为增加融合后图像信息量, 增强图像清晰度, 提出一种基于Red-Black小波图像融合的改进算法。该算法在对源图像进行Red-Black小波变换前, 首先进行PCA变换, 然后对得到的第一分量作多尺度的Red-Black小波分解, 对于高频子带系数和低频子带系数分别采用基于极大值的选择和基于区域方差的融合规则。最后对融合后的各级子图像进行Red-Black重构, 并对重构的结果进行PCA逆变换得到融合的图像。实验结果表明, 提出的算法进行多光谱图像融合可以很好地保留光谱信息并且有较高的空间分辨率, 提高了熵值, 增加了融合后图像的信息量, 扩大了信息丰富程度, 也有效增强了图像的清晰度, 具有较好的视觉效果, 在遥感图像处理领域有较广泛的应用前景。

To increase the amount of image information after fusion and improve image clarity, an improved image fusion algorithm based on the Red-Black wavelet is put forward. The PCA transform is performed to the source image before Red-Black wavelet transform. Then the multi-scale Red-Black wavelet decomposition is carried out to the obtained first component. For the high-frequency and low-frequency sub-band coefficient, the selection based on the maximum value and the fusion rule based on the regional variance are adopted respectively. The Red-Black reconstruction is carried out to the sub-images of each level after the fusion, and PCA reverse transform is carried out to the reconstructed results to obtain the fused images. The simulation results show that, the multi-spectral image fusion performed by the proposed algorithm can effectively preserve spectral information, has a high spatial resolution, improve the entropy, increase the amount of information of the fused images, and improve image clarity effectively with better visual effects. It has a wide application prospect in the field of remote-sensing image processing.