【摘要】 针对全色和多光谱遥感图像的融合,本文提出了一种基于快速IHS融合技术的小波融合新方法。为提高融合结果图像的光谱保真度,采用了基于改进主成分分析的低频小波系数智能自适应算法融合策略,其主观视觉分析和客观评价结果显示,与传统加权平均、以及平均与选择相结合的低频融合规则相比,本算法在保证能够大幅度提高融合图像清晰度的同时,较大程度地降低了它的光谱畸变度;另外,由于将快速IHS融合技术引入了基于小波分解的融合方法中,大大降低了程序的运行时间,因此本算法在实时性方面也得到了显著提高。将其应用到更大尺寸的巨幅遥感图像的融合中,它的实时性效果更显著。更多还原

【Abstract】 In view of the fusion of Panchromatic and Multispectral remote sensing images a new method with spectrum information of high fidelity and fast speed is proposed,in which the wavelet fusion method is combined with the fast IHS transformation.In order to enhance the fidelity of spectrum information of fused images,a kind of intelligent and self-adaptive fusion strategy for low frequency wavelet coefficients is studied and built based on the reformative principal component analysis.Subjective qualitative visual effect and objective quantitative evaluation that indicate,the new algorithm proposed in this paper can reduce the spectrum distortion observably and enhance the definition greatly as well,compared with other methods such as the conventional weighted average and the combination of weighted average and selection.In addition,the new method is also characterized by well real-time performance due to the introduction of fast IHS transformation into the wavelet-based fusion algorithm,and its realtime processing capability is especially notable in fusing of large remote sensing images.更多还原