通过对实际获得的多光谱影像的光谱特征分析,提出了可用于中巴02B卫星多光谱影像的LBV数据变换公式,使得LBV数据变换方法在国产中巴02B卫星数据上的应用成为可能;运用该变换公式处理得到的中巴LBV数据变换图像比中巴原始数据假彩色合成图像颜色更鲜艳,地物类别更易区分,具有更好的目视解译效果。将LBV变换图像与中巴数据假彩色图像分别用最大似然法进行分类,分类图像和精度检验结果表明:用LBV变换公式得到的LBV结果图像能很好地提高图像计算机分类的精度,该变换方法在中巴02B卫星数据的应用中具有很大潜力。
LBV is a new data transformation method which is recently proposed for remote sensing dada transformation processing,however,it is can′t be directly used to CBERS-02B multi-spectral data since there are relative transformation equations data until now.Based on detailed ground spectral features studies of multi-spectral images from CBERS-02B,new LBV data transformation equations for CBERS-02B multi-spectral image were specially proposed.These new LBV transformation equations made it possible to use the LBV transformation method on the domestic satellites and provided a new method to process CBERS-02B multi-spectral data.Moreover,image transformation and classification experiments were carried out,which results showed that the LBV transformed result images were more vivid and the features of them were more easily to be classified compared with the original false color composite images.Finally,in order to evaluate the performance of this proposed transform method,the maximum likelihood supervised classification method was used to the LBV transformed result images and the original false color composite images,which results demonstrated that the accuracy of the LBV transformed image was obviously better than that of the original false color composite images,which showed that the proposed LBV transform method has good potential for CBERS-02B multi-spectral images processing in the future applications.