遥感卫星多光谱图像模拟真彩色算法研究

Method of Generating Pseudo Natural Color Image from Multi-Spectral Satellite Remote Sensing Data

地球资源遥感卫星ZY-102C和CBERS-04搭载的全色多光谱相机缺少蓝光波段,需要通过波段模拟方法首先生成蓝波段,进而合成模拟真彩色图像。根据多光谱数据特点,提出一种基于波段线性关系模型的模拟真彩色方法。首先用具有蓝色波段的多光谱数据作为参考图像进行回归分析,计算得到相关系数,将结果作为参数应用于待模拟图像生成蓝波段。然后按照地物类别进行单景图像局部区域颜色畸变校正。最后按特定顺序合成模拟真彩色图像。试验表明,该算法简单快速,减少了偏色现象,能够获得较好的模拟效果。

Due to the lack of the blue band of ZY-102C and CBERS-04 multi-spectral data,we need to simulate the blue band before generating the pseudo natural color image.According to the characteristics of multispectral data,we present a method for generating pseudo natural color images based on spectral correction with the linear model.Firstly,the multispectral data with the blue band are used as the reference image for regression analysis,and the correlation coefficients are calculated.The result is used as a parameter to generate the blue band for the image to be simulated.Then,the color distortions of local areas of the single image scene are corrected according to the features of ground objects.Finally,the simulated natural color images are synthesized in a specific order.Experimental results show that the algorithm is simple and fast,and can produce better simulations by reducing the color deviation.