一种反射光谱耦合的遥感自然色影像模拟方法

A Reflectance Spectrum Coupled Remote Sensing Natural-color Image Simulation Approach

针对多光谱遥感影像像元或光谱库地物样本进行多光谱谱间关系耦合模拟时通常会存在样本与目标影像时相不一致、样本受大气干扰等引起蓝波段及自然色模拟模型不稳定、精度难以提高的问题,提出一种利用MODIS地表反射率产品(MOD09)作为参考影像选取样本进行光谱耦合的自然色模拟方法。首先对MOD09选取的样本利用非线性回归模型——BP神经网络来拟合MOD09蓝波段与红、绿、近红波段间光谱关系;然后,基于这种拟合关系,通过已有的绿、红、近红波段来模拟其他多谱遥感影像蓝光波段,并合成其自然色影像。实验结果表明,BP神经网络在很好地模拟出多谱遥感影像的蓝波段的同时,能够消除大气效应对影像的影响;同时,基于模拟出的蓝波段可以合成高保真的自然色影像。

While multispectral remote sensing image pixels or spectra library object samples were used as samples to simulate the spectral relationship for the multispectral remote sensing image,the simulation model for the blue band and finally natural-color image are usually not stable and the accuracy is difficult to improve. The reason is that such samples are not consistent with the target image in temporal, and also the samples are atmospheric affected. In this paper, the MODIS ground reflectance product （MODO 9） was used as reference image to select samples for spectral coupling simulation and natural-color image simulation. Firstly, with the MOD09 samples,a non-linear regression analysis models,a back-propagation artificial neural network （BPN）,was used to fit the spectral reflectance relationship among the blue band and red, green, and near-infrared bands. The blue bands of the multispectal remote sensing image were then simulated with the trained fitting model, and a natural-color image was output. The experimental result shows that the BPN model well simulated the blue band of a multispectral image and even removed the atmospheric effect of the blue band. With the simulated blue band,a more realistic and informative natural-color image was acquired.