Image fusion can be used to improve the image utilization, spatial resolution and spectral resolution, which has been widely applied on medicine, remote sensing, computer vision, weather forecast and military target recognition. The goal of image fusion is to reduce the uncertainty and redundancy of the output and increase the reliability of the image on the basis of the maximum combination of relevant information. In this paper, a multi-focus image fusion algorithm based on WNMF and Focal point position analysis is proposed to improve the image fusion method based on nonnegative matrix factorization. In the imaging process, the Gaussian function is used to approximate the point spread function in the optical system. Then calculate the difference between the original image and the approximate point spread function and get the weighted matrix U. Finally, we apply the weighted nonnegative matrix algorithm to image fusion, and the new fusion image with clear parts is obtained. Experimental results show that the multi-focus image fusion algorithm based on WNMF and Focal point position analysis (MFWF) is better.

中文翻译：

---

多焦点图像的优化融合：融合 WNMF 和焦点分析

图像融合可用于提高图像利用率、空间分辨率和光谱分辨率，已广泛应用于医学、遥感、计算机视觉、天气预报和军事目标识别等领域。图像融合的目标是在最大限度地组合相关信息的基础上，降低输出的不确定性和冗余度，提高图像的可靠性。本文提出了一种基于WNMF和焦点位置分析的多焦点图像融合算法，以改进基于非负矩阵分解的图像融合方法。在成像过程中，使用高斯函数来逼近光学系统中的点扩散函数。然后计算原始图像与近似点扩散函数的差值，得到加权矩阵ü. 最后，我们将加权非负矩阵算法应用于图像融合，得到了部分清晰的新融合图像。实验结果表明，基于WNMF和焦点位置分析（MFWF）的多焦点图像融合算法效果更好。