高光谱图像的同步彩色动态显示

摘要

The spectral weighting envelopes using the color matching function are the current mainstream method of obtaining approximate true-color images for representing hyperspectral image. Meanwhile this method has attracted widespread attention because of its easy computation. An improved method is detailed in this paper. This method changes still images into dynamic images by adding time dimension so that observers can obtain as much information as possible within the same amount of time. This proposed method can obtain approximate true-color images, and this displacement cycle weighted method can also enhance the spectral resolution of the images, so different ground objects present different colors and different rates of change to improve the visual effects. Simultaneously, this meth-od also satisfies the design goal of easy computation, and would be appropriate for real-time observations of hyper-spectral imagery. Experiments with the real ground object data show that the color images produced by the model have good visual effects and good separability, and the method adapts to the hyperspectral imagery′s visualization.%高光谱数据的CMF加权封装方法是目前由高光谱数据获取近真彩色图像的主要方法，同时也因其计算简便而受到广泛关注。针对CMF图像分辨率较低问题进行改进，加入时间维，采用位移循环加权方法，变静止图像为动态图像，建立了一种适用于高光谱图像的同步彩色动态显示模型，使观察者在相同观察时间内获得尽可能多的信息。实验表明，生成的图像不仅能够产生接近于真彩色的图像，而且使得不同地物呈现不同色彩和不同的色彩变化率，增强了视觉感官效果。同时该方法也满足计算简便这一设计目标，适用于对高光谱数据进行实时性观察。