一种基于物理机理的高光谱遥感图像真彩色校正模型

A True Color Correction Model for Hyperspectral Remote Sensing Image based on Physical Mechanism

针对高光谱遥感影像由于各波段光谱范围窄,难以获得符合人们视觉效果的真彩色合成影像问题,提出一种基于物理机理的高光谱遥感图像真彩色校正模型。该模型充分利用高光谱影像在红、绿、蓝反射区的所有谱段信息,通过插补波段并进行波段加权积分重建真彩色合成图像,进而结合实测地物反射率光谱,利用辐射传输模拟的方式,构建具备普适性的真彩色校正模型。利用航空高光谱遥感影像进行色彩校正实验的结果表明,所构建的真彩色校正模型能够很好地应用于高光谱遥感影像真彩色校正。

It is difficult to obtain true color composite image that is consistent with human visual effects, since each spectral range of hyperspectral remote sensing image is narrow. In order to solve this problem,a true color correction model was proposed based on the physical mechanism of color generation. The model fully employed all spectral information in red, green, blue reflection zones of hyperspectral images and reconstructed true color composite images by band interpolation and band weighted integral. Then, in-situ measured surface reflectance spectral data was added into MODTRAN atmospheric radiation model to ob- J tain theoretical color of targets with radiant transfer simulation. Finally,in order to establish correction ma- trix between reconstructed color and theoretical color of targets,and construct a universal true color correction model we solved some overdetermined equations. The experimental results of airborne hyperspectral remote sensing images show clearly that the true color correction model can be applied to true color cd;rection of hyperspectral remote sensing images, and can also be extended to multi-spectral remote sensing images.