雷达与光学图像小波融合方法研究

A New Fusion Algorithm Study for SAR and Optical Images Based on Wavelet Transform

本文基于区域差的绝对值,提出了一种新的双正交小波融合算法,并应用于TM与JERS-1SAR的融合研究中。该方法与传统的小波融合相比,更能充分利用参与融合影像的信息,具有更强的数字图像细节表征能力;与基于区域方差的小波融合相比,具有更小的计算量;与主成分融合、乘法融合及Brovey融合等方法相比,具有更好的光谱保持能力,且该小波融合结果更清晰。实验表明,该算法是一种实用有效的融合法。

A new biorthogonal wavelet fusion algorithm was developed based on the absolute values of the difference of local area of the images. The algorithm was applied to the images of TM and JERS-1 SAR. The algorithm has stronger ability to describe the digital images＇ details and to compound the images＇ information than traditional wavelet fusion algorithms,and it needs less computation than those based on variance. It is a better way to preserve the spectral feature and detailed information than those algorithms of the PCA fusion,multiplicative fusion and Brovey fusion. The results of experiment show that the algorithm is highly flexible and practical.