基于改进拉普拉斯金字塔的图像融合方法

Improved Laplacian pyramid image fusion method based on

提出了一种新的基于改进拉普拉斯金字塔变换的图像融合算法。传统拉普拉斯金字塔变换的重构算法在图像在融合过程中会引入噪声,而改进的重构算法能有效抑制融合噪声。首先将图像进行拉普拉斯金字塔分解,得到各层拉普拉斯金字塔图像,针对不同的分解层图像,采用不同的方法进行融合,最后采用改进的拉普拉斯金字塔重构算法进行重构,得到最终的融合图像。通过仿真实验验证了算法的有效性。

A new image fusion algorithm based on improved Laplacian transform pyramid is proposed in this paper. The reconstruction algorithm of traditional Laplacian pyramid transform may introduce noise in the image fusion process, but the improved reconstruction algorithm can effectively suppress fusion noise. First, decompose the image with the Laplacian pyramid to obtain layers of Laplacian pyramid image. For different layers of decomposition image, different fusion method should be used. At last, use the improved Laplacian pyramid reconstruction algorithm to reconstruct and obtain the final fused image. Simulation results demon-strate the effectiveness of the algorithm.