双密度双树复小波域多聚焦图像融合

摘要

将双密度双树复小波变换引入多分辨率图像融合中，利用双密度双树复小波变换的多尺度、多方向性和平移不变性特征分解多聚焦图像，对分解后高、低频图像系数采用不同融合策略进行融合，应用双密度双树复小波逆变换重构图像。采用多组多焦距源图像进行融合实验，并对融合结果进行了客观评价。实验结果表明双密度双树复小波域融合方法对多聚焦图像具有更好的融合效果，图像的细节描述更加精确。%Double-density dual-tree complex wavelet transform is introduced to image fusion based on multi-resolution, multi-focus images are decomposed by double-density dual-tree complex wavelet transform with multi-scale, multi-direction and shift-invariance of nonsubsampled transform and according to variable characters of low and high frequency coefficients , different combination decisions are adopted to fuse images and composite coefficients are reconstructed by double-density dual-tree complex wavelet inverse transform. Some fusion experiments are done by several sets of multi-focus images and objective performance assessments are fulfilled to evaluate fusion results. The experimental results indicate that more favourable fusion performance is generated and image detail is described more fine.