基于边缘分割的多光谱图像融合方法

Multispectral Image Fusion Method Based on Edge Segmentation

对于高分辨率的全色图像与低分辨率的多光谱图像,通常采用传统的IHS方法进行融合。这种融合方法能很好地提高低分辨率图像的清晰度,但却容易扭曲原始的光谱特性,产生光谱退化现象。针对这一问题,本文采用高分辨率与多光谱图像的边缘分割信息对I分量进行调制,使新生成的I分量保持原始的光谱特性。通过融合实验结果表明,这种方法比传统的IHS方法有效地改善了融合图像光谱失真的现象。

For high-resolution panchromatic image and low-resolution multispectral image,usually adopted the traditional fusion method of IHS.This fusion method can well improve the clarity of low-resolution images,but it easily tend to distort the spectral characteristics of the original,engender spectral degeneration.As a result of this problem,the article adopts edge segmentation of multispectral and high resolution to modulate the component of I,keep the new compoonet of I to the original properties of spectral.The experiments of fusion show that this method improved spectral distortion better than the traditional fusion method of IHS.