基于改进的áTrous小波与IHS变换法遥感影像融合

Remote-sensing image fusion based on modified áTrous wavelet and IHS transform

对áTrous小波分解后的图像高频信息,采用局部方差的归一化加权融合方法融合图像,考虑了局部方差归一化过程中分母为零的情况。最后,融合结果既保留了多光像的纹理信息又保留了全色影像的高频信息,光谱信息更加丰富,目标细节更加精细,轮廓清晰可辩。

For the high-frequency information decomposed by áTrous wavelet, the image fusion was completed using normalization weighted of the partial variance. The case that the denominator might be zero was considered when normalizating the partial variance. The fused image both retain the texture information of multi-spectral images and the high-frequency information of the entire malti-color images. The spectrum information is much more abundant, the object detail is more precise, and the outline is more clear.