基于WNMF和聚焦点定位分析的多聚焦图像融合方法

在非负矩阵分解(NMF)图像融合方法的基础上,提出一种基于加权非负矩阵分解(WNMF)和聚焦点定位分析的多聚焦图像融合方法。该方法利用光学系统成像原理及点扩散函数在光学成像过程中的作用,判定多聚焦图像中的聚焦点的近似位置,并以此为基础构建一个权值矩阵,然后将加权非负矩阵算法应用于图像融合中,最后得到一幅各部分都聚焦清晰的图像。实验结果表明,通过本文提出的方法得到的融合图像效果优于普通的非负矩阵分解方法、小波变换法方法及拉普拉斯塔式方法。

Multi-Focus Image Fusion Based on WNMF and Focal Point Position Analysis

In this paper, a multi-focus image fusion method based on the weighted non-negative matrix factorization is presented to improve image fusion method based on non-negative matrix factorization. The method analyzed the relationship between the point spread function in optical system and imaging processing, and built a weight matrix U by determining the focal point on multi-focus images. Finally, the weighted non-negative matrix algorithm was applied for image fusion and got a new image that all parts of the image were clear. Experimental results showed that the image fusion method is better than non-negative matrix factorization method, wavelet transform fusion method and Laplacian pyramid fusion method.