一种基于形态学梯度的多聚焦图像融合方法

A multi-focus image fusion method based on morphological gradient

多聚焦图像融合作为信息融合的一个重要分支,广泛应用于计算机视觉、医学诊断、数字成像等领域。文中针对现有融合算法存在融合图像易出现“块效应”等问题,提出了一种基于形态学梯度的多聚焦图像融合方法。该算法首先对源图像进行形态学梯度测量得到对应的得分矩阵,然后对得分矩阵进行四叉树分解得到初始分割图。进而利用形态学滤波和小区域滤波,除去边缘信息和小洞得到期望的决策图。实验测试表明,文中提出算法相对于现有的算法在结构相似度、峰值信噪比等定量指标上皆有较好的表现。

As an important branch of information fusion, multi-focus image fusion is widely used in computer vision, medical diagnosis, digital imaging, and other fields. In this paper, a multi-focus image fusion method based on the morphological gradient is proposed to solve the problem of blocking effect in the existing fusion algorithm. The algorithm first measures the morphological gradient of the source image to receive the corresponding score matrix, and then decomposes the score matrix via the quadtree approach to obtain the initial segmentation image. Nextly, morphological filtering and small zone filtering are used to delete edge information and small holes to obtain the desired decision graph. The experiment results show that the proposed algorithm performs better in terms of structural similarity, signal-to-noise ratio, and other quantitative indicators than the existing algorithms.