基于改进的空域相关的多聚焦图像融合

提出了一种简单有效的像素级多聚焦图像融合方法。针对正交小波变换缺乏平移不变性而产生视觉失真的缺陷,采用Atrous算法将原图像分解在不同频率域上。Atrous算法先将滤波器h0(n),h1(n)各点间插入适当的零值后再与低频信号做卷积,故又称为"多孔算法"。将具有抑制噪声性能的空阈相关法作为高频子图像的融合规则,选取相关性强边缘特征显著的点作为最终融合子图像的像素点。实验表明,由此融合的图像能完好的保留边缘纹理信息。融合后的图像在客观评价和主管视觉效果上均有提高。

Multi-focus image fusion based on improved spatial correlation

A novel simply and effectively multi-focus images fusion method is proposed on pixel level.Because orthogonal wavelet transform is short of translation invariant and have vision distortion.Atrous algorithm is used to decompose the original images in different frequency region.Atrous algorithm is to interpose proper zeros between arbitrary points of filter before the filters and convolve with low frequency image.So this algorithm is also called "redundancy algorithm".While choosing high frequency coefficients,the spatial correlation filter method is used,which could inhibit noise and select the pixel which has strong correlation and prominence character as the fusion image wavelet coefficient.Experiments results show the proposed approach could preserve the image edge and stripe detail successfully.This algorithm can improve the quality of the fused image in both objective effect and subjective effect.