【摘要】 建立了同一场景不同空间分辨率图像间的局部空间线性恢复模型(LSLRM),并将其应用于多光谱(MS)和全色(Pan)图像的融合,提出了基于LSLRM的MS和Pan图像融合算法。该算法以HIS变换和冗余小波变换为分析工具对参与融合的源图像的光谱和空间信息以及空间高、低频信息进行分离,并在相应的图像分量之间建立LSLRM,从而完成融合步骤。实验结果表明,该融合算法能够以较为合理的时间开销有效地保留MS和Pan图像中的光谱和空间特征,引起较小的光谱畸变,还能够根据实际需要设定软/硬阈值,以获得更合适的融合图像。更多还原

【Abstract】 In the full paper,we explain in some detail a new algorithm for fusing images,which are of the same scene but whose spatial resolutions are different;in this abstract,we just add some pertinent remarks to listing the two topics of explanation.The first topic is: redundant wavelet transform(RWT).The second topic is: multispectral and panchromatic image fusion algorithm based on LSLRM.In the second topic,we mention the separation of source images in two respects:(1) the separation of spectral information from spatial information using the HIS transform;(2) the separation of spatial high-frequency information from spatial low-frequency information using RWT.Then in the second topic,we establish LSLRMs in order to complete the fusion procedure.Finally we compare the fusion results obtained with our new algorithm with those obtained with the methods in Ref.1 by Hill et al and Ref.3 by Zhu et al.The comparison,shown in Fig.2 in the full paper,shows preliminarily that our new algorithm is better,is capable of preserving spectral features of multispectral image and spatial features of panchromatic image effectively with only reasonable calculation time required, and causes less spectral distortion.Table 2 in the full paper also shows preliminarily that,using our new algorithm,we can set soft/hard thresholds so as to acquire more appropriately fused images.更多还原