基于预处理的金字塔图像融合算法

Algorithm of Pyramid Image Fusion Based on Pre-processing

针对经典金字塔融合算法在重构过程中由噪声叠加引起的黑斑问题,提出了一种新的基于分层预处理的红外与可见光图像融合算法.该算法对金字塔分解的各层图像进行预处理,优化各分解层图像质量后再将各自对应层融合.采用该算法融合的图像质量能降低直接对源图像进行融合时融合规则的复杂度及融合后改善图像质量处理的难度,仿真结果验证了该算法的有效性.与经典金字塔算法、同类金字塔改进算法相比,该算法包含的信息量更多.

Considering that the pyramid infusion algorithm has a problem of black spots caused by noise superposition in re-structuring procedure, this paper proposes a novel algorithm for infrared and visible image fusion based on layered pre-processing, and makes a fusion for those decomposition layers after optimizing each layer’s image quality. The image quality fused by using this algorithm can reduce the complexity of fusion regularity when fusing directly the source images and lower the processing difficulty of improving the image quality after fusion. Simulation results prove this proposed algorithm to be of effectiveness. Compared with the classic pyramid algorithm and other similar improved pyramid algorithms, this novel algorithm can contain much more information.