Abstract: The fusion rule of wavelet decomposition’s high-frequency coefficients based-on local variance’s select and weighted average is proposed.This both preserves the source image’s details information and ensures the fusion results’ consistency.The algorithm is used to fuse multifocus image,infrared and visible light image,respectively.The results show that this algorithm can preserve richer detail information from primitive images and avoid target distortion,have better vision effect.

摘要： 提出了图像小波分解的高频部分信息采用基于局部方差的选择和加权平均相结合的系数融合规则,在充分保留源图像细节信息的前提下,保证了融合结果的一致性。针对多聚焦图像融合、红外和可见光图像融合,试验结果证明本方法所得的融合图像能反映更丰富的细节信息,更小的目标失真,具有更好的视觉效果。