Abstract: The application of wavelet transform in image fusion is carried out widely, but most of the methods often simply replace the low frequency components or weighed mean directly after the pyramid decomposition of wavelet. Thus it can not make good use of the low frequency information of the multisource image. In response to this status quo, this paper puts forward point sharpness method in the integration process of the low frequency components, and combines with the method of local energry in the high frequency part, makes an image fusion experiment of multisource remote sensing image by wavelet transform. Results show that fusion image improves in maintaining details and spectral information and also has good evaluation index.

Key words: wavelet transformation, image fusion, point sharpness method, local energry, evaluation index

摘要： 小波变换在图像融合中已有广泛的应用，但在进行小波塔形分解后通常是直接将低频分量简单替换或者均值加权，不能很好地利用多源图像的低频信息。针对这一现状，对基于小波变换的图像融合方法作了改进，对低频分量融合提出点锐度法进行处理，并对高频部分融合采用的区域能量法进行改进。仿真实验结果表明，融合的图像在保持细节和光谱信息方面都有一定的提高,评价指标较好。

关键词: 小波变换, 图像融合, 点锐度法, 区域能量, 评价指标

CLC Number: 

• TP391.41