和高频子带系数。对低频子带系数采用引导滤波加权融合;对高频子带系数引入最大对称环绕显著性检测算法(Maximum Symmetric Surround Saliency Detection Algorithm,MSSS),归一化显著图得到权重图,进而进行加权融合。把得到的高频和低频子带系数进行小波重构,得到最终的融合图像。实验结果表明,与其他算法相比,所提算法具有更好的清晰度,得到较好的融合结果。In order to further improve the effect of multi-focus image fusion,a multi-focus image fusion method based on wavelet transform and guided filtering is proposed.The source image was decomposed by two-dimensional wavelet to obtain the coefficients of the low frequency subband and the high frequency subband.Guided filtering weighted was used to fuse low frequency subband coefficients,and MSSS(Maximum Symmetric Surrounding Saliency Detection Algorithm)was introduced for high frequency subband coefficients.The saliency map was normalized to obtain weight map,and then the weighted fusion was performed.The high-frequency and low-frequency subband coefficients were reconstructed by wavelet to obtain the final fused image.Experimental results show that the proposed algorithm has better definition and better fusion results than other algorithms.

关 键 词：多聚焦图像融合 小波变换 引导滤波 显著性检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]