基于散度的不同焦点图像融合方法

把散度的概念引入到图像分析中,考虑到图像在不同方向上的性质不同,提出了一种基于散度的相关性拉普拉斯变换不同焦点图像融合算法.首先对源图像进行相关性拉普拉斯分解,获得图像的低频和高频分量;然后对低频分量采用平均能量法进行融合,对高频分量利用图像梯度场的散度作为显著性特征进行融合;最后对融合后的图像分量进行拉普拉斯反变换重构出融合图像.实验结果表明该方法的保真度更高,边缘信息保留性能更好.

Divergence-based multifocuses image fusion

The concept of divergence was used in image processing. In consideration of the anisotropic property of images, a novel divergence-based moutifocuses image fusion is presented by using pertinence Laplacian analysis. The original images were decomposed into the Laplacian pyramid to obtain the low/high-frequency component image sequences. The components with low frequency were fused by averaging energy method, the high-frequency component image sequences with the largest divergence were selected at each pixel location and the fused image was recovered from the decomposed component image sequences. Experimental results show that the proposed method outperforms the traditional approach in the objective fidelity criterion and the objective fusing performance.