多光谱彩色合成图像是直接提取多光谱图像中的R、G、B波段合成的彩色图像。但由 于多光谱成像探测器光谱响应性能的差异,造成合成的彩色图像偏色严重,为了重现真实色彩,需进行颜色校正。针 对目前传统白平衡法与色度学合成法的不足,提出了一种基于光谱信息的改进白平衡法,用于多光谱彩色 合成图像的颜色校正。经实验验证,该方法不仅实现简单,校正过程中不需要标准真彩色图像,也不需要 预测图像中所有地物的光谱反射率,而且改进了传统算法,获得的图像更为真实。

Multi-spectral color synthesis image was extracted directly from R, G, B band of the multi-spectral image. However, due to different performances of the spectral response of multi-spectral imaging detector resulting in the color cast of the synthetic color image, in order to reproduce true color, the color correction was needed. For the lack of traditional white balance method and colorimetric synthesis method, an improved white balance method based on the spectral information was presented for color correction to multi-spectral color synthesis image. It is proved that this method is simple, does not require the standard true color image in the correction process, and does not need to predict the spectral reflectance of all features, but also improves the traditional algorithm, making the images more realistic.