基于多光谱图像选取马铃薯叶片的最佳彩色波段组合

Selecting best color band combinations for potato leaves based on multi-spectral images

为了快速目测识别马铃薯叶片,利用Spectrocam多光谱相机获取健康C-88马铃薯叶片的可见光及近红外通道的多光谱图像。采用波段指数法提取叶片的特征波段,通过真彩色原理及标准假彩色对所提取的特征波段进行彩色合成,在此基础上得到最佳彩色波段组合。实验结果表明,用波段指数法提取多光谱图像的特征波段进行彩色合成,能快速获取马铃薯叶片的最佳波段,并得到680,558,475nm和800,680,558nm为最佳彩色波段组合,为遥感图像的目视解译和更多有效信息的提取提供了可靠依据。

In order to identify the potato leaves rapidly, we use a Spectrocam multispectral camera to capture multispectral images of the healthy C-88 potato leaves in visible light and near infrared channels. The characteristics of the leaves are obtained by the band index method,and then color synthesis is carried out based on the principle of true color and false color. Using the above experiments,we obtain the best color band combinations. Experimental results show that extracting the characteristics of the leaves through the band index method and then finishing color synthesis can quickly yield the best color band combinations. The bands 680,558,475 nm and the bands 800 - 680,558 nm are the best color band combinations. The study provides reliable criteria for the visual interpretation of remote sensing images and the extraction of more effective information.