能量属性与引导滤波结合的遥感影像融合  

Remote sensing image fusion combining energy attribute and guided filter

作　　者：宋加文 朱大明[1] 付志涛 陈思静 SONG Jiawen;ZHU Daming;FU Zhitao;CHEN Sijing(Faculty of Land and Resources Engineering,Kunming University of Science and Technology,Kunming 650093,China)

机构地区：[1]昆明理工大学国土资源工程学院,云南昆明650093

出　　处：《测绘学报》2023年第12期2154-2163,共10页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金(41961053)。

摘　　要：针对多光谱与全色影像融合容易出现空间和光谱信息失真的问题,本文提出一种基于能量属性与引导滤波的全色锐化方法。首先,对多光谱影像进行强度-色调-饱和度变换提取强度分量,强度分量与全色影像通过均值滤波和差分算子得到高频分量和低频分量,其中高频信息通过引导滤波增强;然后,使用像素最大值规则获得决策图;最后,高频图像通过像素加权平均规则与决策图融合。低频分量通过EA策略融合,结合新的低频分量与高频分量替代原强度分量并逆IHS变换得到融合影像。本文在SPOT-6、WorldView-2和Pléiades NEO遥感影像上进行大量试验,并与4种先进方法进行定量定性对比。本文方法的光谱角度映射、相对无量纲全局误差、相对平均光谱误差、均方根误差、通用图像质量指数和峰值信噪比指标值较次优值分别平均提高了77.13%、10.78%、9.57%、12.20%、1.35%及0.39%。试验结果表明,本文方法可以在保持多光谱影像光谱信息的同时充分融入全色影像的空间信息,并在视觉感知、定量指标和时间效率上都取得最优的融合效果。A pan-sharpening method based on energy attribute(EA)and guided filter is presented to solve the problem of spatial and spectral information distortion in multispectral and panchromatic image fusion.First,the intensity-hue-saturation(IHS)transformation is applied to the multispectral image to extract the intensity component.The high and low-frequency components are obtained from the intensity component and the panchromatic image using the mean filter and difference operator.The guided filter enhances the high-frequency information.The decision map is obtained using the rule of maximum pixel value,and the high-frequency image is fused with the decision map by the rule of the weighted average pixel.The EA strategy fuses low-frequency components.The fusion image is obtained by combining the new low-frequency component with the high-frequency component instead of the original intensity component and inverting the IHS transformation.In this paper,many experiments are carried out on SPOT-6,WorldView-2,and Pléiades NEO remote sensing images,and the results are compared quantitatively and qualitatively with four advanced methods.The spectral angle mapping,relative dimensionless global error in synthesis,relative average spectral error,root mean square error,universal image quality index,and peak signal-to-noise ratio values of the proposed methods were improved by 77.13%,10.78%,9.57%,12.20%,1.35%,and 0.39%compared with the sub-optimal values,respectively.The experimental results show that this method can fully incorporate the spatial information of panchromatic images while maintaining the spectral information of multispectral images and achieve optimal fusion results in visual perception,quantitative indicators,and time efficiency.

关 键 词：图像融合 能量属性 引导滤波 均值滤波 全色影像 多光谱影像 

分 类 号：P237[天文地球—摄影测量与遥感]